I thin a lot of us have this issues, especially on side projects. At work therre are expectations, jira tickets and so on, keeping you shipping. I think there is something to be said about getting that quick and dirty POC to the right group of people early for feedback before you add redis caching, kubernetes, auto scaling, disruption budget, distributed nodes, high availability, backups, disaster recovery. At work you kinda have to have the right person to shoot ideas by that can understand that you probably need some of these complex things for your app and it will take time to get right.

I would love to have a browser based video editor I could throw on a server and do quick edits from anywhere. I tried to get this one to work and struggled to get front end to send api requets to backend. I think the root of it was their redis wants to run on 80, this caused a permission error so I tried to run 8880:80, but redis was still unable to start due to a config permission error.

The ability to query s3 buckets so seamless looks like such a pleasure to work with if you have a use case for that. Kedro catalog takes care of this most of the time for me, but I wonder if there are some cross project searching use cases I might find for this.

I love this idea of tiny useful apps for yourself. In fact I'm working on a project to built out [[ tinyapps ]] for myself to replace my common needs. I absolutely love that all of the state is stored in the url bar, nothing is stored server side. As much as I love to hate js, I really appreciate that things like this can be built to just live on the web, be accessible from anywhere, and live practically forever as they require such little hosting demand.

This looks like a very useful formatting tool to keep in the back of my mind. I do a lot of python and our tool tends to be pre-commit, named after the git hook pre-commit. It specifies a bunch of tools to run, you can run them in ci, manually, and opt into doing it before commit. I like the simplicity of this one not needing a whole ecosystem, but rather just leveraging the cli commands from those tools. This would probably be something that would get in the way of setup for new devs and not something I would throw on one project by itself, its another thing for everyone to figure out how to install and run on every platform, I'm sure its not hard, but being on python teams pre-commit just fits in.

This is one of the greatest pycon keynotes I've ever seen, bookmarking this to come back and leave better thoughts on later.

Focus on the joy, not the suck. Nothing you do in life will be absolute pure joy with no downsides forever, life does not work that way, your brain does not look that way. Look at anyone who ever got massive billion dollar payouts for something like minecraft and how much their life is not glorious when they have nothing to really look forward to.

Prime talks about it in almost a cliche way, every boring ass task is an opportunity to grow. This is so real though, if you look at every task ask a shit you gotta do to check that jira ticket off and make bossy lady not scream at you its going to be a hell. If you rather look at it as opportunities to implement new features in new ways or learn something to better yourself and watch yourself grow you are going to take a big dopamine hit. I think prime talks about this in the sense of larger projects. He as talked about his experience being much less of a daily standup, but more of a ok we got three months to figure this out lets go boys. When you are stuck in that daily jira grind it's harder to see that larger picture of the learning and growing you are doing over the course of 3 or 6 months.

Timestamped to the part of the video that really dives into gratitude and finding the joy in the work that you have to do on purpose, but the whole vid is fire.

Should I go to college? Was my education worth it? Should I keep going. A question that comes in all too often accross most industries that require some level of education. DHH has such great takes on it, some I had never fully thought about. He starts out with should we have people study niche topics (using Russian Poetry as an example). Yes the world deserves people who can make their life works out of something that brings them and many other so much joy, but no you probably shouldn't go 100k's into debt to do it. Should I get a software engineering degree, or become a doctor also have similar answers, it needs to be somewhat justified and not outrageous as has become the norm.

We used to listen in to Dave Ramsey on long car rides and he would have people call in and say, they went half a million dollars into debt to become a dentist, only to discover they did not want to do dentistry. At this point it's too bad, you gotta suck it up and pay that off with something that makes some serious cash, and the only skill you probably got that can bring in that level of cash is ... dentistry.

They dive into the college experience, learning to have adult debates with classmates about which direction a project should go or how to complete it. Great stuff!

Something I had never put thoughts to is how do you feel when class is cancelled or an assignment is removed. Are you excited to "not have to do it"? I'd bet that most people do. That's probably a good sign that you are there for the wrong reasons. What other industry do you pay a f*** ton of money for a product and be like nah ill just take the receipt, you keep the thing.

This all just gets your foot in the door. If you spent none of that time learning it will become apparent really fast and you will not climb the ladder. Yes climbing the corp ladder is cliche, but at some point you probably want paid more and get opportunity to make more decisions. And if it turns out you really suck, you're going to find yourself looking for a new job with no skill and a big pile of student debt sitting behind you.

Student debt is probably one of the biggest issues of our generation that only appears to be getting worse, everything signals your need to go to college, presige is important. It feels like the other side is filled with glory lots of money and you will have all that debt paid off in no time. I get it when you're 18 and you see 50k or 100k in student debt, but 80k or 150k salaries on the other side its like damn I make nothing right now I'll have that paid off in no time. Then you get out and realize everything costs serious money, housing, kids, cars, broken AC units, and you cant pay off 50k in a year and a half making 80k.

webtui, looks like a pretty sick design aesthetic. I like the keyboard driven nature of it, the look and feel is on point to a terminal interface, sadly it looks like it is not a 2 way street, you don't automatically get a tui our of your website, just one that looks the part in the browser.

I've never heard of niri, or a scrolling window manager, it looks quite interesting. I think tiling window manager misses out on named sessions and hotkey straight to tmux sessions, Brodi mentions not using tmux right before this segment. Niri looks quite interesting, but looks like it suffers specificity. maybe there are other tools that allow me to jump straight to something like brave, or steam, but I don't see how I could jump to a specific terminal.

I'm way behind on my notification game and need to pick it up. maybe I'll look into hcio as well. maybe I'll look into something that goes straight to signal or just get things working on ntfy. An 80GB log file is massive and the kind of thing id like to see notifications more.

Taking this as inspiration to do more non-tech on my blog, I've branched out into [[gaming]], but need take it to the next step. excited to watch pype.dev evolve as well.

I've long avoided running ads on my blog for the same reason. For a few months I ran an ad above the fold. It was a "Your Ad Here" kind of thing, and in the messaging I was looking for content relevant to my content, not google driven ads. This resulted in nothing, no hits, not a one. I'm kinda with Will on this one beer money is not worth degrading the project for. I seriously thought some of the big projects with a moderate level of success got a good cut for these sponsorships. Some of the companies are big companies, like how do they even go through meetings and decide who gets beer money without spending more than that in decision making resources. Maybe they have a guy with more autonomy than I would expect.

Browsing for the minio tag that I have running right now I discovered that you can do minio --version and you get the same version that matches the docker tag, this is super convenient and helpful. I also notice that they use timestamped version numbers. I kinda dont mind this. It feels easy to understand how far behind it is. I really appreciate that the version in the container matches the version inside the container.

It's not as pretty or flexible as semver, it does not communicate trees of majors and minors, but how often do we continue supporting/patching older majors and minors, in my experience only really big teams or teams with sufficient motivation are doing this.

food for thought.

I am going to start trying to employ this rhythm to my writing. I'm not very sure how I feel about it, there is something almost too assertive about it. It's giving me a (i'm great and you should too) kind of vibe. I want to become more assertive in my writing.

I'm giving this a shot and see what I learn, you might notice in my tils.

This talk about live store really made me think about database transactions in a new way. They are talking about live-store, and the complexity of distributed applications like a notes app with the ability to go offline and continue working. The complexity of resyncing each instance is not simple, conflict resolution accross all the possible installs that may or may not even be online is a really hard problem. They go deep on discussing an event driven paradigm that is driven off of a log of events and how this changes how we deal with databases. Using the event log as the source of truth we can do things like forget about database migrations, we can replay all of the events onto a new database. Its very interesting to rethink in terms of a log system that speaks in terms of understandable events (not table operations) as the source of truth for an application.

I need to find this podcast, was DHH this animated through the whole thing?

You don't need a mentor. There's no secret sauce left inside anyone's head any more. It's all been tapped, bottled, tweeted, and shared a million times. Sample some of that, but also guard your ignorance. You'll lose it soon enough.

It takes work, one on one hand holding is a shortcut. Sometimes one that we need. Sometimes we need to level up quick, hence why your job might pair you up with someone for the first few months, but it is not something you need, you can figure shit out on your own with hard work. These days we have things like gippity to bounce ideas off, and you can generally get the sense of the direction the average of the internet it was trained on. Always add your own experience and make a choice for yourself.

The object storage (S3-compatible) platform MinIO created a bit of a stir this week

I had not heard about this before it came in through selfh.st. I use minio a lot, and did not know there are so many great alternatives out there for it. I might be looking into some of these options such as garage.

Its hard to tell from this article what mino dropped, but luckily for me it seems to be all ui related. I use the UI for debugging/feedback/sometimes learning, but at this point I've got good flows for setting up new access keys, buckets, and everything with the cli.

I suffer hard from NIH, I'm cheap, I like building things, I hate reading the docs, the perfect recipe for some bad NIH. I really like DHH's take here. If no one builds anything new we get stuck with the same old shit. I think theres a lot of things that as far as my use case is concerned feature complete and needs no more. I would just build with it or on it, but not re-invent. It's a slippery slope.

Jim Nielsen fetches his hacker news ranked articles for his home page.

this post by Jim Nielsen, lead me to this commit where I found that he was including posts of his that wound up on hackernews. I really like this idea and might take it, even though i have very few HN linked posts.

I really like the idea of Jim's Eternal Links, and really want to take it for myself. To expand here I want to be able to look for common places for rss feeds, and be able to scrape out rss feeds for sites that I tend to link to often. Also if they have something like a /blogroll it might be a good place to find new great people to follow.

Maybe we need a little more friction in the world. More things that merit our time. Less things that don’t.

I can resonate with this post, less friction feels like it leads me to thinking less, having less skin in the game, understanding less, feeling less fulfilled. Vibe coding is a new trend of 2025, it feels like the future, but it does not quite feel like the present yet. It's riddled with errors and I only get frustrated when it doesn't work. I like having some friction that leads me to think and pay attention. There might be a future where this is not required for some things like coding up crud apps, but that does not feel like today.

Some of the best things from the old internet are still preserved with RSS. Content is shared via simple files, which means the slow-loading, ad-stuffed and tracker-filled clutter of the modern internet are mostly absent.

There aren’t any algorithms. RSS readers are wonderfully dumb. There’s no AI sifting through content to find whatever will outrage you the most. You just get new posts and mark them as read. It’s a calmer world.

With RSS I follow lots of people writing about normal people things. People blog about getting back into playing the drums, a fun book they just read, a tough problem they’re working through and the other day to day things of life. This type of content tends to get buried on social media — it doesn’t get the clicks and sell ads like fear and outrage do.

I feel like a curmudgeon, but i feel all of these things. I dont think that the new web is completely terrible, what is terrible is that the options of an algorithm ran by companies with differing goals is seemingly the only option. RSS still works, its fantastic, I personally love it, but theres only a small fraction of the internet that it reaches both ways. Few people have a reader, even fewer have a feed.

This is a very interesting cli, its so simple. I stumbled accross the gi command awhile back and was like pfft, I dont want to install something for that. Didn't even realize that you don't install it, its just http. Their install instructions lead you to putting a curl funtion in your bashrc.

        
function gi() { curl -sLw \"\\\n\" https://www.toptal.com/developers/gitignore/api/\$@ ;}

This now has me wondering "What else can build like this?"

linkarzu has a way to navigate his entire mac using a hyper key. Everything looks so tight and polished, also a lot to remember! Lucky he has a system of mnemonics that make it easy to remember. His setup is very Mac focused using mac only apps, so this would not work for me, though I'm sure I could get something similar on linux. He did mention Kanata which is cross platform.

What I do

I use a far different system that is fast loose and easy. On every system I run I have 9 workspaces that let me put 9 applications, I can easily move apps to different workspaces and have a side by side if I need. The core of what I do is terminal, web browser, and chat. Those go on workspaces 4,5,6, whch are home-row keys. If I'm running obs, that is on 8, steam goes on 1. but I have some freedom to move. Sometimes 2 will be an image editor or a video editor, sometimes something else all together, but I can quickly go to each app.

What I like from Linkazru

I do like his layered approach. I run a 42 key keyboard so things can get a bit cramped quickly. And when thinking in mnemonics you only get 26 letters in the alphabet, but prefixing these with another layer this number goes up exponentially. Sublayers sound sick to be honest.

Oh, I feel this. I go through the effort of removing dum ai comments so the ai looks less ai.

you're not allowed to write comments in your code anymore, because if you do everyone will just think it's ai generated.

This has me wondering if I need to really learn more patterns, data structures, and algorithms. This looks particularly useful when trying to combine several objects that you dont have full control over and make them behave similarly.

wow looking at how this is done kinda draws me towards jekyll a little bit, I did not realize some of the similarities that it has with markata.

css if() just landed, I'm struggling to understand what I an do with this that I can't do with something as old as classes. I can get it if I don't have control over html creation or js to add classes. The example that Una shows includes data that could directly be a classname with a set of styles in css rather than this crazy css variable unpacking out of a data attribute and an if statement.

Allen Carr1 on quitting smoking: [Carr] recommends working to really notice and internalise that disconnect [between what we want and what we enjoy]. He tells smokers to pay attention to their next cigarette. It’s like mindfulness but for noticing the unpleasantness.

I can appreciate the restraint here, theres something about the mindfulness behind it all.

I rolled out the blogroll today, nothing pretty, but is one single page of the rss feeds I follow.

Markata got a shout out part way through the latest episode of LNL, I will go back, re-listen and take some of the feedback. His thoughts on Markata were interesting. On one hand it really is a thing for me that works for me, and as a person with too many side projects I don't have the focus to really give it polish. On the other hand it really confirms why listen to podcasts, news, finger on the pulse, opinions and how often these guys are wrong, they are not the expert they probably look at 6 things like this a week. He said that it was some sort of javascript thing, that maybe he could fix or customize with javascript if he wanted, kinda shocking, I thought maybe I accidentally added node modules or something dumb, nope, I have a whopping 1.4% js. So most of the comments were plain wrong. I get it he probably peeked at it for 30s and realized it wasn't the thing for his problem. At the same time I should probably do a better job at marketing what it really is, cleaning up the docs and demo.

Such a great message right now. I feel like everywhere I turn is negativity, especially social media. It feels like so many things are trying to divide and create hate. "This" is what we should be doing with social media. There are a lot of elements of "there are two ways to have the biggest building in town, tear down all the bigger buildings, or just build the biggest fucking building", If you want to be successful in X then surround yourself with others successful in X. This is a catalytic skill that everyone needs to have in their belt.

Great conversation with Billy Basso the creator of Animal Well on the code architecture of Animal well. It's all hand crafted C++. He talks about early games he tried to build being heavy in oop, and really got lost in oop. Animal well is very flat, there is no inheritance, just lists of entities that all implement similar methods in their own way. Layering and order of entities becomes very important. Its crazy how much he had to think about hardware and MS build being very helpful with this, but needing to know all of the console apis.

This is great, beautifully captures a modern backend view of https://motherfuckingwebsite.com/. I honestly resonate with almost all of this. I have found myself in more trouble than help when trying to fully vibe out a project. It never refactors, it leaves it shit everywhere, it mostly does what you say, until you get to something that seems easy, so you try to do it yourself, but you break its brittle piece of shit into pieces any time you try to touch it. AI coding help is great, mcp seems like it really has some game changing abilities, but hands of vibe coded crap aint there yet for me.

I did not realize half of this, and it took me at least 4 watches through this to catch everything.

k8s-monitoring requires setting imageregistry and pullsecrets twice

        
global:
  image:
    registry: my.registry.com
    pullSecrets:
      - name: my-registry-creds
  imageRegistry: my.registry.com
  imagePullSecrets:
    - name: my-registry-creds

Bugs exist when your test-enforced implementation fails to match the behavior described in your documentation. Without documentation a bug is just undefined behavior.

