lucideer

5 ms·

31.
by lucideer 3mo ago
Ah! I misunderstood - they were talking about forgejo & I skimmed too quickly & read it as Tangled being a forgejo host (akin to e.g. Codeberg). Turns out it's their own stack - very cool, I'm more interested now.
32.
by lucideer 3mo ago
I get you on unix philosophy (I'm not into it myself but I respect those who want that) but uv, for all its extra features, does actually do that particular "one thing" a lot better than pip. You've summarised many of pi
33.
by lucideer 3mo ago
> not in the least because that is completely unenforceable You've completely missed my point. It's ALL completely unenforceable. That doesn't mean the directive is useless - it's a guideline for contributors towa
34.
by lucideer 3mo ago
A lot of people pointing out that Proposal A is untenable/ at philosophical odds with the widespread upstream LLM usage, etc. etc. I think that's an unfortunate framing & I think A would be the best option for the opposite rea
35.
by lucideer 3mo ago
<unrelated side note for the author> I would never normally mention this, but I'm reading this repo on mobile & it seems Forgejo doesn't wrap .txt files on mobile. Any reason not to use .md for a new project in 2026?
36.
by lucideer 3mo ago
Wherever you fall on that debate (I have my opinions but they are not strongly held), a bigger red flag here for me is the `pip install ...` opener. The fact they later have a `uvx ...` example in the docs just makes defaulting to pip even
37.
by lucideer 3mo ago
I guess it depends on your source of info but I've found what came before was pretty comprehensively & accessibly documented, at least as well as systemd, which - while well documented - suffers from sprawl & overwhelm of the d
38.
by lucideer 3mo ago
First I agree that bash scripts should only be used for quick or small (& ultimately mostly personal unshared) tasks. There's absolutely no need for it to be either "dirty", & as mentioned in my post, brevity has no m
39.
by lucideer 3mo ago
I've read through all of the examples in the article & they all seem to serve to sole purpose of turning readable multiple line code into one-liners. One-liners are a cool little artifact of early shell culture & are sometimes
40.
by lucideer 3mo ago
The solution to social engineering can't be to remove useful features - you can socially engineer people to do literally anything, you can have someone walk to their bank, withdraw cash & fly to you with it with a dating scam: ther
41.
by lucideer 3mo ago
This. I'm a developer, I've published Android apps, & I've still managed to lose access to a device that had dev settings enabled purely because remote adb enablement is a (extremely fiddly) toggle that happened to be off
42.
by lucideer 3mo ago
Who uses the word size to mean mass? The typical colloquial analogue to mass is "weight" - which nobody would ever conflate with "size".
43.
by lucideer 3mo ago
I'm not fully sure what this means but perhaps you're right about Python given its use by data engineers, academics, SREs (though they're mainly Go these days ... which also lacks fully stable native SIMD support). Otherwise
44.
by lucideer 3mo ago
> Every developer should know at least that much SIMD. > This [...] applies to any programming language. Support for SIMD instructions varies by programming language This is a very pedantic nitpick because this article is good (&a
45.
by lucideer 3mo ago
I think this is a case where it's a net positive to have both. If Codeberg's opinionated mission isn't aligned with your personal approach to code, you definitely shouldn't host there, but it's genuinely hugely valu
46.
by lucideer 3mo ago
You're definitely referring to fully autonomous harnesses as I suspected - these are generally quite well suited to VM deployment (quite possibly in the cloud) if you're concerned with sandboxing as they don't typically run i
47.
by lucideer 3mo ago
It might be overused elsewhere but I don't think its use here is inappropriate given it's qualified: it doesn't say "frontier models", it says "frontier open models".
48.
by lucideer 3mo ago
That's a pretty oddly narrow & specific definition of what LLM harnesses are for & it certainly doesn't match my experience. In fact I'm genuinely struggling to think of any scenario where an agent CLI would need to
49.
by lucideer 3mo ago
> allowlists are not there for security I don't honestly believe that the allowlist isn't (naively) intended to be there for security. It's certainly not a very useful ux for providing LLM guidance: prompts, memories &am
50.
by lucideer 3mo ago
It's an allowlist so not-git would be disallowed by default. Renaming it to a binary that is already listed in the allowlist should fail due to a cryptographic binary signature mismatch.
51.
by lucideer 3mo ago
I've used VMs because dealing with tool access is infinitely simpler & less fraught than tailoring sandbox profiles, but ultimately using sandbox-exec has yielded better workflows (if & when I've gotten an appropriate prof
52.
by lucideer 3mo ago
This is very doable - it's a shell you're creating within the cli so you have a lot of control over the environment. If you look at tools like sandbox-exec & apparmor, they specifically target binaries, not string input. E.g.
53.
by lucideer 3mo ago
I've run agentic CLIs in a sandbox (using MacOS sandbox-exec) & in VMs - both solutions have an inherent advantage over any integrated sandboxing by virtue of their independence (even before we get into configurability). Claude CLI
54.
by lucideer 3mo ago
> - Sensible permission system - when the agent runs git diff && rm -rf /, what do you think will happen in your current coding agent? It will treat it as git *. Maki uses tree-sitter to parse the bash command and figure ou
55.
by lucideer 3mo ago
I agree generally but I'd moderate this to say this is an "anti- the current state of AI" post insofar as none of the issues listed are inherent to LLMs, they just seem to be ubiquitous in the current generation of agentic CL
56.
by lucideer 3mo ago
Good to see Codex as a datapoint: I haven't used it. I can confirm though that this is certainly what Claude does, with the addition that Claude's adherence to its own string-match based allowlist seems even less deterministic tha
57.
by lucideer 3mo ago
This is a good summary of the dangers of using agentic clis, but the title & general focus on opencode is odd for two reasons: 1. Most obviously & importantly this is a complaint without a straightforward suggested alternative. A si
58.
by lucideer 3mo ago
This is a generally great article with a lot of truths, but the title & overarching narrative throughout is a little off & feels like the author is trying to shoehorn one story into another. In truth, the phenomenon they're des
59.
by lucideer 3mo ago
I'm not sure I see the value in "showing the team it's possible". I would presume the team are intelligent enough to be well aware that it's possible, given tradeoffs. And as it's clear those tradeoffs have bee
60.
by lucideer 3mo ago
The gp feels like less of a targeted individual criticism & more of a general musing about a trend. Nobody is saying this author hasn't demonstrated discipline & won't maintain this project, but the statistical average

More ›