#programming · textlog

I remain ambiguous. This morning I used pi+gpt5.6 to create a color scheme based on evergarden.moe. It works, and I haven’t checked how well it stay true to the style used in vim-bundled schemas. Before I wouldn’t have done this, and now I can. But I don’t know if it is good. github.com/…/evergarden.vim

How do you define slop in relation to ? For me, it is how quickly you get to an unmaintainable state- the acceleration into something that becomes difficult to work. I talk more about it here: hellodavid.dev/blog/slop

Not that I think we're close to a canonical definition (as if we'll ever all agree on one) but my team has defacto adopted the definition of code that the developer responsible for it doesn't understand, approve of and essentially has now honestly reviewed. Basically anything which leads to cognitive debt.

I agree, I don't think that there will ever be one canonical definition, but I like this definition as well, as it places the onus on the actual person creating the slop. Thank you for sharing!

ai coding feels like it has reached diminishing returns for me. luna is my default orchestrator, routing work by complexity. in practice, usage lands around 16% luna, 48% sol, 30% astra, and 6% terra. at this point, i care less about each new frontier release. the system already finds a capable model for the job. the frontier is moving beyond my day-to-day work, so i’m happy to settle into predictable models for a while.

I’ve found the same, albeit with Opus — it’s hard to find uses for Fable. I don’t need a model that works for longer or works on harder problems, I need one that consistently produces simple, easy to understand, code; which at this point none of them do.

I've been there since I settled on a combination of GLM, DeepSeek and MiMo for planning, implementation, and verification flows. I'm a lot more interested at this point in having cheaper models that can consistently do the work I need. It's been a while that I hanven't noticed any improvement when I try new frontier models. I only notice they are very expensive.

…

same here, for almost an year. At this point, I could easily live with models that were released 6-12 months ago and continue doing so. also possibly the reason why the current focus has been around agentic workflows or agentic anything, than the model capabilities.

Coding probably yeah, but still they are bad at generating novel information, new ideas and such. They’re also not very good at seeing the bigger picture, they will narrowly perform the task but they’re not taking into account future revisions, which ends up in refactor after refactor in order to move forward.

I find myself building bespoke tooling for my own and workflows these days, not only because of the obvious reason of software being so much cheaper to build, but also because there are SO MANY options for solving any given problem that it feels impossible to evaluate all of them. If it's not a crucial piece of tech for a work problem then it often feels like paving my own way is both cheaper mentally (kind of amazing that's true these days) and probably a better fit.

..then of course I have to contend with - "Is this going to be a maintenance nightmare in the future?" and then I still usually conclude "no, the maintenance will be aided by the same approach that the construction was". It's interesting how dramatically and quickly the scales have changed in this equation - and for such a wide range of software.

…

For example, I am looking for a coding agent coordinator with a particular (but not all that insane) set of requirements. I used both Claude and Perplexity to show me options and break them down by fit against my requirements. I got completely different lists! No overlap! And no perfect fits.

It will become increasingly harder to find niche markets for software when those people can just build it themselves exactly how they want it. In the end, people want to accomplish tasks, and the AI is going to be building the software on-the-fly exactly for that task, why do we need generic software that may or may not be a fit?

…

Yeah. It makes me think that protocols for exchanging data between apps will become even more important in that landscape. But maybe, not - maybe each exchange is another bespoke AI-driven implementation?

Protocols, infrastructure, containers, those kinds of things. I was actually playing around with an idea for a minimal markup language + runtime that can be utilized by AI to spin up apps more easily, and brainstorming it with ChatGPT, I have a feeling something like that will appear soon to complement MCP.

I needed to process 20,000 json files & 20,000 JXL files. I'm pulling together a website from them all. Anyway, started with emacs elisp to handle it all, just because I've not done that before. Went kinda ok, and I'm gonna use org-mode to export it all to a website (done that bit loads of times before). In the end, I ended up breaking out my C programming skills (late 80s->nowish). It was a delight. Nothing more refreshing than writing raw code, that doesn't rely on 3rd party code/libs.

Looking for a new daily-driver language. Criteria:

interpreted (e.g. can use #!/bin/bang)
batteries-included stdlib (json, subprocesses, fs, etc.)
not sh, not perl, not python, not javascript

Shortlist:

Any other ?

I use , extended with as my daily driver. Since it's implemented in Go, it's straightforward to extend with anything from Go's standard library (or third-party ecosystem) that isn't already incorporated. If you haven't used an array language before, then you get the added bonus of learning a ton about an interesting, powerful area of programming.

Not familiar with array languages but it looks… interesting. Thanks for the suggestion.

I sometimes use , but I use V-lang as well. It is kind-of interpreted. I’ve never heard of rye. Looks nice.

I did not know V has scripting support. Thanks for the suggestion.

It does not, but fakes it well, since it is fast and compares built binary with script mtime to automatically recompile. The downside of using it is that you end up with a binary, but the upside is that it is blazing fast.

…

Love the pipes in revo, reminds me of elixir

Babashka

cool new for ! autotag action will enrich your post with relevant hashtags so you don’t have to do it manually. It uses an to do it so it is quite good. Try it!

How do you avoid the creation of too many tags? Does it check existing tags in order to favor them? Or it tags each post independently of how other people are tagging theirs?

I simply tell the do fewer than 5 . But it does exceed them often.

I mean, a profiferation of tags in the system, not on a given post. I mean, maybe you would want to use the same tags for the same topics in different posts. I'm not sure if I'm clear enough.