This is quite an interesting thought, so does this mean that, none of my undocumented side projects have bugs? no I think there is still some implied behavior that naming things covers. a function get_bucket_contents implies doing something wtih s3, getting stuff from your local filesystem or crashing would be considered a bug. I think the argument here is that if I start mining bitcoin when you call get_bucket_contents and I have not documented it that this is a feature not a bug. If I were to take this a step further, now do I need to document that this does not also start a bitcoin miner? maybe this is more of an unwanted feature than a bug, I'm convincing myself more and more.

So many small details go into making hollow knight such a great game, but it starts with such good controls, every thing is so fluid and predictable. I knew about coyote time, but not some of the other details that Juniper covers, such as hang time, and faster decent than jump.

Excited to hear this story, they have so many key players in the trailer this will be fantastic.

Interesting take on kubernetes from a front end perspective. All valid arguments to me, and really the answer to any do you need to any specific implementation of tech is probably no. We got along just fine before k8s ever existed and you still can, but its really nice in a lot of cases. If your skills lean toward backend or infrastructure I encourage you to give it a try.

k8s distros

There are a lot of beginner friendly k8s distros that you can setup with relative ease, kind and k0s are great for single node, If you want multi-node k3s is what I generally use. If you want a very lightweight OS that you only interact with through an api, and has a very small attack surface talos is an amazing product.

When else might you want k8s

Internal, on-prem, self hosted. If you are trying to avoid the cloud for cost, rules, regulations, red tape, kubernetes is a great option to manage your container workflows yourself without needing to have a cloud budget, get approvals and sign offs on running workflows in a public cloud.

Just listened to this as I am really starting to get into grafana and feel like there isn't a mountain of setup this time around realizing how much of my stack is brand new. Drill Down and Alloy are both new and key to my setup. The Ai integrations at the end sound wicked good, I will be interested if you can do similar things with an MCP vs how much proprietary stuff needs grafana cloud.

Ultimately though a business needs a product. Textual has always been a solution in search of a problem. And while there are plenty of problems to which Textual is a fantastic solution, we weren't able to find a shared problem or pain-point to build a viable business around.

I can totally see this. Finding a marketable business idea is not easy, working in the developer space where everyone wants to do it themselves is no better. Textual specifically I could see, I really wanted to build things on it as it came out, I had ideas, it was hard to use at the time and changing, so I took a break, got busy with far too many other things, and really I 'm good with rich most of the time.

I daily use k9s, its absolutely amazing at what it does and appreciate that I could build something like it in python, its just hard to justify the time investment for the things I tend to work on.

Which is why Textualize, the company, will be wrapping up in the next few weeks.

Damn, that hit hard, its been an adventure watching textual get built, things like posting look fantastic! Glad to see that its still alive with open source and Will is not leaving, but there will be no company. This likely means significanly less velocity, but it might mean that not being tied to making money other opportunities open up.

Thanks for giving us amazing terminal products Will.

So it’s back to plan A: taking a year off. I plan on using this time to focus on my health–something I haven’t prioritized while working as a CEO / Founder of a startup.

Wish you the best Will, you have given us textual and rich, and from what I can tell left it in some great hands.

All I can say for certain is that I would like to write more. Writing scratches many of the same itches as software development, and it is a skill I’d like to nurture.

Go get em Will, write to your hearts desire, and resist the urge to make an SSG company this time.

Now “too much magic” is not the same thing as “bad magic”, although they are often conflated. Bad magic is when the implementation details leak out from the level below. This can manifest itself as cryptic errors that reference the magic’s implementation.

Dang Strong takes against markdown here with a strong push for bespoke content models/structures. This idea is completely foreign and wild to me. I get it that markdown has its issues with flavors, add ons and what not, but overall its mostly transportable, its a skill that works most content sites and writing tools. I am so far on the other side that I seek out tools with markdown as an option and lean away from wsiwyg tools with specialized data formats on the backend.

I'll end with, I'm also a dev that creates very simplified content and maybe seeing the backend of a site with lots of custom fields would be very eye opening for me.

Last year I attempted to do some newsletter-style link aggregation… that good intention imploded spectacularly. But I switched to Obsidian this month and now I have a better system for aggregating links (post on that coming later). Inside this issue you’ll find some games, some homelab server hardware, some AI discourse™, some musical instruments, and more.

This hits so close to home, I even went through the effort of making a weeknotes script, one weeknote post. I also was inspired by obsidian but it didn't work out for me, so my script uses data from markata.

Do you remember regression models from college: given some data, you find a best fit line that allows you to predict Y given X. At the end of the day, ChatGPT, and LLMs in general, are the same thing as the regression model – it’s just that ChatGPT is the largest and fanciest model we currently have to model language and information.

I really am coming to the idea of calling it a "word calculator", this seems to be the most succinct description of llms that the lay person can comprehend and relate to.

ChatGPT does not hallucinate or become unhinged

I think Steve goes much deeper on this in his intervew on fafo.fm. They describe it more as a pleaser or "yes man" essentially all the companies that are building these models want to give the "best" answer, better than their competitors. With this comes the risk of it being completely wrong, they are designed to always give an answer.

On that episode they talk about how each word has a statistical probability tied to it and in some visualizations you can see the confidence, in almost every example of "hallucination" you will find some set of very poor confidence followed by BS. Thats not to say that every time there is low confidence there is BS though, that is why this is such a hard problem.

ChatGPT is not a replacement for search

I've become really bad at this. This is slowly becoming my starting point for a lot of research, not my only research, but the first wet finger in the air. Most often done with low effort, low risk tasks.

In ChatGPT, just like in the shelf full of the Encyclopedia Britannica I had when I was a kid, information on topics has been reduced to good summaries and presented in logical format. In this case we have removed the human authors and replaced them with a very large neural network.

I really like this description and it matches up with my last comment.

I'll triple down on the link-blog chain here, see this one going around all over this week and finally had time to read through when it hit my rss reader via Chris.

It should come as no surprise that nearly every vibe-coded app on the Internet struggles with security issues; look no further than the vibe-coded recipe app that leaks its OpenAI keys. Every time one generates code by prompt, they create a new stillborn program; vibe coding is the art of stitching together their corpses into Frankenstein’s monster.

Damn, that is a strong statement, stitching together the corpses, strong statement here. The OpenAI key thing feels kind of obvious to me, every set of docs, blogs and examples on the internet need to be runnable for people to learn and try out new tech easy, putting secrets in the wrong place is easy, putting them somewhere that you can decode them without sharing them is hard team specific, app specific, and so nuanced to your architecture that its rarely included in public examples. I imagine there are a lot of good code example out there that follow good practice, but it feels like that might have missed the training data here.

It alrso reminds me of thought-625-the-rule, where prime talks about small apps being happy and easy and at some point every change is hard, and likely impossible to vibe code.

The whole point of making creative work is to share one’s own experience - if there’s no experience to share, why bother? If it’s not worth writing, it’s not worth reading.

llms are not capable of being creative, they are word calculators, statistical models that predict the next likely word based on their experience, they are not capable of new discoveries.

Under 2000 everything is happy, green field. Any decision you have made is relatively easy to back out of (barring you making a library with downstream users), but as you go, regret kicks in. Regret we didn't make that pydantic 2 upgrade earlier, as new features become more apealing. Regret that we chose sqlite for simplicity, speed, agility, and now we might need robust and distributed. Regret that you chose a front end framework, or to have a front end at all to a backend problem. Regret that you put 6 layers of abstraction on your db early on and now that you understand the problem you want different abstractions, but all of your endpoints deeply depend on the current one.

Vibe coding will not save you, it will only make these wrong decisions for you without the context that you have. You will hate it's decisions more because you had no input into some of them.

"Gradually roll out your releases to a small group of people"

~ roughly what prime said (I'm listening live)

This really hit home with me, tests can be so good at making sure that we dont repeat bugs and that laser focused things work, tests are generally small and focused, but this does not replace some sort of integration testing. These days very few things are written as a monolith, and hence there are a lot of interactions that really need to play well together accross various systems.

They call out Crowdstrike here, which took down the world blue screening critical windows systems everywhere in 2024. It was revealed that a small changed was rushed through and skipped critical rollout paths since it seemed like a small change. Crowdstrike also runs at a super low kernel level of access and a small memory bug can kill the system.

I'm trying to level up my sre game. I'm trying to set up grafana dashboards for everything and it is such a wide surface area. It's never just one thing you have to have 3 or more things hooked together in order for the data to flow.

I'm really getting not invented here vibes, and thoughts that I can just build this myself. Not grafana and it's scalability necessarily, but small components of observability.

This is a really great guide to setting up kubernetes monitoring with helm, it uses loki as a log datasource and alloy as a collector of kubernetes logs, events, and nodes. The charts are setup really well to start collecting logs from all your kubernetes pods.

Really helpful article to getting tempo datasource setup in grafana, this enables you to see span and trace data within grafana. This data helps debug and work through issues that you might come into with performance and need to see the timing of requests along with logs.

Steve is such a great listen, the neurospicy 🌶️ rambles this episode goes on is so relatable. I feel like I really missed out on some great takes on intellij vs neovim, but got some really great knowledge about vector db's, embedding, text compression, similarities to vector algegra like infinite craft.

Just popped open infinitecraft and I've definitely played this with my kids before, super fun, just could not remember the name of this one. I do remember an android one as well that is alchemist or something like that, which we have also played a lot.

This episode really got me thinking about the difference between HA and DR and my approach to each one. They talk about it from the perspective of a cach cow kind of app rather than a homelab or internal tooling, but think of HA as 9's how many 9s are we willing to pay for, tink of DR as dollars how many dollars will we loose during the period of recovery. So much more in the episode, a lot of talk around cloud vendors and what they give you vs a purpose build platform with HA and DR in mind.

Astral is doing great things in the python industry. They are disrupting entire categories of tools with extremely fast, easy to use, and feature rich alternatives that make it really hard to keep using the incumbent. So far I am seeing no signs of evil, sometimes with such a disrupter there is some sort of downside that make it hard to want to do the switch. In the interview they even mention things like leaning on lsp so that it works across all editors rather than building out vscode integrations that work for most developers. As a neovim user I greatly apreciate this.

ty, has a playground running at types.ruff.rs. You can edit code in there and see what the type checker results would be in browser. This looks good, excited to see it running in my lsp.

Here is an example where a Optional may not be defined.

Checking for existance before using it resolves the issue.

Astral is working on some great things around python, they have created a high standard for python tooling built on rust that works really well, runs fast and covers everything in the space it resides in. ty appears to be their linter coming soon.

This is madness that Wes Bos made this with manifold.js and no openscad! Yes, I have these stupid brackets everywhere, yes, I hand model my own brackets. No I don't do it enough. I don't like that these model generators like openscad cannot make fillets and chamfers, but I appreciate the heck out of the speed and automation you can make iterations of things.

Link to the promo video. https://bsky.app/profile/wesbos.com/post/3lo4h7unk6s2i

Chris hit me where it feels about 10 minutes in. He said he has not been writing on his site as much lately and how hard it is to get back in. He mentions having a baby idea of a post, but then having the thought do you really want to come back from a long break with this!

Momentum is a b**** when you got it you cant stop, and when you don't you can't stop.

How is usability and it doing the thing I paid for it to do a selling point?? Any time I've touched a windows machine in the past 7 years has felt awkward, I have no idea where things are now, but they look so much worse.

How is usability and it doing the thing I paid for it to do a selling point?? Any time I've touched a windows machine in the past 7 years has felt awkward, I have no idea where things are now, but they look so much worse.

Interesting how confidently he says we can easily go to the top. really makes you wonder what we the normies are leaving on the table by using these general purpose models and what could be achieved with really tuned in models. Could I make an automatic blog tagger more accurately, maybe smaller, maybe tuned so well it runs fine on cpu?

The web is everywhere, its the one true write once and run anywhere platform. Millions sunk into browser performance and things like the v8 engine allow us to run our shitty websites anywhere and it still runs good.... most of the time

I didn't realize that postiz had a helm chart, I just hand rolled mine based on the compose file they provide. I went from running the compose stack locally to running in my homelab with kubernetes. I am using cnpg rather than a postgres container which I really like the workflow of as far as backup and restore. The one hiccup I ran into was changing the domain from localhost to my homelab domain killed all of my integrations and they needed the redirect url updated.

Great talk from Lous Rossman! TLDR you don't own it, and stop pointing the finger calling everyone else an idiot for supporting the other brand, cause your's probably also has different issues.

This is a wild concept for a slicer, essentially he didn't even make a slicer just a crazy pre-process and post prossess to cura slicer, deforming the part until it doesn't have any overhangs, creating a normal planar slice, then undeforming the output from cura. He also mentions that the rapid moved needed modified as well. I'm assuming this is because they are generally long distances and not short, without breaking these long lines up we would still end up wtih a straight line after deform.

Intersting markdown presentation tool, Looks very simple. I really like split on --- much better than by h1 or h2. Their theme looks really nice in the screenshots.

How to make an entire clickable without presenting the entire content of the card as the link title. These videos are great, I've ran into these types of problems so many times, and definitely did not know about things like isolate to keep the z-index scoped to one element.

• isolate - scope z-index inside this element so that it does not leak out.
• [.relative [.absolute, inset-0, z-10]] - the inset zero is a modern shorthand for zeroing all sides, top-0, right-0, bottom-0, left-0.

This is an absolute banger of a review by prime and Dylan Beetle. I love the similar takes with different perspectives, would really like to see them podcast together, but this one way style interview does really well to cover a lot of issues in open source, rug pulls, version pinning, thankless maintainers, what its like to open source from a large company.

Interesting takes on Diun here. I agree that I like to be in control of updates and pinning not to latest. both seemed like they weren't going to run it because they can look up the latest version. Maybe I need to be less aggressive on keeping things up to date and its a me problem. I just got diun setup and hooked into ntfy, and I kinda like the automated checklist of new images that I can review and update.

To be a bit more clear, having control over changes coming in from others, even if I dont care to see the changelog, it is nice to roll out an update, have it in your git history, watch it deploy and work like before, if not roll back and read the changelog. For internal applications I'm down for automated releases like argo image updater give you, this thing has already gone through review, launch the damn thing at least to a dev space.

Davids blogs always have so many links that send me down new rabbit holes. Interesting that his experience with smart home is turning away, I've been somewhat interested for awhile, but never fully pulled the trigger on buying things.

I really hope tailscale enshitification does not take off, but really for me, I barely use it even as a homelabber. Idk why, but every other homelabber praises it so much and I just dont find myself using it.

xeiaso, has the coolest characters on his blog. Definitely something I'd like to replicate. I really appreciate how each one has its own sprite sheet, and they have conversations with each other.

I want to go build these now, these are suspension setups I had no idea you could do in trailmakers.

keeping this in my back pocket for now. I just moved a few TB's of data in the homelab and I am expecting a lot of duplication to show up.

Diun, looks like a very interesting tool to monitor for image updates, it does not make any change, it only makes notifications. This feels like an easy start to getting image updates started with low effort, keep git ops, but requires manual updates. I see this as a tool that would be a great start and pair well with automated image updaters to ensure they are working as expected.

Keel looks interesting, I might give it a try as a simple image updater. I'm unsure if it fits my gitops patterns though. I like to keep everything defined in git, I don't like drift outside of that so Keel might not be the thing I want.

Damn he makes this easy. I did not know about hx-select. yes there is waste in requesting the entire thing every 5s, but damn that was easy to get life reload. I've only done very specific backend endpoints, built pages up from partials, made endpoints for partials. keeping this one in my back pocket.

I'm just kind of amazed that he could do this all in html without touching the backend or js, typically things like this require one or the other. Yes js is running, but no other js library I'm aware of lets you do this.

This is a handy line to reset your admin password in nextcloud.

redis has all of their default self documented configs hosted here. You can pull the default redis.conf for any of the major releases.

nice dataset to use for example / test projects. I'm using it to play with duckdb currently.

Good report, make notes later

looking into trying these Mill-Max pins on a handwired 3d printed build to see if I can get away from specialty hot swap sockets. Damn they aren't exactly cheap, I really want the nice short ones but they start at $20 per 60ct and you need two per key, that adds up quick.

jina reader is a pretty sweet tool to convert a site to ai compatible text. There are other web to markdown types of tools, but the convenience of just adding r.jina.ai to the front of any page makes it so easy to grab for one page of docs.

the racked up 4 framework mainboards sound wild. connected with usb4 and 5gig ethernet. they said they can run big models quantized down from 600Gb to within the 512GB limit they have. This seems wild to bring this level of capability to such a low price point. It will be really cool to start to see demos come out.

I think I'm getting really close to having a good workflow setup for using pyapp. Such an amazing project to allow developers to create applications in python without passing on the hassle of python and managing installs to the user.

Watching Wes fiddle through this with what a sane person would write in a normal day application and not applying the tricks for this kind of battle is how I feel when trying to do leetcode.

This is such a cool idea, I tend to not use laptops at all because they are so uncomfortable I just wait till I'm back at my desk. This solves two main issues I have with laptops, the posture to use them is shit, the keyboards that come on them is not what I want to use. I've solved the latter with my own custom keyboard.

OpenTools is an index for mcp servers to work with new agentic workflows like roo code and windsurf.

I really like this idea for a homelab jbod. Hardware Haven builds out Just a Bunch Of Disks using some pretty affordable hardware and has up to 16 disks added to his homelab with the os having full access to use with zfs.

Nice overview to getting started in zmk

Zuberios Mantic clamp, would ya look at it. This thing looks like a handy tool for soldering. Excited to give it a try.

Damn this looks good, I've been casually keeping my eye out for something like this for quite awhile, I think this will come in handy for keeb builds. Printing one out as I post this, damn I love 3d-printing.

Zach's site looks sick colors are all on point, the fonts are so good. I really like the idea of a style-guide. I think I might be renaming my [[ sample ]] post to style-guide now.

Astral uses just in CI, kinda cool to stumble into this setup in the wild.

        
run: just release-run ${{ secrets.GITHUB_TOKEN }} ${{ github.event.inputs.sha }} ${{ github.event.inputs.tag }}

And her is the accompanying justfile. you can see how it accepts arguments, and starts calling out to other just recipes.

        
release-run token commit tag:
  #!/bin/bash
  set -eo pipefail

  rm -rf dist
  just release-download-distributions {{token}} {{commit}}
  datetime=$(ls dist/cpython-3.10.*-x86_64-unknown-linux-gnu-install_only-*.tar.gz  | awk -F- '{print $8}' | awk -F. '{print $1}')
  just release-upload-distributions {{token}} ${datetime} {{tag}}
  just release-set-latest-release {{tag}}

fixing more ahrefs issues on the road to fixing all major issues within my control I found a ton of urls pointed to an url with a double slash, turns out I wasn't properly referencing slug with post.slug.

I found that I had Structured data has schema.org validation error on essentially every single page on my blog, turns out I had made some changes and have never tried to validate it. Damn json and its hatred towards trailing commas.

Long live RSS! Rss is not dead David, you are right there. I really agree with David that learning a topic well enough to form thoughts and write about it really help learning. You don't need to be an expert, but forming your own thoughts, putting ideas in words takes a lot more than surface level knowledge. When you try to write or speak about something you quickly realize where your holes in understanding are.

Blogging helps me learn. When I commit knowledge to writing it reinforces what I know and shines a spotlight on what I don’t. Most topics require additional research. Even then, I occasionally get things wrong, or miss different ways of thinking, and I welcome corrections. I’ll often update and enrich my posts based on feedback. Without my blog I’d miss other points of view.

As they say, the best way to get an answer on the internet is not to pose a question, but to assert the wrong solution! Most feedback I get is constructive. Sometimes it’s blunt but I try not to read into unspoken sentiment. Some people are more direct. If the end result is positive learning, I can take a hit or two.

In fixing a bunch of meta tags, I introduced Open Graph URL not matching canonical on every page by having trailing / on canonical and not on the og:url.

This commit will fix the error.

I like Davids idea for cotton coder here, reminds me a lot of [[ thoughts ]], which turns out to be mroe commonly called a linkblog. I can relate to David heavily on gathering too many side projects and soem collecting more digital dust than you would really like them to. I use thoughts for quick publishing, very similar to David's notes. I have tags and titles, but the titles are a reflection of the post I'm taking a note on. They are short and sweet, I put just enough thought into them without overthinking them. They live as a separate server hosted website, but the data gets pulled into my blog at build time, so they end up in the same place eventually.

valkey appears to be the largest open source fork of redis that was forked just before their transition to the new source available licenses.

One notable thing missing from the readme is how to run with docker, which I saw in the valkey-py docs.

        
docker run -p 6379:6379 -it valkey/valkey:latest

You can install the python library with

        
python -m venv .venv
. ./.venv/bin/activate
pip install "valkey[libvalkey]"

python bindings for valkey, forked from redis.

one notable difference I see from redis is that you can install with libvalkey to autmatically get faster parsing support.

For faster performance, install valkey with libvalkey support, this provides a compiled response parser, and for most cases requires zero code changes. By default, if libvalkey >= 2.3.2 is available, valkey-py will attempt to use it for response parsing.

        
pip install "valkey[libvalkey]"

I can't believe I've never see this Tim Berners-Lee quote, but I can't unsee it and will be required to reference it from now on.

eventually every URL ends up as a porn site

I had a friend let his blog domain expire, within a short period it was scooped up and was hosting porn. I don't know why, but my best guess is that they were holding it ransom with the most embarrassing content to have your personal site replaced with.

This is a super cool reference for htmx snippets. I really like how he has a couple of errors on the page as examples with examples that fix these common errors.

Interesting thoughts here on blog post titles, do we need them? They are so ingrained into everything.

It makes me think about markata.dev. I don't require you to add any meta data to your post, you don't need a title at all, but you do have to name a markdown file, and this does end up being your title if you don't set one.

Titles are a lot of pressure! I think there is a reason that the big text-based social networking sites (Mastodon, X, Facebook, Threads, LinkedIn, Bluesky, etc.) don’t have titles. Especially for short posts, the title just isn’t necessary. Just say the thing.

Interesting observation what rss readers do without one.

My own favorite[rss reader], Feedbin, shows the author of the post as the title if it’s missing. Eh, not great not horrible.

Hilariously he puts a title on the OG image for the post. I was interested in seeing what would happen in signal, it appears to be showing the author name as well.

Confirmed the page has no title

or does it?

It makes me wonder where Signal gets this title is it from the title block or the site_name meta data

cool article for setting up talos linux with an nvidia gpu. What a wild world it we are living in where these devices that started out being only for hardcore gamers are becoming commonplace in servers and slowly entering the homelab space.

Damn prime makes an interesting point near then end of this video. He's seen a bunch of people able to just throw down charts and shit at their company and end up being "the coding guy" cause they proompted something once. In a way I can relate, I got into software in a similar way, but at a time that it took a lot more hard work, understanding , and copy past from the right stack overflow. Based on some of the people around me at the time I can only imagine how some people must feel like they got pushed into it without wanting it, and now are building something they don't know anything about with no care about it or care to build any expertise. Is the future proompted charts from enterprise chatgpt or do we only continue growing more need for software from here.

Tailwind has the best color system, very well done. Even if you don't use it, it serves as a great color picker.

So proud of Wyatt for writing in his own blog!

Big fan of Primes setup. I was not far off of his setup before he really came on the scene, but I've picked up a ton of nuggets from him and how he operates. I took his first developer productivity course on Front End Masters as it came out.

It is interesting to see him roll back his ansible scripts for bash scripts here. I converted my setup to ansible after watching his first, but have also since rolled back to bash scripts for quite similar reasons. Ansible is great for remote tasks that need to be done on a fleet of machines, but like he says here overkill for this purpose and ends up something that you need to read the docs for every change to your dotfiles.

Unlike prime I've really leaned harder on installing everything in a docker image and developing out of a docker image. I've long built docker images of my dotfiles with the idea that its nice to be able to just use them on other machines, but it rarely happened.

In the past year I've moved bazzite, an immutable distro. It comes with podman and distrobox, so I install very little on it, a few flatpaks from the store for brave and signal, but most of what I really use day to day comes from my devtainer. It's nice that I really have one install target for all of my scripts so they become quite stable. I don't need to worry about arch vs ubuntu vs fedora, no matter where I am its the same base image.

I've also really started to lean on kubernetes, it is so useful to just be able to start a pod in k8s using the same exact develop setup as I would have locally. Nothing needs installed, I can just bring my dev setup to the cluster where the network and data I might need to debug is.

Jhey has the coolest webdev demos!

Today I learned that cloudflare free tier universal certs do not support multilevel subdomains.

By default, Cloudflare Universal SSL certificates only cover your apex domain and one level of subdomain.

Migrating from kedro 0.18.4 to the latest version involves handling the deprecated OmegaConf loader. Switching over does not look as bad as I originally thought.

1. installing kedro 0.18.5+
2. set the CONFIG_LOADER_CLASS in settings.py
3. swap out import statements
4. config must be yaml or json
5. getting values from config must be done with bracket __getattr__ style not with .get
6. any Exceptions caught from Templated config loader will need to be swapped to OmegaConfig exceptions, similar to #3
7. templated values must lead with an _
8. Globals are handled different
9. OmegaConfig does not support jinja2 sytax, but rather a ${variable} syntax

Prime mentioned on stream that Whites were his favorite switch. I tend to like lighter switches and want to give it a try. I really like my Durock lupine's at 55g, the box whites are 45g, that feels like it would take quite a bit more control, floating over the keys.

A good reference of common screen ratios. I just realized that 16:9 is also 1.78:1. I've been putting some images on my blog again, and thinking about using some 2.39:1 ratio on them.

There is a glimmer of hope out there that normal people can scrap together enough gpu to really run the latest models themselves. The ui really appears to be having huge leaps forward such that doing things like rag is no longer such a research project that it was just a few years ago. So excited to see Prime go through this homelab exercise.

Oh, this kills me to hear it. RSS is the OG way to subscribe and share content out to others. It gives you control of what you subscribe to and reminds you when new content lands on your favorite sites. It is a huge component of web 1.0 and I feel is the most decentralized social media can ever hope to be.

I fully believe in our right to repair, ewaste reduction, and bringing a second life to still good hardware that is not up for it's originally intended purpose. This is a sick console like experience you can strap to the back of a tv, throw in your back to take on a trip, or leave stuffed in your vehicle to game in the backseat. Sucks that it cant do 4k, but I've used mine on large screens, and it does quite well for a lot of games, maybe not AAA, but the cartoony multplayer games I play with my kids do quite well.

Damn these deepseek memes go hard. Wild to see openai get played by their own game.

It's crazy that the normie news that I have seen on deepseek shows that the Chinese made what the Americans did at a fraction of the price, without taking notice that they are building on the shoulders of openai.

👏👏👏 This one is really good. I'm right there with him on most of this. I am very hesitant on subscription models, and all the ai tools feel like they are getting ready to be the next round of death by a thousand cuts, this time with pretty limited free tier and relatively high prices to run. I'm sure we will see companies get taken by huge bills soon by building off of someone else's service.

On the flip side I'm definitely the guy that gets in a rut of just copy paste to the ai, wait for codeium to to inject. I feel like I have issues of momentum more than anything. When I'm on one side or the other I tend to stick it out for too long, but less so on going without because that llm drug is calling you when you hit a hard problem.

I'm excited to see him build out a homelab for llm stuff that he mentioned at the top. I'm interested, but probably not building one out for myself until we start to see some cheaper maybe used hardware to do it.

Kelsey says several times in this interview, you don't need kubernetes. If you are running one node you don't need kubernetes. My question though is, would you use kubernetes? Ya I get it if you are a web developer, data scientist, backend dev, but if you are looking to bee a whole ass engineer, or infrastructure engineer, you know kubernetes, Should you use kubernetes on single node?

I came accross from_attributes today it allows creation of pydantic models from objects such as a sqlalchemy Base Model or while nesting pydantic models. I believe in the past I have ran into some inconsistencies with nesting pydantic models and I'll bet one had from_attributes set and another did not.

Arbitrary class instances¶ (Formerly known as "ORM Mode"/from_orm).

Pydantic models can also be created from arbitrary class instances by reading the instance > attributes corresponding to the model field names. One common application of this functionality is integration with object-relational mappings (ORMs).

To do this, set the from_attributes config value to True (see the documentation on Configuration for more details).

The example here uses SQLAlchemy, but the same approach should work for any ORM.

Dang strong stance that tmux and zellij should not exist. I really do get his point though. Theres a good number of terminal features I often miss out on because I run tmux. Its an app that runs apps, and doesn't let all of the signals back to the host. But its fantastic at what it does, and brings so much to the table that the little bit of downside it brings is well worth it to me. The other thing missing in this discussion is that I can take my hotkeys and session workflow to any machine just by running tmux. I do not need to run a certain terminal, or install it headlessly on a server to get special features just for it.

Really good listicle of new modern top python libraries from 2024. Very well done article with images, links, and an actually quality listicle with many things I've never even heard of.

Good overview of seaborn color palettes. They have all sorts of different types, some designed to purposfully give each color the same weight for catecorization. Some designd to give linear differences in value, some have a parabolic feel with a diverging nature.

I've never seen or needed to use a serversideapply in kubernetes before, but I ran into this same issue in my k3s homelab while installing cloudnative-pg.

You can do it with argo

        
apiVersion: argoproj.io/v1alpha1
kind: Application
spec:
syncPolicy:
syncOptions:
- ServerSideApply=true

and you can do it with kubectl

        
kubectl apply --server-side --force-conflicts -f cnpg-1.25.0.yaml

Nerdfont cheatsheet is a fantastic way to copy paste icons into your shell. I just used it to juice up my starship prompt with my current $NVIM_APPNAME managed by [[ nvim-manager ]]

For my next drive upgrade in my homelab I am gong to be using one of these factory recertified drives from serverpartdeals.com. Found them on an LTT video awhile back. They are some lightly used and recertified, fully burnt in drives.

Shop for drives that are certified once again by the manufacturer to work like new. Factory ReCertified drives are cost-effective alternatives compared to factory-sealed new counter parts. Additionally, unlike in mass production, the re-certification process involves closer attention to the overall operation of the hardware so that the re-certification will not have to happen a 2nd time

dust is one of my favorite rust rewrite tools. Its so useful for narrowing down file system bloat and cleaning up some disk space on your nearly full disks. It runs right in your terminal and gives you a nice bar graph on the top directories in use.

Keycloak looks like an interesting way to setup sso. It's part of the cncf so it's got a good backing. I want something better for argo workflows and this might be it. I'm curious what else I can tie into it.

Don't stop learning! Stop trying because you have a doomer outlook on ai, llms, industry and think they are taking over. If you have no hope for the future, if you stop now you are cementing in that you will be no good and the ai will be better. Many, maybe most of us in this industry go here by hard work, long nights of learning, trying to solve problems that our job had. If llms take over then the world is going to be a whole lot different, it will be a world you cannot predict or plan for. For now put your head down and succeed in the world we have today.

TEEJ has some great thoughts on this whole sentiment, put this on for you morning walk or whatever you do.

I like the charts that Theo brings to to these videos. Shout out for a positive k8s reference and not shitting on it.

Htmx brings html/css just a bit further down the complexity graph with little to no extra effort, while react allows us to go all the way full complexity at the cost of build and dev complexity to go from zero to 100 as soon as its introduced.

htmx brings us back to the ease of jquery ajax without any complex swapping or json parsing, all of the object parsing and html templating is done in the backend, the front end just tracks where to put it. HTMX couples the frontend and backend much tigher, since all of the front end html is generated in the backend, done correctly it is not possible for the front end to get out of sync and try to do things that the back end does not know how to handle, vice versa.

It's interesting how many people in tech maintain a blog. I think part of this brings us back to web 1.0 days when so many individual websites owned the web it was a free for all unindexed land and you got to own a small piece of it.