Ah I understand now. No I’m not passing the current tags in the prompt, but we have a tag aliases system in case we want to group tags with similar meaning. For example I’ve grouped into also — we also use WordNet to canonicalize tag names, stuff like become #tag or becomes .

Cool. If you run every post via this auto tag feature, then the process is vulnerable to prompt injection attack. Isn’t it? Do you have any special protection in place? I guess yes. Curious.

It doesn’t run on every post, you run it yourself via more → autotag, and it’s safe, it doesn’t have access to anything other than the post content and we simply use its response. Or do you have any other case in mind?

I just tried and i do believe it is possible to craft malicious prompts. I can share an example. How do i DM you?

Check the IRC at the footer or you can send to ‘abuse@textlog.cc’ so I find it.

Sent email. .

…

Thanks! Fixed it by passing it through OpenAI’s moderation endpoint first, so it now blocks at least the problematic/illicit ones now before reaching autotag.

Makes sense. Cool.

Which do you think is ?

Clean Code

I had so much pointless discussions about this ... but when you say "Clean Code" , do you mean the book or the general idea that such a thing exists?

…

The book, but actually ideas, like tiny functions and names. Aposd is much deeper book. Hilarious read

I assume this is not original at all at this point - but (extreme) DRY is not a good practice. My experience is that a bit of copying yourself and your colleagues in the code base often make maintenance of the code base easier.

Many have adopted the term WET (write everything twice)

Getters and setters for everything. Certainly, if you want an object to be doing more than just getting and setting a value right now, get them in there! Planning to add functionality later without breaking code? Stop it! That should break code! If I'm consuming a library and have been using this setter method for 2 years, I don't want it to suddenly introduce brand new validation, caching, or logging behavior!

Are getters and setters on everything even considered a good practice? To me it screams junior developer who hasn’t yet fully understood encapsulation and abstraction — if you’re just exposing getters and setters as the interface then what is the class/module/etc providing beyond what a standard variable does…?

I was followed by here and wonder if this the one of ari fame? If so, I’m thankful for their work. Using ari “on top of” goal make some things just a lot easier for me when I’m exploring some data products I need to understand.

Hi, yes, I develop ari. I'm so glad it's proven useful for you! Any feature requests or changes you'd like to see? I'm curious what bits of ari have made it easier for you to explore data, too. UI or additions keep tempting me, but I haven't settled on what I want there. If you have thoughts on that front, I welcome them.

…

The two main things I use are the duckdb support and of course the built-in resty client. Not needing to shell out to curl just makes life easier. I should probably add that I’m mainly prototyping and explore things that later eventually will be developed further by others in our organization, so using array langs where I don’t need to type a lot is a big plus for me:) I’ll let you know if I need anything!

Thanks, that’s helpful. Those are the two things I use the most, too. One development tool I use frequently is the util.ac function (for “auto-complete”) that takes a glob pattern and returns all the globals whose names match. I didn’t want to make people have to define or update their Goal lib path, so in recent versions of ari I’ve been packaging all of the Goal libs and ari's into a libs FS that you can use with import.

created a poll:

, do you have warnings as enabled at ? How is your so far?

yes0%

no0%

0 voted

One idea that I've been thinking about is an or where it is a app so you get a new file everyday right except the app keeps track of all the files for you. the idea is that anytime you type anything it it basically makes a commit and use so anytime you type something it commits automatically after a quick pause after you stop typing.

That sounds ! Could also create summary titles with for the filenames in addition to their date so you can find what you were writing about easily. Also full text search.

Yes! Great idea. Not sure how I would implement it but the main idea is that you should feel free to delete text through out the day because everything is tracked in a git repo. 🫣

This #tome about #Tcl/Tk has me re-evaluating my plans to move to the woods to be a carpenter hermit. Now I want to move to the woods and be a programmer hermit.

Hello to friends. Holla for tools. 👨‍💻 : , , and Pragmata Pro 💪 : NotePlan 🎧 : Overcast 🧠 : Allume 🖋️ : Mark Two (EDC), Pen Type-B 🌳 : Anisota

I'm doing experiments with short attention-span students and code. Starting with python, and, oddly, now working in BASIC. As a C programmer of 40+ years, I find this aesthetically and pedagogically displeasing, but I am delighted to see them getting fully absorbed in writing 130 PRINT "You have been eaten by a bear!" : GOTO 900. Yes there is spaghetti, but it's their spaghetti and they seem to take delight in debugging it. Python will come later.

replied to(deleted account):

I was talking about this with some alumni students, specifically their experience learning python with us a few years earlier (these are college students who are not really using python any more). What it seems to boil down to is that with python you have the feeling that there is this huge forest of solutions to any problem you are having -- but you will never know the whole forest. BASIC feels like you "know" it after a day or two of playing around, evidently.

good ! it’s time for and .

replied to(deleted account):

Good morning to you both 🌞

i miss real coding

me too, but at the same time i'm enjoying getting things done, otherwise i would get lost in details like bikeshedding naming variables and making the code look pretty, and little would be done.

…

whats stopping you from it?

Nothing, I still write code by hand for most personal things, but for work, it seems AI has been adopted uniformly.

…

I feel you there.. and fear the craft will vanish.

The essential work of is building the theory of the problem + solution. Theory then enables creation of working code. In the end, the code is necessary, but it is not sufficient. Theory is indispensable. (cf. Peter Naur, 1985)

join the communityorbrowse more notes