I agree with most of Brittany's points here I write a lot to keep my skills sharp, and to refer back to. Brittany mentions keeping all her old posts, even the cringy ones. I'm all with you here, I'm just wodering how you look back at anything you wrote in the past and not get a bit of that feel, maybe its just me, but I see cringe and mistakes gallore, but it all makes me better moving forward.

nice overview of availability measurements and what they really mean. The crazy world we live in today depends on so many things runnig, its also so hard to measure your uptime, The uptime metrics can mean a lot of different things. The site is up and accepting traffic, but can users make changes or submit orders, there is a lot more to it than just up or down. I really appreciate Brittany's story from Nike nested in there.

The SuperMini nrf52840 is a sick controller for building keyboards, affordable, easy to get, and compact. Bluetooth and wired setup just works in zmk. This page has a nice image of the pinout.

interesting UI for RAG based workflows, i.e. chatting with your documents. It looks like it can run a number of models, feels like ollama with RAG and a nice web ui.

Bluesky is almost excatly like twitter was when I joined years ago. It's gone crazy lately bogged deep in politics, bots, and ads. I've seen like two scroll pages of ads in a row, its nuts. What I did not know before Joel pointed out here is that the feed I am looking at is my following feed, its only feed of people I follow in descending order. On bluesky you get to pick your feed!!! This feels like tweetdeck did back when we were able to run that. You could tune in search terms and save them it was glorious. Bluesky has some really interesting ones that you can use like popular with friends, only posts, my bangers, that have a pre defined algorithms.

This man is responsible for making gaming on linux what it is today. Such a heartfelt story to hear that reviving his dad's machine was at the core of what drove him to do what he has done for the wider gaming on linux community. Update on your schedule, remove all the tracking and bloatware, this is what drove him to start using linux before making it accessible for his Dad.

But really do update, this is not your 2002 PHP box, things need updated and regular updates help the process.

Just tried using my twitter api key for the first time in quite awhile. Apps now need to be tied to projects in order to work. It looks like projects are where pricing comes into play. Thankfully they still give a free tier for doing small time things for myself. You can really see the effect that llms have on these things though as it is 5x more expensive to read posts than to make posts currently. Data is the new gold for these kind of companies.

Wild that the podman-compose github readme calls out k3s as an alternative.

compose definitely has its place, especially for local development on a developers machine, its so much easier to stand up and get things like hot reload up and running smooth.

Intereresting someone built a blog generator in bash. it comes with normal markdown to html, static content, robots.txt, sitemap, rss, and tags. It uses pandoc to take markdown to html and mustache for page templates.

Css is getting so good, new things like interpolate-size are making things that use to require some deep expertise and hacks intuitive and easy.

        
/* Opt-in the whole page to interpolate sizes to/from keywords */
:root {
    interpolate-size: allow-keywords; /* 👈 */
}

Looks like a great start to a rules file for fastapi.

• Place the happy path last in the function for improved readability.

I have never heard anyone say this. It feels weird to me. The other early return, find and handle errors early all make sense to me, but happy path last is new to me.

Really interesting way to generate a rules file for agentic workflows based on your current repo. John uses gitingest here, looks like a fantastic tool, but probably not useful for most private repos. I'm sure you can replicate the same thing in a private repo wtih a small amount of effort the few times you need to do it. gitingest looks like a great way to pull in some extra context for some open source dependencies that you have though.

Gitingest has a python package on pypi that you can run with uvx, and it accepts the same arguments as the web version, right in your terminal

        
⬢ [devtainer] ❯ uvx gitingest --help
Usage: gitingest [OPTIONS] SOURCE

  Analyze a directory and create a text dump of its contents.

Options:
  -o, --output TEXT           Output file path (default: <repo_name>.txt in
                              current directory)
  -s, --max-size INTEGER      Maximum file size to process in bytes
  -e, --exclude-pattern TEXT  Patterns to exclude
  -i, --include-pattern TEXT  Patterns to include
  --help                      Show this message and exit.

Replace hub with ingest in any github and get a prompt friendly codebase ready to feed into any llm. It combines the entire codebase, based on a gitignore style glob that you pass in, into a single TXT file.

Definitely need to give codecompanion.nvim a try, it looks like a competitor to windsurf but in nvim. It looks so feature complete that its hard to grasp all of what it does.

Great panel of software folks at open sauce. It was interesting hearing from all these creators from the perspective of an open sauce panel.

New release out for nvim-manager that supports installing pre-configured distros. It's such a breeze to install these now, its been fun to go through each of them. The currently included distros are.

• LazyVim
• AstroVim
• kickstart
• NvChad
• LunarVim

First release of nvim-manager is out. Your dotfiles change a lot, sometimes it's hard to manage all of the places you have installed them and potentially made hand edits to. nvim-manager allows you to easily make static releases of your dotfiles, and keep your nvim install from breaking by leveraging NVIM_APPNAME and pinned releases of your dotfiles stored in ~/.config. In this directory you might have many nvim configurations installed, nvim-manager automates the process of installing and updating from your dotfiles, while keeping previous pinned versions untouched.

Nice list of url escape codes. I did not actually know that to get a literal sequence like %2D you can use $2D.

This setup fixed my nvim syntax highlighting in helm templates.

imagegenius has made an immich all in one setup that looks much easier to use than immich.

I can say I had the same kind of feelings when I first saw something called "Own Your Web" being run in Buttondown. I totally get it. It takes time and effort to build your own stuff, email sending is hard, not done right ends you in the spam folder. There is something about the name though that I think needs to set an example and self host as much as it possibly can.

The changelog has covered this several times, do they need to go to the crazy lengths they do to run their site, no probably not, but it keeps them in the loop. They are using the tech they talk about in a very real and production critical way to run the show.

Cant wait to see more from ownyourweb.site

Gyroscope better than a mouse?? Nerd nest really sells how having two gyros in the way they have done for noise cancelling changes the game on it, and makes it a contender to replace a mouse.

It really makes me want to try it. I love how repairable this controller looks. I've got to imagine that the fact that it comes as a kit, and all the parts are available that this hits S tier repairability. My current controller of choice is a PS5 and I've had stick issues I wish I could fix.

No analog triggers, I'm out. Maybe they will make it an option in the future idk. I don't play shooters where I need a hair trigger, this won't work on session or driving games.

micro usb, seriously, that kinda kills it for me too.

edit I looked on their website and they have usb c in the latest version

postiz looks like a very polished way to automate and schedule posts to all the social services.

Reminder to myself, look into self hosting postiz with this helm chart later.

Chris Coyier had a small re-align on his site, some good nuggets in here.

I like the idea of having a photo of myself prominently on the site, so you know who you’re dealing with here.

I really like this after thinking about it and I think I am going to make sure I get my face back on my posts. I do have my 8bit style pixel art image of me that I use on social media, but no real picture.

I feel like a lot of people redesign their entire website when it’s time to update to the latest list of social networks and I’m no different. Once you touch it you gotta keep going.

I can totally relate to this, once you open the thing, you get the build tools greased up, and your confidence high that re-deploying isn't going to mess something up, I tend to start digging in to other things.

I've only recently learned what colophon means, and I really like to read through site that use it. If you don't know its about how the site is built. I've always liked peeking under the hood of things to understand how they work, it's what turned me towards an engineering degree.

I love how he mentions that he chose the name when he was 17 and he is stuck with it. I particularly like the name, it has something special to it. Hats off to you for doing something that has lasted so long for you. I fully understand though, I have projects that I made a year ago that I think why did I name it that. At the same time when I try to think of a name I end up with the I don't have anything good and I'd rather build the thing so fuck it, its going to be what it is.

Very interesting way to catalog games, I need to make a catalog of mine, I'll probably start adding some blog entries for games I'm in and have completed. Wonder if there is a way to hook into steam with python to get achievements and progress live.

Tons of cool people came out with their rss feeds here, again will need to browse more closely later.

Sturobson has a ton of rss feeds here, I recognize quite a few, will definitely need to poke at some of these later.

I'm a sucker for good own your own shit on the web blogs, and Matthias Ott has a top notch one here. The archive has been a great read so far, I've discovered things like slashpages.net.

What a great first guest for FAFO. Kelsey is always such a great listen. He talks about killedbygoogle and how engineers get no incentive to work on old projects, google had no incentive, and you got the thing for FREE in the first place. He talks about end of career and having love you money, having so much in the bank you can say no. If you are presented with a project that does not align with your values you can say no.M

Justin even mentions how Google has more killedbygoogle projects than Amazon has total projects.

If we knew how hard it would be, we would never build it.

Autumn's Fav quote from ep1 of shipit

Well done changelog with some really good inspirational nuggets. Many slash pages I want to check out and an 11ty contribution graph.

A nice list of slashpages you might want to consider including / aliasing / 301ing. These feel like nice things to setup and keep in the back pocket for obsidian style wiki link to easily. I get kinda bad at wiki-linking as much as I would like to, mostly because it does require some amount of work to make the page, and keep it up to date over time, then remember that you even have it.

Some are serious, some very common, some quite useful.

Theo does a fantastic history of serverless here.

Kubernetes shit

Theo can't have an infra video without shitting on k8s. Specifically people who have never touched k8s pushing fear of k8s to large audiences of people who have never touched k8s. If you are a webdev who solely lives in webdev space and never touches as much as a dockerfile listen to him. If you touch infra at all try it before you take his opinion at face value.

Serverless shines in high variance

If you plan on having traffic spikes 10x your regular traffic for something like black friday, serverless might be right for your use case.

stateless programming

He argues that targeting a stateless deployment of serverless leads to better code. I'd like to see more examples here. Maybe most of the code bases I work on already do this. I've never targeted a serverless deployment, but I've targeted horizontally scaled deployments many times and they feel like they have the same targets. For instance if I spin up 8 pods for my application or uvicorn with 3 workers I have to target statelessness, all of the state must live in the database and cannot live in memory. Even if I target 1 instance in a containerized environment I have to be ready for restarts at any point in time.

I might be missing something here, but I don't see how this point applies to serverless.

Scale to Zero Services

https://www.youtube.com/watch?v=_VQl_HTk9PM&t=2607s

Companies that can run on scale to zero can allow your shitty side projects that have no users run for free indefinitely because it costs them nothing. He compares planetscale running mysql vs Turso running SQLite stored in s3.

Everything he covers is really cloud to cloud

He compares early deployments of LAMP and MEAN stack running on one server then jumps to serverless. It feels like he is missing the angle of owning your own hardware. This might just be the experience difference between theo and DHH. DHH works on one focused company, Theo is jumping around between startups.

Ok I get some of it

At the end he covers a dumb side project that has branch deploys. It had a 100s of deploy still running, some very old, some never even touched, vercel can do this because it costs them nothing if its not running. What I didn't think about right away is that there is probably a backend component to this.

In my python backends I often have deploy environments, but since they don't scale to zero they have at least some cost even if its small, with this I cannot just keep hundreds of them running because that cost would add up.

I would argue that this is hardly useful, because you can checkout any old version and run it locally, and you rarely are going to need to poke through these old branch deploys. But in that case that you need to quickly do this serverless is definitely superior.

linkding looks like an interesting alternative to thoughts. Thoughts is focused on the note being a value add tweet length blog post that you share to the public. This seems more focused on fire and forget with some note taking and search ability. I should definitely level up the search and tag discovery in thoughts.

cool video on expanding vim with cli.

piping commands into vim

write a healthcheck

Lane from boot.dev madde this fantastic video about serving files on the internet. It has me wondering if I need to rethink a few of my things that I have built. I have a few things I am serving media from, but I have very aggressive cloudflare cache rules on them, so each file should only be uploaded about once per year.

My problem going straight out of minio right now is how do i set headers for cache control on it. If I can't set the cache control and everything is coming out of minio this does not solve my problems.

I went back and played with presigned urls and you can in fact control and set response headers, this is definitely the way and I have been wrong.

This is a cool new feature coming to bash, I can't think of a use case I have out of the gate, but it looks cool. I'm thinking this might be good to keep in the back pocket for something like CI where I don't have a hightly tuned bashrc file, and I want a dynamic curl request based on some state that exists as a fille.

Nicely worded Cassidy! Javascript can be too much, it can bog down low powered devices, we can ship so much that its untenable on poor connections. It can be argued that its a bad language and putting it on the server is a mistake, but it brings sites to life. It makes it possible to extend the static nature of html/css with just a little special spice only your site needs.

I'd add to the argument that a lot of js should go away over time. Over time libraries such as jquery have fallen out of use, not because they are bad, or have been replaced by new libraries, but because the browser has adopted most of the functionality that jquery brought.

As a primarily python dev I'd really like to see htmx die a very similar graceful death. There have been several iterations on this idea, and the crux of it is very similar, give the ability to use HTTP verbs right inside html with some instruction of when to apply them and what to apply them to.

No js should not go away, it never will, we will always find new patterns that the browser should take, sites will always need a small bit of spice to bring them to life.

/p warp will warp your party to your current server in hypixel skyblock

Today I learned how to configure the baseurl for htmx using the tag. This is pretty handy to be able to configure different baseurls.

        
  <base href="<scheme>://<netloc>/api/v1/">
  <button hx-post="clicked"
       hx-trigger="click"
       hx-target="#parent-div"
       hx-swap="outerHTML">
    Click Me!
  </button>

WTF, you can just change a server response from devtools and update a vuejs app? Just tried with htmx, and my GET requests are not showing up in the sources tab. I'll keep this in my back pocket for next time I'm supporting a vuejs app though.

Gary Vee, is the goat. leaving this image as a reminder for myself

Atuin v18.4 is out with an atuin wrapped command.

Here's Ellie's Wrapped.

This is a pretty sick command, I've only been using Atuin about as long as I've been on bazzite, so maybe 4 months, but here is my wrapped.

Comparing to Ellie, I'm surprised that I even have 126 package related commands, being on an immutable distro most of my packages come from the container, I don't need to run packaging commands. My cd is also super low, I use tmux sessions to take me where I need to be most of the time, which is a project's root directory, It's pretty rare to see me outside of a project root directory.

I'd venture to guess that nvim would be higher if it didn't auto start on every tmux session as well.

Red Hat has donated the whole open alternative to docker to the CNCF, the hosts weigh in with thier opinions all being pretty positive as they seem to be a legitimate donation and not dumping crap on open source.

• podman
• podman-desktop
• buildah
• compose-fs
• podman
• bootc
• scopio

Personally I've been running podman exclusively at home since switching to Bazzite in August. I'd like to use buildah in ci. I gave it a try, but was unable to swap out my use of kaniko for buildah, I will get there, but it was not quite as drop in as I wanted.

Holy shit these AGI models are incredibly expensive to run, require lots of wild hardware that there is not enough to go around, and requires shit tons of power to run.

Now more than ever is time to distinguish yourself with deep expertise, jack of all trades is being eaten by ai. People with deep expertise are getting a jack of all trades bump from ai, not o3, just the regular stuff.

Theo weighing in on the 2025 job market. It's no 2018 out there right now, the ratio of jobs to engineers in the market has flipped big time. Theo as usual really focuses on community, being in a community, and being a good citizen. At the end of the video Theo weighs in on his experience hiring, and generally it starts with we need someone to do x does anyone know someone, then goes to a more formal internal post, then more formally asking internally does anyone know someone, then maybe to his community, and if he really still needs the person it might become an external post. These days there are so many good engineers on the market that very few good jobs actually get a posting for in his opinion.

kitze hating on python packaging in new ways. Python packaging has a lot of quirks that can make it infuriating. Not once have I thought "you know what this needs, quotes and braces"

This tip of using tinkercad to do boolean operations on an stl of a solid gridfinity bin and an outline is absolute fire 🔥🔥🔥. This feels like a relatively simple operation, but to do it to a generated stl proves hard to do in most modeling software, at least harder than it needs to be. Somehow tinkercad got it right and made it a very basic operation to do.

It took me a minute to find the Merge button that Uncle Jessy mentioned, they call it a group in TinkerCAD.

Damn Glorious Eggrolls is still making gaming on linux better. Of course its containerization that drives everything on linux these days. This is a pretty badass talk. Umu is already running in steam and bazzite. Bazzite gamemode uses this to get a gamescope session running.

Great list of 4 tips for running fastapi applications.

Keep routes small

Fat routers with all of the logic built in makes them hard to test, hard to refactor, causes lots of duplication, and makes it hard to reuse the business logic code later in something like a cli application.

Deploy Early

I really like this advice! He reccommends deploying as early as you can get a healthcheck live in your application. I've found too many times developers build something that is really hard, or impossible to deploy, when if they had tried to deploy early they would have spotted some easy to fix issues. This is less important if you are building out of a template that your team commonly deploys from, but very important with new patterns.

https://youtu.be/XlnmN4BfCxw?si=ks1wvmgDyoQLgrv2&t=1093

This is really interesting, the lazy uv scripts are really becoming quite appealing, especially for something like this to just pop out of an llm ready to run. The article features several examples of these one-shot prompt ideas that I suggest you give a try, and a prompt for creating them.

Depot's uptime seems to be great. I definitely hit some issues with it this afternoon 12/24/24 that were not reported. I wonder if my issues were with the fly integration. Maybe fly ran out of credits to depot or something.

Here the integration to depot appears to be opt in using the --depot flag on fly deploy. This must have changed over time though because today it was giving me issues and I had to opt out using fly deploy --depot='false'. Looks like a great service and I just learned about them on their bad day.

Just learned about depot today ironically because it seems to be down and fly is using them under the hood to do the container builds, seems like a really great service for fast builds accross your team.

This was an eye opening video into agentic editing workflows.

setting rules

Dfferent ai tools use different rules files, windsurf uses .windsurfrules.

testing out rules

Test out your rules file by having it say something at the beginning of the output to verify that the rules are being applied correctly.

First line

He suggests to use this key rule for debugging purposes, otherwise you are guessing to what rules if any it is following.

        
Every time you choose to apply a rule(s), explicitly state the rule(s) in the output.  You can abbreviate the rule description to a single word or phrase.

Dax talked about this in a recent How about tomorrow podcast https://thoughts.waylonwalker.com/post/461. He is using it as his dev machine, he just ssh's in and devs on it. Feels like quite an interesting workflow, their prices seem competitive, but as a cheap ass homelabber I see their prices and think I could grab a used optiplex for the cost of a month or two of these and probably wouldn't know the difference. DAX mentions longer compile times so maybe he does notice.

This is a pretty great episode talking shop with typecraft. They talk setups, cameras, content creation. I found them talking about their linux setups particularly interesting. Dax talked about his flow from building his own machines to using reliablesite.com.

tmux default leader

Dax hates on c-a, both typecraft and dax use c-s, which normally freezes a terminal, we can all agree that is useless. I use the default c-b, it seems fine for me.

type crafts setup

• Ubuntu
• Ghosty
• Tmux
• Nvim
• Ruby

Selling coffee without a web front end

Dax talks about terminal.shop and how they originally planned to have a web front end, but after they had so much success they stuck with it. now they are leaning harder into it and are building out integrations with a bunch of languages and an api, but no front end.

PEP 723 is what is inspiring all of these lazy self installing python scripts, Authored by the author of hatch and pyapp. This is a really cool thing that uv has picked up and made python packaging just a bit easier.. maybe... dependency resolution still sucks.

I've kinda fallen out of using direnv now that a lot of my projects use hatch, I generally just hatch shell into them. I just need to make sure I go through all of them and make my installer uv. Now I've been thinking about making uv my only needed dependency to run a python project and leaning more to something like uv run --with . uvicorn myapp --reload

I really like Trey's steps to making an executable python script with uv

his old process seems to be the same with a new shebang

1. Add an appropriate shebang line above the first line in the file (e.g. #!/usr/bin/env python3)
2. Aet an executable bit on the file (chmod a+x my_script.py)
3. Place the script in a directory that’s in my shell’s PATH variable (e.g. cp my_script.py ~/bin/my_script)

And here is the new format the the shebang followed by the metadata comment block defined in PEP 723.

        
#!/usr/bin/env -S uv run --script
# /// script
# requires-python = ">=3.12"
# dependencies = [
#     "ffmpeg-normalize",
# ]
# ///

I still cannot believe I thought I had such a unique idea with thoughts only to find out shortly thereafter that Simon beat me to the punch by about 20 years!

Simon put a lot of work into this post please check it out. Its not only filled with tons of little nuggets about blogging, it has just a ton of links to other posts I'm itching to read now.

https://simonwillison.net/2024/Dec/22/link-blog/#thoughts.waylonwalker.com

a linkblog proves that you read

Simon mentions it in the post as well, but one thing about having a link blog does is that it proves that you actually read articles that you share. This is really more about proving to myself, and reminding myself that anyone else, it's a good feeling.

Trying to add soemthing extra

• include names of people, and add tags for common people
• something extra - the whole point is to bring something to the post without stealing thunder, but rather linkbuilding for the author that put the hard work in.
• copious amounts of quotes
• It should bring the author up not put them down

A post a day

well .98 posts per day

Simon mentions starting in Nov 2003, which puts that at 7722 days since then. That is so close to a post a day for 21 years, even though they are often as short as one sentence that is impressive to keep up over time.

So far I’ve published 7,607 link blog posts and counting.

What's my rate

I just check and I am at 513 days in with 413 posts (0.8 posts/day). Not quite up to Simon, but close enough to be proud of. My only grief is that I didn't start sooner.

My stack

I have a full post explaining more about how I do mine and the stack behind it here.

https://waylonwalker.com/thoughts/

See Images are hard!

This is one of the main reasons that I built dropper for myself, and plan to include it in fokais as a product to help others with their blog. This is definite evidence that I am not the only one with this issue!

I've made images easy form myself

I've tried half a dozen or so easy techniques from obsidian to various github repos and none have really hit me with inspiration to really add screenshots like I do with dropper. I write all of my thoughts posts in a chrome extension drop down local first style and when i hit paste it uploads and gives me a markdown image link, it cannot be easier for me.

When I store them I compress them and convert them to webp. I serve them by streaming them out of an s3 bucket with fastapi and some very generous cloudflare cache headers, so I should only get a request for that image about once a year.

Warning, I built dropper for me, no guarantees that it works forever, foakis will be for more users.

Bluesky does not penalize for links

I am loving bluesky more and more, learning this just gave it another level up.

Series

I just noticed that Simon has Series on his blog in a very elegant way. One I need to get back to blogging more, I did not do so good in 2024, and two this is awesome and I want it.

test

Ok, second post on this one. I am sending only head requests, so I want to see the first request for the image, which happens in the chrome extension after pressing submit. It will not yet come from cloudflare so I am interested in what it looks like just streaming out of object sorage. This time I have the inspector open on the tab.

As expected I got a cloudflare cache miss on first hit.

But the overall performance of 351ms to get the image was not bad considering it takes several seconds to get the image fresh.

Now closing the chrome extension popup, and hitting the main thoughts page again gives me a cache hit from Cloudflare's CDN.

And the total request time goes from 351ms to 21ms with the Cloudflare CDN, quite impressive!

Sorry with how meta these last two posts were, I needed a second post to get that fresh request graph from.

This is a test post. I just updated the chrome extension for thoughts to take a the thumbnail image on open of the chrome extension so that it is ready when i publish it. let's see what happens.

I really like this idea of keeping a set up contacts in a markdown repo, and being able to wikilink them to different tickets / meetings and what not. I'm imagining the graph you can build, it feels quite interesting. Even more relevant as we see things like teams rolling out retention limits to messages.

Here is Cassidy's format, I like it but I'm probably not going to track the birthday of most people I work with, thats just not the relationship I tend to have with work friends. It might be a midwest or non tech thing, but I am not even aware if any of my co-workers have social media, and I assume that if they did it would not include anything work related but more football and other sports.

Dang context can really cause you to pull your hair out. Context seems so freaking convenient, but I've avoided it and just ssh in for these reasons. Maybe I'll come around eventually but for now ssh is my friend.

Reminder to tune in later, can't watch now, but saw a link on Dave's blog.

Cheers to the Unique brains dave. I can say I am 100% with Dave on all of these, except the clean environment thing, lucky Dave. If I have enough room to see what I am doing and walk around a mess, I probably wont even notice it. I'm so hyper focused on what is right in front of me that mess could be a huge pile of cash and i'd never see it.

Damn I'm loving using dropper already, I am able to copy paste images right into posts on thoughts. This is my first post that really utilizes them heavily to show how different styled rss feeds look in the post. These were super quick to screenshot and paste in, this is a DREAM workflow for me.

When I paste an image into my thoughts chrome extension it automatically uploads it to dropper, and places a markdown image tag in its place. Ya, I know the alt text could be better, but it feels fine for quick and dirty posts like this.

check out https://thoughts.waylonwalker.com/post/448 for some sweet styled rss

I love me some styled rss, it's how the OG internet was made to be. You choose what you want to read and when. There is no middleman aggregator inflating the reach of things they want you to see or suffocating things against them. It's just you and your internet friends.

Cassidy's Styled rss

Cassidy has a quite lovely and easy to read rss feed, with an open source style sheet, that is part of her open source blog template for astro blahg, love that name by the way!

Dave Rupert's Feed

I first learned of styled rss feeds from shoptalkshow.com, specifically from Dave Rupert.

Dave uses a pretty bog standard styled rss feed with pretty-feed-v3. This repo is probably the best place to start if you are looking to style your rss feed.

https://github.com/genmon/aboutfeeds/blob/main/tools/pretty-feed-v3.xsl

My RSS Feed

My feed comes from my static site generator markata, a plugins all the way down static site generator written in python, made for customizing, lacking just about any and all docs so you are on your own.

Dropper is a place for me to drop files for quick posts like this.

Nice example of adding a healthcheck to fastapi, and integrating it with docker. Don't forget to include curl in the install, nice touch.

new versions of just now come with color variables already set.

        
[group('manage')]
version:
    #!/usr/bin/env bash
    version=$(cat version)
    echo current version {{BOLD}}{{GREEN}}$version{{NORMAL}}

I found this nugget in thechangelogs justfile, it lets you add color to your justfile with variables quite easily.

        
# https://linux.101hacks.com/ps1-examples/prompt-color-using-tput/

_BOLD := "$(tput bold)"
_RESET := "$(tput sgr0)"
_BLACK := "$(tput bold)$(tput setaf 0)"
_RED := "$(tput bold)$(tput setaf 1)"
_GREEN := "$(tput bold)$(tput setaf 2)"
_YELLOW := "$(tput bold)$(tput setaf 3)"
_BLUE := "$(tput bold)$(tput setaf 4)"
_MAGENTA := "$(tput bold)$(tput setaf 5)"
_CYAN := "$(tput bold)$(tput setaf 6)"
_WHITE := "$(tput bold)$(tput setaf 7)"
_BLACKB := "$(tput bold)$(tput setab 0)"
_REDB := "$(tput setab 1)$(tput setaf 0)"
_GREENB := "$(tput setab 2)$(tput setaf 0)"
_YELLOWB := "$(tput setab 3)$(tput setaf 0)"
_BLUEB := "$(tput setab 4)$(tput setaf 0)"
_MAGENTAB := "$(tput setab 5)$(tput setaf 0)"
_CYANB := "$(tput setab 6)$(tput setaf 0)"
_WHITEB := "$(tput setab 7)$(tput setaf 0)"

Usage

        
echo:
    echo {{_BOLD}}{{_GREEN}}hello there{{_RESET}}

Hurl was mentioned by @gerhard on the latest changelog and Friends. Looks like a feature rich easy to use testing tool that is tested via what looks like a config file.

Hurl is a command line tool that runs HTTP requests defined in a simple plain text format. It can chain requests, capture values and evaluate queries on headers and body response. Hurl is very versatile: it can be used for both fetching data and testing HTTP sessions. Hurl makes it easy to work with HTML content, REST / SOAP / GraphQL APIs, or any other XML / JSON based APIs.

Simple, modern and high performance file watching and code reload in python.

For some reason the ublue ecosystem does not prompt you to set your hostname on install and you get a hostname like bazzite showing up. Looks like this is the fix.

hostnamectl --static hostname babyblue-aurora

Great examples for working with s3 buckets with ansible.

On reboot of my opnsense router it did not tailscale up. I'm not sure if a key expired or what happened. The fix was to first enable ssh, then ssh in and run tailscale up.

enable ssh

In opnsense System > Settings > Administration > Secure Shell > Enable Secure Shell

tailscale up

        
ssh <opnsense ip>
8 # to select shell
tailscale up

Follow the link to log in.

disable ssh

now uncheck secure shell to lock down the opnsense machine.

In opnsense System > Settings > Administration > Secure Shell > Enable Secure Shell

This guy fit 2 3.5" drives and an sdd in a dell 7050 SFF. It looks tight, and there it probably not much airflow running in it, but its an impressive project.

I tried this fix on my secretlabs chair, but unfortunately it did nothing for me. Mine ended up being an issue in the tilt mechanism and not the hydraulic piston. I greased the crap out of it, it can only help the longevity. I did notice that the two washers on either side of my bearing were quite worn in.

This is definitely happening to my secretlabs chair. I did the fix this morning, its still not perfect, but it's better now than it has been in a long time. Mine use to be so bad that it would pop and creak every time I sat down or moved the smallest bit. It's now quiet through the mid range of its motion with the occasional pop when I sit down.

switching cloud providers, theres no easy way. K8s was supposed to get us there, haha, the deep integrations with each vendor just keep locking us in

Dang, love this guys branding. Hooks to get stuff offDaBench, what a freaking cool name.

This is a really amazing documentary of argocd. I got into k8s pretty late in the game. Which is pretty typical for me. As I went to use k8s for the first time i was using workflows, then cd. both of these tools had a level of polish that made them seem like they had been there forever and not quite as young as they actually are.

I thought it was interesting how they focused on how the name must be two syllables or less, start with a or b, logo needs to be cutesy funny and recognizable seemed interesting, but puts them at the top of lists and makes them look like they've been there forever.

bootc is the underlying technology that enables OS's like bazzite, and Aurora from the uBlue ecosystem.

A new completion plugin that I might give a try. Readme makes it sound like its built on some fast teck that allows them to handle a lot of items and run more frequently. The videos look like they don't have some of the same issues cmp does for me. Maybe its my configuration, but I'm pretty sure it does not update when you backspace and things like that.

These css anchors are crazy that you can do this with html and not a bunch of js, probably requiring a library or framework.

Staically makes creating OG images so easy, you can just pop this in your section with some og tags.

        
    <meta name="og:image" name="og:image"
        content="https://cdn.statically.io/og/{{post.title}}" />
    <meta name="twitter:image" name="twitter:image"
        content="https://cdn.statically.io/og/{{post.title}}" />
    <meta name="og:image:height" content="640" />
    <meta name="og:image:width" content="1280" />

I hit an interesting error after updating my nvim plugins today. I'm sti not even quite sure what a ShaDa file is, but I found min in my nvim state directory, unlike this issue that mentions it being in share.

The Error.

Error while reading ShaDa file:

The Fix

        
mv ~/.local/state/nvim/shada/main.shada ~/.local/state/nvim/shada/main.shada.bak

Hynek has a sick dockerfile setup for using uv in python projects.

SQLModel models ship with an is_, and is_not that you can use to compare to None without pesky linters complaining.

This comment summed it up quite well.

I believe this is concerned entirely with SQLAlchemy, not with SQLModel, and has to do with the required semantics to construct a BinaryExpression object. Hero.age == None evaluates to a BinaryExpression object which is eventually used to construct the SQL query that the SQLAlchemy engine issues to your DBMS. Hero.age is None evaluates to False immediately, and not a BinaryExpression, which short-circuits the query no matter the value of age in a row. From a cursory search, it does not seem that the is operator can be overridden in Python. This could help explain why the only possibility is by using the == operator, which can be overridden.

so rather than using Team.heros == None we can use Team.seros.is_(None) which checks for itentity not equality.

Non Gridfinity Rugged boxes

gridfinity rugged box openscad

Cluster API book

Gridfinity generator uses open scad, so you can make rugged boxes, bins and base plates with form input. not fully custom fit to things, but you can custom size square bins, hole cut out sides and all. From what I can tell, no bento box either. so as long as what you are looking for is square this generator has you pretty well covered. I'm definitely using this for simple bins and rugged boxes.

The BethDeck is a sick looking DIY steam deck competitor built on top of the framework 13" mainboard. No soldering needed, lt with all off the shelf parts. Controllers appear to be an off the shelf controller with the shell removed and plugged into a usb hub.

    I really like how Spencer walked through how to make quite a few unique setups with only a few simple obstacles.  It does not take anything fancy to have a cool setup.

Today I learned that you cannot pass negative integers as values to typer. in this case context_settings={"ignore_unknown_options": True} is required so that the - does not look like a flag.

        
# script name: main.py

import typer

app = typer.Typer()


@app.command()
def failing(value: float):
    print(f"{value=}")
    

@app.command(
    context_settings={"ignore_unknown_options": True}
)
def working_good(value: float):
    print(f"{value=}")
    
    
if __name__ == "__main__":
    app()

Aaron Francis is a database master, pumped for thsi dude and all that he is able to accomplish.

This new demo of textual is wildly awesome, so many widgets and features being added into the main library. The themes and animations are on point and far surpass my expectations of a tui.

Loving this 6 pink with blueforward multicolor combo the porta john has going right now. Also my son printed me a cute black pumpkin.

ploopy just dropped a trackpad that runs qmk. This sounds so cool. I'm excited to see some videos on it. It would be sick to get this incorporated into a porta john running zmk, one bluetooth device to run all the peripherals.

Zulip was just featured on a changelog clip as a self-hosted chat offering. I'm interested to give this a go and see how it compares to matrix. glancing at it, it looks a lot like slack.

weazyprint was throwing me some errors, turns out that it's currently not compatible with the latest pydyf package.

my error

        
TypeError: __init__() takes 1 positional argument but 3 were give

I fixed it by locking in pydyf at 0.8.0

        
pydyf==0.8.0

helix inspired treesitter select outwards and select inwards.

Helix officially made it in my devtainer. I am installing the binary right out of the github release with wget and tar. I can't use installer because their release format does not match what installer is looking for.

!!! note: I also move the runtime over to /usr/local/share/helix/runtime.

!!! note: I use /usr/local/bin rather than my home directory as it works better with distrobox.

Here are some really great keybinds to make helix vim-like. It feels very antithetical to use this whole sale and forego the helix motion-verb language, but there are some nuggets like G in here that I find useful.

helix multicursor has the ability to do what vims :s/hello/world/gc does by using ) to cycle through cursors, and n to discard unwanted cursors.

Here is a really good vim substitute with regex capture groups, saving this one for a rainy day.

        
* Reading 1: This is a title to a link
* Reading 2: This is another title

:%s/\v(: )(.+)$/\1\[\2\]\(

        
* Reading 1: [This is a title to a link](
* Reading 2: [This is another title](

Interesting take on refactoring the whole codebase from your editor. In this thread the-mikedavis suggests multicursor as a replacement for substitute, and later in the thread takes the side that larger workspace wide edits are outside the scope of helix and he would reach for a refactoring tool like fastmod to do the job.

I am a heavy user off substitutions in vim, helix does not substitutions built in, rather it leans on multicursor support.

to replace every instance of hello with world in vim

        
:%s/hello/world/g<CR>

and in helix you would

        
%shello<CR>cworld<ESC>,

How to make helix themes transparent. You can make any built-in theme transparent in helix with one line, a few extras and you can make all the pop ups, help menus and status line trransparant as well.

        
mkdir -p ~/.config/helix/themes
hx

:o ~/.config/helix/themes/dracula_transparant.toml

        
# ~/.config/helix/themes/dracula_transparant.toml
inherits = "dracula"
"ui.background" = { fg = "foreground" }
"ui.menu" = { fg = "white" }
"ui.popup" = { fg = "white" }
"ui.window" = { fg = "white" }
"ui.help" = { fg = "light-gray" }

"ui.statusline" = { fg = "gray" }
"ui.statusline.inactive" = { fg = "black" }

:config-edit

        
# ~/.config/helix/config.toml
theme="dracula_transparant"

This post shows how to set up multiple LSP's in helix, the example uses pyright and ruff-lsp for python.

Add this to your ~/.config/helix/languages.toml a

        
[[language]]
name = "python"
auto-format = true
language-servers = [
    {name = "pyright"},
    {name = "ruff-lsp"},
]

[language-server.pyright]
command = "pyright-langserver"
args = ["--stdio"]
config = {}

[language-server.ruff-lsp]
command = "ruff-lsp"
args = []
config = {}

How to set your python formatter to black with helix. The following snippet lays out how to set the helix editor to auto-format on save with the black formatter.

        
[[language]]
name = "python"
language-servers = ["pylsp"]

[language-server.pylsp.config.pylsp]
plugins.pyls_mypy.enabled = true
plugins.pyls_mypy.live_mode = true
plugins.flake8.maxLineLength = 88
plugins.pycodestyle.maxLineLength = 88

[language.formatter]
command = "black"
args = ["--line-length", "88", "--quiet", "-"]

Simon shared a really cool new utility tool for sqlite ispired by rsync. It checks hashes of each sqlite page and syncs pages. So if nothing in the database has changed it will only require 0.5% the bandwidth as a copy would.

porta-john design update, working on moving the seams to line up with an edge to eliminate long noticable seams. These seams are required because I don't have a printer big enough to print this in one shot.

hotswaps are going in. I'm getting really excited for hotswaps as they will make keebs more repairable, and users can self change out the switches if they want. I can also pre-build them ready to go, and drop in switches at the time of order. I did learn today that these are hyper specific to gateron, damn all these switch manufacturers and their special pinouts.

Last thing I gotta make more room for wiring the microcontroller and running wires out to the rows and columns, my first prototype build took waay to long to build.

Is Jesse going to get sniped by the helix community?

This low profile design almost made it to 9. I think blending together the custom brim to make them all one piece really made the final difference here. The custom brim is two layers thick, only touching on the top layer making it very easy to peel off, but gives better adhesion to the bed, and does not loose one like this print did.

really excited for how these caps of my own design are working out. Having the extra control is the first time I've been able to print more than one at a time without some hacky gcode. I've got a set of nine here that I have printed 4 times without a single failure.

MX keycap stem dimensions. I used these to create my new caps and all these dimensions worked great. The one dimension I had issues with was the outer diameter off the cap stem, not shown here. 5.3mm ended up being a good outer diameter for me. I print them at an angle and the bottom of the stem can get a bit of sag, pressing out on the outer wall of the switch and can actually stick the key.

This page is gold. It lays out all of the distrobox assemble api with some good examples of how to get access to things like podman and kind from inside of containers.

Especially this example.

        
[tumbleweed_distrobox]
image=registry.opensuse.org/opensuse/distrobox
pull=true
additional_packages="acpi bash-completion findutils iproute iputils sensors inotify-tools unzip"
additional_packages="net-tools nmap openssl procps psmisc rsync man tig tmux tree vim htop xclip yt-dlp"
additional_packages="git git-credential-libsecret"
additional_packages="patterns-devel-base-devel_basis"
additional_packages="ShellCheck ansible-lint clang clang-tools codespell ctags desktop-file-utils gcc golang jq python3"
additional_packages="python3-bashate python3-flake8 python3-mypy python3-pipx python3-pycodestyle python3-pyflakes python3-pylint python3-python-lsp-server python3-rstcheck python3-yapf python3-yamllint rustup shfmt"
additional_packages="kubernetes-client helm"
init_hooks=GOPATH="${HOME}/.local/share/system-go" GOBIN=/usr/local/bin go install github.com/golangci/golangci-lint/cmd/golangci-lint@latest;
init_hooks=GOPATH="${HOME}/.local/share/system-go" GOBIN=/usr/local/bin go install github.com/onsi/ginkgo/v2/ginkgo@latest;
init_hooks=GOPATH="${HOME}/.local/share/system-go" GOBIN=/usr/local/bin go install golang.org/x/tools/cmd/goimports@latest;
init_hooks=GOPATH="${HOME}/.local/share/system-go" GOBIN=/usr/local/bin go install golang.org/x/tools/gopls@latest;
init_hooks=GOPATH="${HOME}/.local/share/system-go" GOBIN=/usr/local/bin go install sigs.k8s.io/kind@latest;
init_hooks=ln -sf /usr/bin/distrobox-host-exec /usr/local/bin/conmon;
init_hooks=ln -sf /usr/bin/distrobox-host-exec /usr/local/bin/crun;
init_hooks=ln -sf /usr/bin/distrobox-host-exec /usr/local/bin/docker;
init_hooks=ln -sf /usr/bin/distrobox-host-exec /usr/local/bin/docker-compose;
init_hooks=ln -sf /usr/bin/distrobox-host-exec /usr/local/bin/flatpak;
init_hooks=ln -sf /usr/bin/distrobox-host-exec /usr/local/bin/podman;
init_hooks=ln -sf /usr/bin/distrobox-host-exec /usr/local/bin/xdg-open;
exported_apps="htop"
exported_bins="/usr/bin/htop /usr/bin/git"
exported_bins_path="~/.local/bin"

This is a very well thought out zmk config featuring many macros, numword, and timeless homerow mods. The build system to build locally looks on point, I really need to give that a try!

Today I learned that you can use init_hooks to access host machine commands from inside a distrobox container. This is super handy for things that you cannot get to from inside the container and need ran outside (docker, podman, flatpak, xdg-open).

        
init_hooks=ln -sf /usr/bin/distrobox-host-exec /usr/local/bin/podman;

Tailscale comes with a feature called taildrop that lets you easily share files between machines on your tailnet. If you have tailscale on ios/android it shows up as a share target when you try to share something, and you can pick the machine to share with.

What was not obvious to me was how to receive the file on linux. The linux tailscale service does not automatically receive the file, which can be kinda nice that you can put it where you want, but was not obvious to me at first. Use this command to receive files.

        
sudo tailscale file get .

The Heawood42 is an interesting diodeless keyboard that is not direct wired. According to the repo this is the only keyboard to be diodeless and not direct wired. It does this through the use of a graph.

This is a wild key cap that uses a lever to convert horizontal key presses to vertically press the switch down. The leveret v2 uses two of these on thumbs, one to press outward, and what appears to be one straight forward which feels like it would have to be a wrist motion, but who am I to judge without having it in person.

A nice set of blacks to use in web design. Subtle variants off of black or white like this can really make your design look nice and modern.

This is a pretty incredible use of css grid to overlay items overtop of each other without needing to resort to position: absolute and the side effects that it brings.

        
.wrap {
  display: grid;
  & > * {
    grid-row: 1;
    grid-column: 1;
  }
}

Some sick looking icons no attribution needed.

someone has created a knock off of the ltt screwdriver and made it printable

uv now can install python for you. uv is inspired by rust's cargo, an all in one fast package and installation manager. uv is so fast and becoming feature complete, it's really changing the python packaging ecosystem.

        
uv python install
uv python install 3.12
uv python list

Switching between nvim configs can be really easy to do since they implemented the NVIM_APPNAME Environment Variable.

        
NVIM_APPNAME=nvim-lazyvim nvim

Now config will be loaded from ~/.config/nvim-lazyvim

Michael lays out some aliases in the full article.

        
alias v='nvim' # default Neovim config
alias vz='NVIM_APPNAME=nvim-lazyvim nvim' # LazyVim
alias vc='NVIM_APPNAME=nvim-nvchad nvim' # NvChad
alias vk='NVIM_APPNAME=nvim-kickstart nvim' # Kickstart
alias va='NVIM_APPNAME=nvim-astrovim nvim' # AstroVim
alias vl='NVIM_APPNAME=nvim-lunarvim nvim' # LunarVim

This is the coolest scrappy build, no printer or modeling required, no fancy tools, just straight chisel and claw hammer in soft pine. Im here for it, create the thing that you want to create.

Obsidian has a go to definition like feature, the keybind is alt+enter, I would have never guessed this one.

Very inspiring video Jason, these people faced adversity and crushed it. The market is so hard right now, its not easy to get a job, move around, and recover from a layoff. Everyone in this video just showed that you can make it in this industry right now, you just have to be willing to work harder than the rest. The bar is raising, a lot of great people are already putting in hard work to make it.

Community is Key, Theo mentioned that a lot of the companies he works with don't even post openings, they go to their community first and are often filling spots through lower risk community hires.

This is such a cool 3d printed cornhole game.

Thor always comes in hard with fire takes in his shorts. Especially if you are early in our career take this in. If you dont yet have kids, but plan to, double take this in.

Docker no Docker, what!!!

So fly.io uses Dockerfiles to deploy your app, but no docker. They use containerd to download your docker images into firecracker microvms to run your app. Firecracker is the same tech that runs aws lambda functions.

Fascinating short post on the beans under the hood at fly.io and how they scale your app globally.

This is a pretty sick design, touch screen macropad, trackball, ring, and knobs for days. I just want to play with it and see what I can make it do.

As I'm typing I think I just realized that its modular, each rectangle is it's own section!

Damn Johnny is always such a positive vibe, just getting steamrolled by this braille shit. WTF are these brand sleeping on. Happy to hear that Andy and Brian are supporting him well.

Also kinds shows what shoes are good when he goes to buy shoes cause no one will send him any and he goes with Etnies.

Johnny keep it going!

Wyatt is again inspired to leave his thoughts on the world while watching me take notes on the zettelkasten method.

Niklas Luhmann created the Zettelkasten method of notetaking, He was the Da Vinci of writing, holding a high standard that we can take inspiration from even today in this digital era.

As many times as I have seem the Zettelkasten covered, I never knew that it roughly translated to little cards in a box.

Four Principles

Mischa lays out four principles of Zettelkasten.

• Atomic - each card conains one idea
• Link - cards are linked together
• Structure - comes from the links
• Own Words - You write the note in your own words

My blog, tils, and thoughts make up my own zettelkasten. I am not going to say that mine is a perfect representation of his method, but the components are here. rather than building books, my thoughts and tils build up to blog posts. The rate of blog posts I write definitely changes with the seasons of life, as seasons change it becomes a bigger or smaller part of the season.

I really like Mischa's most important note to Write in your own words, this is what really promotes learning, and proves to yourself that you can form a thought around the topic. Up until recently I've wrote everything, but lately I've taken to copy pasting some quotes into my thoughts posts, but I think this is fine, because generally its used to build on the narrative and give me a place to talk about the linked post.

Even after switching to my hm and ht behaviors I am running into some issues where sometimes I am still accidentally hitting mods(&hm) and layers(&ht) while typing and it's been getting frustrating. My main issue has been on &ht, they are configured the same so I suspect that my pinkies just move a bit slower over the keys than my pointer/middle finger.

I just added quick-tap-ms and require-prior-idle-ms to my &hm and &ht behaviors, and a few intentionally sloppy passes through monkeytype seem to show that its working well. A few days of trying this will tell whether it was a good fix or if I have maybe gone too far the other way.

The end goal here is to be able to roll over keys faster without worrying about hitting other layers or mods.

I've been working on something, I'm gong to combine my mechanical engineering skills with my love of great keyboards and start a custom keyboard shop for building custom keyboards for exactly what people need with an emphasis on portability.

Damn Prime covers this so well from all angles. Can't overstate the importance of that last step. Look at the issues, and raise an issue if there is not one before putting in a bunch of hard work. Make sure that the maintainers are open for your changes and no one else is already working on it.

Sqlite is getting rust extensions now, and datetimes make it totally worth if if they work well and and fast, two things that don't always go together in datetime libraries

50 minutes into this episode they start a discussion on seo, and naming posts. They give Jerod the task of googling for tmux and see how high up the ranking they find Nick, expecting a number one post, and he cant even get to Nick.

The problem? Jerod is such a beginner he doesn't even know what to search, he starts with a long query about getting started with tmux, but all of Nick's videos are why you should use advanced features of tmux, but Jerod has yet to discover that these exist.

As experts we often come up wtih a post title for the exact thing that we are trying to teach someone how to do, but yet, they haven't discussed that they need that feature yet, they just want the getting started guide, or tell me why I should start using tmux. Not why sessions are better than windows, and tabs suck all together.

Definitely an eye opening conversation listening to two experts sit in shock as the beginner just gets dropped big tech articles, and Nick Jani-freakin-takis doesn't even hit the query page after page.

Also hats off to Nick for keeping a 2014 machine alive, I got you beat by a few years on my 2011 free give away machine, but I stopped daily driving that two years ago, and its purely a homelab now.

This is a wild keyboard test. This is exactly what I want to see from ticklemykeys.com. I want to make ergomechanical boards that you can use and abuse. I want you to not be afraid to take them with you or to eat a nutrigrain bar in the same room (cough 2015 mac). And If something does go out, I want users to be able to repair it themsleves, key caps, switches, batteries, and microcontrollers should be replacible parts they can get at a good cost from the shop if they have paid for a board.

Today I swapped out all of my keys that are used dual purpose for letters and layers to homerow layers. This prevents goofy things happening when rolling, and prefers-tap makes it so that keys that are rolled over get hit as letters instead of as layers. This was one of my biggest hurdles jumping into zmk, lt as a homerow key just does not behave the same as the ht/hm behaviors with tap-preferred set.

!!! seealso See previous commit where I added the hl https://github.com/WaylonWalker/zmk-config-42block/commit/9522c859cdf024a2c2b73931c130ddc907c09ffc

        
        hl: homerow_layer {
            compatible = "zmk,behavior-hold-tap";
            label = "HOMEROW_LAYER";
            bindings = <&mo>, <&kp>;

            #binding-cells = <2>;
            tapping-term-ms = <150>;
            flavor = "tap-preferred";
        };

Deep breath, the deepest you can take in

Smile, a real fukin smile not some pansy bullshit

Say I'm not going to give a shit about this when I'm 90

This fixed my bazzite update issues after the signing key was rotated recently. This team is killing it with such a great user experience.

        
curl -sL https://fix.universal-blue.org/ | sudo bash 

This one space killed my whole config and held me back from learning zmk.

The trackball spinner looks sick here. I can imagine using that spinner like a scroll wheel.

They had split ergo boards back in 1983??? WTF who the heck keeps these row stagger boards going. This board looks like endgame material, If this thing was more normal, it'd kill a whole section of the ergo mechanical keyboard industry for good reason.

This keyboard layout looks weird af the mix of column staggar and row stagger is wild. Not sure if its genius or an abomination.

That solenoid though is absolutely wild though, I kinda want one TBH.

and the clear plate with the diodes laid out on it in a herringbone pattern is a very nice touch.

today I learned that there is an accessibility feature in chrome that allows you to place a text cursor anywhere on the page. I had accidentally done this and it drove me mad that it was there.

Hard to argue this take, happy to see that its at the top. With it being such an old language its amazing that it still holds this position, and not surprising that it has warts, and thing that have turn users off from wanting anything to do with it.

timestamped in the link

This is a pretty sick set of guidelines to help you write better cli programs, I'm definitely coming back to reading this one more in depth later.

You can explicitly make a script render blocking, nothing will be rendered until this js is ready.

        
<script blocking="render" 
        src="important.js" 
        defer></script>

I've long been a user of remove.bg, and I just discovered that you can run this transformer right within your browser with no api limits.

Very interesting approach to htmx and fast api. It uses separate decorators for returning template partials and json that can be stacked to include both options on a single route. The templates are explicitly set in the decorator. Separate decorators are used for full page and partial pages. I don't see an example of full and partial pages being combined. I think the demo app must be behaving in a spa like fashion where it does not get all of the data when it calls index and index will ask for user-list.

Definitely going to keep my eye on this project and ponder on it.

        
from fastapi import FastAPI
from fastapi.templating import Jinja2Templates
from fasthx import Jinja
from pydantic import BaseModel

# Pydantic model of the data the example API is using.
class User(BaseModel):
    first_name: str
    last_name: str

# Create the app.
app = FastAPI()

# Create a FastAPI Jinja2Templates instance and use it to create a
# FastHX Jinja instance that will serve as your decorator.
jinja = Jinja(Jinja2Templates("templates"))

@app.get("/")
@jinja.page("index.html")
def index() -> None:
    ...

@app.get("/user-list")
@jinja.hx("user-list.html")
async def htmx_or_data() -> list[User]:
    return [
        User(first_name="John", last_name="Lennon"),
        User(first_name="Paul", last_name="McCartney"),
        User(first_name="George", last_name="Harrison"),
        User(first_name="Ringo", last_name="Starr"),
    ]

@app.get("/admin-list")
@jinja.hx("user-list.html", no_data=True)
def htmx_only() -> list[User]:
    return [User(first_name="Billy", last_name="Shears")]

Very interesting approach to htmx and fast api. It uses separate decorators for returning template partials and json that can be stacked to include both options on a single route. The templates are explicitly set in the decorator. Separate decorators are used for full page and partial pages. I don't see an example of full and partial pages being combined. I think the demo app must be behaving in a spa like fashion where it does not get all of the data when it calls index and index will ask for user-list.

Definitely going to keep my eye on this project and ponder on it.

        
from fastapi import FastAPI
from fastapi.templating import Jinja2Templates
from fasthx import Jinja
from pydantic import BaseModel

# Pydantic model of the data the example API is using.
class User(BaseModel):
    first_name: str
    last_name: str

# Create the app.
app = FastAPI()

# Create a FastAPI Jinja2Templates instance and use it to create a
# FastHX Jinja instance that will serve as your decorator.
jinja = Jinja(Jinja2Templates("templates"))

@app.get("/")
@jinja.page("index.html")
def index() -> None:
    ...

@app.get("/user-list")
@jinja.hx("user-list.html")
async def htmx_or_data() -> list[User]:
    return [
        User(first_name="John", last_name="Lennon"),
        User(first_name="Paul", last_name="McCartney"),
        User(first_name="George", last_name="Harrison"),
        User(first_name="Ringo", last_name="Starr"),
    ]

@app.get("/admin-list")
@jinja.hx("user-list.html", no_data=True)
def htmx_only() -> list[User]:
    return [User(first_name="Billy", last_name="Shears")]

Pinout for nice!nano boards. Note that P0.15 means gpio port 0 pin 15, they can be referenced in zmk when setting column and row pins.

        
#include <dt-bindings/zmk/matrix_transform.h>

/ {
    chosen {
        zmk,kscan = &default_kscan;
        zmk,matrix_transform = &default_transform;
        /delete-property/ zephyr,console;
        /delete-property/ zephyr,shell-uart;
    };

    default_kscan: kscan {
        compatible = "zmk,kscan-gpio-matrix";
        label = "default_kscan";
        diode-direction = "col2row";

        col-gpios
            = <&gpio0 31 GPIO_ACTIVE_HIGH>
            , <&gpio0 29 GPIO_ACTIVE_HIGH>
            , <&gpio0 2 GPIO_ACTIVE_HIGH>
            ;

        row-gpios
            = <&gpio1 15 (GPIO_ACTIVE_HIGH | GPIO_PULL_DOWN)>
            , <&gpio1 13 (GPIO_ACTIVE_HIGH | GPIO_PULL_DOWN)>
            , <&gpio1 11 (GPIO_ACTIVE_HIGH | GPIO_PULL_DOWN)>
            ;
    };

    default_transform: matrix_transform {
        compatible = "zmk,matrix-transform";
        columns = <3>;
        rows = <3>;
        map = <
            RC(0,0) RC(0,1) RC(0,2)
            RC(1,0) RC(1,1) RC(1,2)
            RC(2,0) RC(2,1) RC(2,2)
        >;
    };
};

today I learned that /dev/pts is a pseudo-tty. It amazes me how much linux is still built around things like hardware terminals.

miniserve is a sweet http server, replacement for python -m http.server. It's fast, runs off a small binary, but why would I want to use it over something that already exists on most machines, because it includes a bunch of features like qr codes, pretty themes, and uploads. I've used python -m http.server many times to transfer files from one machine to another in a pinch, like at a family members house. But what if they have an android, windows, or something not easy to get a python repl running on, you can run miniserve and upload from their device rather than hosting from their device.

The arch wiki is always full of good content, and pacman tips and tricks does not disappoint. Today I discovered this command to remove orphaned dependencies on my system.

        
pacman -Qdtq | pacman -Rns -

Interesting to see that curl is used in so many places. I often think of things like games being so windows centric and curl being so linux centric I don't even think of these things crossing paths as much as they do.

interesting, seems like such a simple way to completely remove the need of a whole other cli. No offense to anyone working on wget, but generally I use it out of lazyness or something wierd is happening and I am looking for a second opinion. Cool to know that wcurl exists and will start shipping with curl.

I recently started seeing email-decode.min.js show up on my blog posts, and I wondered what the heck ? I didn't put it there. Turns out that cloudflare put it there from pages to safely serve email addresses for me.

inspecting the page without js running we can see that the mailto email is swapped out for email protected. Neat feature.

        
❯ curl --silent https://waylonwalker.com/diskcache-as-debounce/ | grep email
<a class="decoration-pink-500 hover:decoration-pink-300 hover:text-pink-100" href="/cdn-cgi/l/email-protection#a4ccc1c8c8cbe4d3c5ddc8cbcad3c5c8cfc1d68ac7cbc9" rel="me"><span class="__cf_email__" data-cfemail="630b060f0f0c2314021a0f0c0d14020f0806114d000c0e">[email&#160;protected]</span></a>
<script data-cfasync="false" src="/cdn-cgi/scripts/5c5dd728/cloudflare-static/email-decode.min.js"></script></body>

Looking deeper into this article it looks like this feature comes from Scrape Shield and enabling Email Address Obfuscation.

fastapi comes with a concept of background tasks which are functions that can be ran in the background after a function has been ran. This is handy for longer running functions that may take some time and you want to have fast response times.

Here is an example from the docs

        
from fastapi import BackgroundTasks, FastAPI

app = FastAPI()


def write_notification(email: str, message=""):
    with open("log.txt", mode="w") as email_file:
        content = f"notification for {email}: {message}"
        email_file.write(content)


@app.post("/send-notification/{email}")
async def send_notification(email: str, background_tasks: BackgroundTasks):
    background_tasks.add_task(write_notification, email, message="some notification")
    return {"message": "Notification sent in the background"}

markdown it py running in rust claims to be 20x faster. I'll definitely look into this if markdown it py is ever a bottleneck in my performance. At first glance it appears that plugins are written in rust not python, and there is no admonition plugin, so I'll keep my eye on it for now, but I can't use it.

diskcache has a peekitem method that allows you to lookup the expire_time of a cached item without changing it. I recently used this to implement debounce for fastapi background tasks with multiple workers running. since all the workers I care about are on the same machine, but running in different processes diskcache was a great option. All workers have access to the same disk, but not the same variables in memory.

Great intro into kustomize. This helped me get started with kustomize.

Huge list of micro controllers tried and used in keeb builds.

Not gonna lie, kinda pumped about this one. I manually did one, jotted down the coordinates, opened the gcode in vim, added markers between setup/teardown and print. then added the wipe, the copy pasted the print+wipe section a bunch of times.

My printer tends to run a bit better on single prints than printing a dozen at once as it has less issues with retract start and stop.

another great use and demo of the one eyed fighting kirby

Damn, supply chain vector attacks are wild. Random guy in Primes stream was getting $40k offers to buy their open source project while in university and they have never made anything from it. What a social engineering attack this is. It would be so easy to make it look like a good deal and that the package was going to a good new owner who has real resources to maintain it.

This is almost exactly what I have been thinking about building for quite awhile. I am going to give it a good try with a few $$ in filiment some cheap switches a microcontroller and go pretty bare bones, maybe open wiring on the bottom.

Absolutely sick post. This is top tier animated blog posts. This posts demos how different queuing systems work with fantastic interactive demos.

Supply chain attacks are so big these days engineers definitely need to take these into consideration. It's wild that such a simple attack vector hit some really big applications. This particular vector is so easy to avoid. You are already hosting web content, just curl the file and self host the script, then you own it. That eliminates this attack vector all together, but doesn't completely remove supply chain attacks, the js file can still hit external apis internally.

What I see has happened in this case is that the owner of the domain polyfill.io changed. so anyone who directly linked to them got a malware injected script used.

I can only imagine the number of applicatons that are not even being maintained anymore getting hit by this. TLDR, if you are taking something to production, where you are goind to deploy it and let it run, host the js yourself. these cdns are great for prototyping, but tread with caution.

Well spoken prime. Co-pilot is a fantastic tool for templating, boilerplate, or just getting something done that is just outside your knowledge or care to know. Bit if you really want to learn something, you need to use your hands. Reading the solution only gives you the happy path. This does not give you that path that most people would take, and completely misses debugging. If you are very new and learning it's unlikely that you can write large blocks of code without running a test or a script to try it. Taking these shortcuts will hinder your ability to do it yourself.

If I were to devils advocate here, copilot is still not the greatest at the big picture, you still have to drive, and tell it I want to use python, fastapi, I want to load settings from .env. taking these shortcuts can help you move quick and get the big picture. This is especially helpful for someone who might know one language or framework and is trying a different one, but at some point you need to do the work and learn if you are going to be effective.

pyannotate is a tool that uses runtime types to suggest type annotations to use. Test runners such as pytest can be used to generate the types.

Interesting, linux supports 32 programmable buttons, special buttons that do nothing that hotkey programs like xbindkeys can pick up and run things, seems like it would be very intersting to use on a macro pad. You could easily do dynamic and complex tasks without clashing that keybinding with something else.

I'm going to look into this for my next obs setup. No matter what I try to do with the hyper key it always tends to do odd things depending on the application I am in. typically its tmux, and it starts trying to do something I don't want it to.

Create virtual environments for any Python version, without needing Python installed.

Love how Daniel just brushed off the number of netflix tv installs as small beans for curl installs. overall great episode on open source, owning a major component to technology with such a small team.

Great keyboard design guide for all the different mounting options that are normal for keyboards with pros and cons of each.

Super cool idea, kinda breaking the rules by talking about it here, but Its a cool idea to throw out posts to the core audience. Apparently its been around for quite while Dave started in 2018.

A linter for linting kubernetes manifests and help charts.

Full reference to all of the qmk keycode list. Handy reference to refer to all of the keycodes that qmk has.

Fantastic guide to building handwried keyboards. I used this guide to build my first hand-wired macro pad.

Two inspirational people in one podcast, its cool to see how adam thinks about code, css, webdev, and building businesses.

Great listen for anyone interested in productionizing python code with docker. Itamar brings up some

Don't trust base images for security, upgrade your packages. Vulnerabilties become published and solved giving the bad guys istructions how to wreck your day and these fixes wont come to your docker application for up to two weeks due to image build tatency.

For job based containers pre-compile your pyc for faster startup.

Alpine linux is probably not what you want for python. Many packages such as postgres ship pre-copiled binaries that work for most linux distributions wich use glibc, but alpine uses musl so the binaries will be incompatable requiring you to need to install a bunch of build dependencies.

Homelabbers have been some of the best co-workers I have ever had. Typically have a get shit done, If there's a way I will find it kind of mentality. If you are struggling to get a job in tech right now its tough. Starting a homelab on a pc you pull out of the trash is a good way to get some experience that you can talk about in interview questions.

Linked video has some great points!

podman requires qemu-system on

        
❯ podman machine init
Looking up Podman Machine image at quay.io/podman/machine-os:5.1 to create VM
Extracting compressed file: podman-machine-default-amd64.qcow2: done
Error: exec: "qemu-img": executable file not found in $PATH

The fix to this for me was to install qemu-system before podman machine init.

        
sudo apt update

sudo apt install qemu-system

svgbackgrounds is a really awesome resource for svg things recently featured on https://shoptalkshow.com/618/

Currently Posting is a super early postman like tui (terminal user interface). It looks so good. Darren is really getting into a groove, and textual is getting to a place that is allowing him to really make these beautiful.

I am so impressed with the progress that he has made so early, it looks so close to postman in the freaking terminal, and the ui is so good. He already has think like jump that give you a way to around the ui like easymotion plugins. Theres themes and a command line, oh man its so good. star it.

Oh I kinda like the name blogmark, as opposed to thoughts like I have chose for the same thing. Aparantly Simon beat me to the punch by 20 years on this one.

THIS! is the same reasons that I built thoughts{.hoverlink}. Simon has bee a big inspiration along the way. He defintely changed the format of my posts as I watched him build out his quote posts.

Link blogging is a pleasantly low-pressure way of writing online. Found something interesting? Post a link to it, with a sentence or two about why it's worth checking out.

Ditto! just make a post.

Kellan brings some interesting thoughts on where the internet is headed in 2024. Interestingly I see myself headed in a similar direction. Feeling like I know just enough to say fuck it and build my own platform for me to me me, from thoughts{.hoverlink} where I link and make thoughts on posts like this, to reader{.hoverlink} which is my rss reader replacement that I wanted in 2013 when it was killedbygoogle{.hoverlink}

And particular with the collapse of the social spaces many of us grew up with, I feel called back to earlier forms of the Internet, like blogs, and in particular, starting a link blog.

Ai has really had quite the two sided effect since chatgpt launched and set the world ablaze. Suddenly you can get any answer you want as a custom fit blog post for free without effort, thus killing the traffic to any of these sites.

Setting up DNS overTLS in opnsense has made my dns just a bit more secure and reliable. I recently had an outage of half the internet within my house. This also hit some of my friends and not some. It did not hit my mobile network. What seems to have happened is a dns issue with my isp not resolving some domains. This setup corrected my issue and I was back online more securely.

!!! Note I did try to setup the family resolver and found it was blocking some sites I am ok with. I decided to drop back to the vanilla resolver and let other services within opnsense control blocking where I can caontrol the whitelist myself.

Argo events is an event driven automation framework for kubernetes that can create kubernetes objects among other things based on events. I've been using native kubernetes cronjobs to kick off jobs based on a cron trigger.

For instance I am running reader.waylonwalker.com every hour, to rebuild the site and re-deploy it. It takes about two minutes to fetch every rss feed, so this is a nice application of a job compared to a web server fetching the feeds live. Now my posts may be up to an hour stale but they load fast.

Argo events takes event drien architecture to the next level allowing to be triggered by many more things, and do many more things than creating a cron job. I'm definitely thinking about dropping this in my homelab.

I am using this theme for Ulauncher in arch and it looks fantastic! One line install (read it first).

        
python3 <(curl https://raw.githubusercontent.com/catppuccin/ulauncher/main/install.py -fsSL)

Just discovered this really cool launcher from the DHH distro omakub. github.com/omakub{.hoverlink}.

Ulauncher comes out of the box looking good, supports extensions, and shortcuts like I have a lot of mac launchers have. I installed it plus a theme and in no time It was looking good and launching applications.

In the past I have leaned in on rofi for task launching, it is good. I just felt it was harder to configure to get right or look good.

This is a sick steam deck dock! With a corne built in laptop style.

This is DHH's linux startup script. Call it a distro if you want, but he doesn't. It's made to take a vanilla ubuntu install and configure everything the way he likes it. There's a number of great nuggets in here to pick up on.

No need to write bespoke configs for every essential tool just to get started or to be up on all the latest command-line tools. Omakub is an opinionated take on what Linux can be at its best.

Check it out and give it a ⭐ basecamp/omakub{.hoverlink}.

This is a really cool animated visual representation of how sine waves work, how they relate to circles, pi, rad, and how to add up a series of waves to make square and sawtooth waves.

This looks like a sweet tui postman clone. Darren is really rolling with these tui's. Cant wait to see where this one goes.

Sometimes I struggle to get my os to report dark mode to chrome, luckily there is a way to force chrome to always use dark mode.

I've never really gotten into dark reader and extensions like this. For some reason they all make websites look really weird to me and I don't really care for it. What I want is websites designed to be in dark/light to always go dark, if the designer didn't design dark just let it be light.

xxHash is an extremely fast non-cryptographic hash algorithm, working at RAM speed limit. It is proposed in four flavors (XXH32, XXH64, XXH3_64bits and XXH3_128bits). The latest variant, XXH3, offers improved performance across the board, especially on small data.

I hit an issue with markata where even though a bunch of articles were cached, the site build was still slow because I was hitting hashlib.sha256 so hard for cache keys. I was shocked when this popped up in my profiler as a significant portion of the time spent. I swapped out for xxhash and that issue completely went away.

This is so easy compared to self hosting stable diffusion yourself. It even has a nice api that you can hit with curl or python. The pricing seems competitive as well. Bookmarking this to try next time I need something like it.

Great set of tips here!

No waiting. No “waiting until tomorrow” or “It’s Friday, let’s wait until Monday” to deploy. If your deploys are so slow that deploying an hour before the end of the day is a risk, that’s a separate problem. If you’re afraid of a Friday deploy, your system is too brittle, or you don’t have foolproof rollback procedures, or you don’t have people you trust on call to resolve it. Each of these is a problem that you can fix.

This one I find interesting I think there are some industries where customers come in large waves over the weekend, and a weekend bug can not only ruin someones day off, take longer to fix, but also cost a lot of money.

Not deploying on Friday is totally what that team should be doing.

Most of us are not that team. Most of us work on small teams supporting some sort of product that Should be able to be tested and rolled back. I completely agree with Jake here, if your not willing to deploy on Friday you are not confident in your product. You need to work on better testing, rollback strategies, and feature flagging is what you need.

I've also been on teams where we don't deploy in the middle of the day or right before we leave for vacation because we are early, our customers are important to us, and we accept that we have not hit our marks yet on the ability to swiftly recover from some downtime so risk is high.

Be Fast, Practice, Hone your craft. There's a lot to be said here about honing your craft for editing text, picking up a few extra WPM, learning vim shortcuts.

Also just build shit. The more you build new and different things the more not only your text editing will just roll out, your skills to see patterns in code and architecture will flourish.

Read their bios, and the answer is always no. They loved to play, sure. They had some base talent, typically. But they also invested an absurd amount of time into that skill set.

This! is actually what turned me on to Post Malone. I remember hearing his story in how he was just known as the guy with a guitar because he was always playing it between class and everything.

This one is too funny. opening a sixel image in the terminal when logging into prod.

Tailwind calls strikethrough line-through. This caught me off guard and took me a minute to find.

Control how text is decorated with the underline, no-underline, and line-through utilities.

Maggie is a fantastic proponent to the digital gardening movement. In this article she proposes 3 ways for someone to start their own digital garden with low friction and no code.

This is one of my favorite changelog episodes of all time. I had no idea all the work that has gone into the right to repair and ifixit. They talk a lot about apple and its trend to be less repairable from unservicable air pods to serialized components within iphone.

A lot of legal talk that was far more interesting that I thought it would be. Recently winning the right to repair case against John Deere, and creating repairability scores for devices to be placed in stores like energy guide is.

Get gnome running in distrobox.

This is a pretty nice gear generator. I printed a few gears today and it worked great so far.

This is kinda sick, its a tool to clean up container images in a k8s cluster.

distrobox gives you distrobox-host-exec to run commands on the host. This is handy to get access to host level clis that you probably wouldn't want to run from the container like podman, docker, flatpak.

DESCRIPTION distrobox-host-exec lets one execute command on the host, while inside of a container.

Under the hood, distrobox-host-exec uses host-spawn a project that lets us execute commands back on the host. If the tool is not found the user will be prompted to install it.

A comprehensive community built index of anti-cheat support for linux very similar to proton, but specific to anticheat support.

ublue-os makes a github template for making your own git repo with actions that build out your own personal ublue iso.

Distrobox is so interesting and cool, I've only recently started realizing how much it can do especially related to hardware and graphics, this is quite an example that runs obs in a distrobox. I had no idea distrobox would let you connect to cameras and gpus so seemlessly, and give you a gui to work from. And with distrobox you can export so that it just looks like an app on your system.

2024 has been a wild year for infra with going "back" to on prem being made popular by @dhh. Well it looks like ahrefs saw right through the cloud trends an decided to ride the anti cloud train until it came back around to the station.

Being just a bit critical of the article it is impossible to get an apples to apples without actually running something of this scale and spending too much to find out. I cant imagine raw ec2 and ebs being the cheapest route into aws. They used no serverless tech in their article, but I digress, because I like this own your shit and build good product train.

What about People?!

This follow up does dive into the typical gut reaction that people cost a lot of money, you must account for them. You see when you hire people who are actually good at what they do, and run lean a lot of cost goes away, you have levels of management that disappear, levels of tooling that don't need to exist, departments of IT don't need to exist.

Colo's are the new hotness, and will continue to grow!

I just notices that they didn't even account for egress, support and taxes, egress costs would make a difference.

When using justfiles each line is ran separately from the last, unless you specify the file to be ran by something other than just such as bash. If you want variables to persist you need to set a shebang.

Also if you are using your script i a way that you want it to exit when it fails you need to set -e and -o pipefail. This is critical if you are thinking about using just for production scripts like ci/cd. I've hit too bugs where ci passes, but no artifacts were created issues for this exact reason.

        
foo:
  #!/usr/bin/env bash
  set -euxo pipefail
  hello='Yo'
  echo "$hello from Bash!"

This is a dope ass cheat sheet for justfiles. It's filled with good examples that are short and to the point, probably all from the docs, but anyways I need to do some like this for myself.

Damn this Tim Paul quote finishes hard and such a good point. None of the stuff around llms just work. Good ui's, front end, back end, infrastructure, product. All these things still need to exist, and in fact for ai to be good we need to still go hard on them otherwise everything will die in a heaping pile of ai slop

I’m no developer, but I got the AI part working in about an hour.

What took longer was the other stuff: identifying the problem, designing and building the UI, setting up the templating, routes and data architecture.

It reminded me that, in order to capitalise on the potential of AI technologies, we need to really invest in the other stuff too, especially data infrastructure.

It would be ironic, and a huge shame, if AI hype sucked all the investment out of those things.

— Tim Paul

PopSql looks like a very innovative product to bring collaboration to data exploration and visualization in a way you would expect from something like vscode liveshare. This looks far more appealing than a traditional BI data tool.

interesting pricing model from popsql (pronounces Popsicle). At a glance you pay for data retention, want the abiltiy to recall all the queries you ran within the last year, run at a higher frequency, you jump a pricing tier.

such a sick episode with dax.

SST's free tier will be free as long as aws allows a free tier, their free tier literally costs them nothing.

They talked about keeping SST small, the limitations that brings, but also the number of problems that just go away when you only have 3 people building. Lots of process disappears, everyone can trust everyone, no one needs to wait for approval, everyone is their own PM and just builds cool shit. They don't have to worry about big costs and making payroll because they are profitable so much higher than their costs.

If they can get through phase one of just being the go platform for a very specific audience of users, and gain marketshare, the ideas of offerings on top of this are endless.

I had no idea that you could just drop an msi installer right in steam. This worked for me, and was much easier to install pokemon tcg live in 05-2024 on ubuntu 22-04. I added the msi to steam from my downloads, hit start, failed right away like he said it would. changed compatability to proton experimental, and it opened right up.

• Download the official installer from pokemon tcg
• Add non steam game
• click show all file types
• navigate to downloaded msi
• click gear icon > compatability > proton experimental
• click play
• profit

exclude=True and repr=False is a good pydantic combination for secret attributes such as user passwords, or hashed passwords. exclude keeps it out of model_dumps, and repr keeps it out of the logs.

        
from pydantic import BaseModel, Field


class User(BaseModel):
    name: str = Field(repr=True)  
    age: int = Field(repr=False)


user = User(name='John', age=42)
print(user)
#> name='John'

Hatch be flyin.

This new release of hatch includes support for the new package installer uv which is just mind blowing fast compared to anything else we have in python right now.

        
[tool.hatch.envs.default]
installer = "uv"

The other features are cool too, check them out. I'll probably be using the test runner, but I've been waiting for the uv support since uv launched.

I had a boot issue on my sons fresh ubuntu 24.04 install and journalctl came in clutch.

        
journalctl -p 3 -xb 

• -p 3 gives me priority 3
• -x gives me extra catalog information when available
• -b gives me the current boot.

This Thread saved my son's ubuntu 24.04 install. His was failing to start with the following error.

        
Gdm: GdmSession: no session desktop files installed, aborting...

https://twitter.com/_WaylonWalker/status/1785825677079441482

        
sudo apt install --reinstall ubuntu-session

Super neat tool to convert html to markdown

        
>>> import html2text
>>>
>>> print(html2text.html2text("<p><strong>Zed's</strong> dead baby, <em>Zed's</em> dead.</p>"))
**Zed's** dead baby, _Zed's_ dead.

It even plays nicely with rich.

        
from rich.markdown import Markdown
from rich.console import Console
import html2text
console = Console()
md = Markdown(html2text.html2text("<p><strong>Zed's</strong> dead baby, <em>Zed's</em> dead.</p>"))
console.print(md)

Fantastic interview with Uncle Bob, really it was more of a 2 way conversation. So many of Uncle Bob's takes have been taken out of context, its cool to hear some of the well actuallys, and the experience behind some of the reasoning.

Imagine waking up to a $1,300 for running an example project! That sounds like peanuts for a cloud bill but for an individual trying to learn that hits my monthly budget real hard.

That's what happened to Marciej, make sure you check out the full article and give them a 👏 on Medium if you have an account.

The more I see things come out about aws, the more it makes me sick, and confirm my feelings that I cannot possibly use them for a side project without some real $$ planning to come out of it.

Yes, S3 charges for unauthorized requests (4xx) as well[1]. That’s expected behavior.

They offer no DDOS protection against 4xx or 5xx requests against your bucket. Absolutely bonkers that you have ZERO control over this.

This response just feels absolutely gross.

I notified the AWS security team. I suggested that they restrict the unfortunate S3 bucket name to protect their customers from unexpected charges, and to protect the impacted companies from data leaks. But they were unwilling to address misconfigurations of third-party products.

It sounds like this guy followed some default instructions for an example site, HOW MANY OTHERS have done this or will do this? And AWS has no response other than to take thier money.

After contacting them he was able to get it cancelled, but this is no guarantee. We've seen other cloud vendors stick users like this with a few thousand dollar bills after cutting their bill to 5% of the original.

AWS was kind enough to cancel my S3 bill. However, they emphasized that this was done as an exception.

2024 is wild on the cloud hosting front, own your shit or be careful.

A full list of standard Accept types.

This is a handy reference.

pydantic has a nice built in email validator EmailStr

It requires an optional pydantic dependency

        
pip install email-validator

Then you can validate email addresses.

        
from pydantic import BaseModel, EmailStr

class Model(BaseModel):
    email: EmailStr

print(Model(email='contact@mail.com'))
#> email='contact@mail.com'

This is my go to rich response container for clis written in python. It creates a nice box around the content on the screen and provides some nice separation in the output. It can be overdone, but comes in clutch when looking for that print statement in a long output.

This page shows how to customize your fastapi errors. I found this very useful to setup common templates so that I can return the same 404's both programatically and by default, so it all looks the same to the end user.

        
from fastapi import FastAPI, Request
from fastapi.responses import JSONResponse


class UnicornException(Exception):
    def __init__(self, name: str):
        self.name = name


app = FastAPI()


@app.exception_handler(UnicornException)
async def unicorn_exception_handler(request: Request, exc: UnicornException):
    return JSONResponse(
        status_code=418,
        content={"message": f"Oops! {exc.name} did something. There goes a rainbow..."},
    )


@app.get("/unicorns/{name}")
async def read_unicorn(name: str):
    if name == "yolo":
        raise UnicornException(name=name)
    return {"unicorn_name": name}

This post sat in draft for months. I stumbled upon it again and found great success returning good error messages based on user preferences. the default remains json, but if a user requests text/html it will be an html response, and text for application/rtf or text/plain

Joe has a sick cli.labs site for deploying tui applications.

        
ssh cli.lab.joe.codes

html can preserve newline \n characters by styling an element with white-space: pre-wrap;