sReinwald

5 ms·

1.
by sReinwald 11d ago
Yes, but this is exactly the sort of thing that should be a user-level configuration. A personal scratch directory has nothing to do with the repository itself and doesn’t belong in a repo’s .gitignore. If it's only needed for one part
2.
by sReinwald 24d ago
OpenCode also defaults to ask for external_directory, which applies to bash commands touching paths outside the workspace. sre on archlinux 10.0.0.193 ~/opencode-test took 0s ﬌ opencode run "use bash to touch an empty t
3.
by sReinwald 24d ago
"That's the model running a command nobody asked for, and OpenCode doesn't stop to confirm." Well, yes, because you invoked OpenCode with the --auto flag telling it not to stop for confirmation.
4.
by sReinwald 29d ago
That is simply incorrect. In Germany, truth is a complete defense as far as libel and defamation cases are concerned.
5.
by sReinwald 2mo ago
Framing this as whether he can convince you the mistype was accidental gets it backwards, and "would I believe X or would I think Y" is preponderance, not reasonable doubt - entirely different standard. Reasonable doubt doesn&#x27
6.
by sReinwald 2mo ago
I'm from a region bordering the palatinate region and I'll absolutely back you up on being able to understand them pretty well. There's an amish guy who sometimes pops up on my TikTok feed, and it's always fascinating to
7.
by sReinwald 2mo ago
> at a breakneck pace for no apparent reason, lately. This isn't surprising to me at all. The World Cup is on, and it draws attention away from politics. This has been a pretty common observable pattern for as long as I can remember
8.
by sReinwald 5mo ago
You clearly haven't played WoW in a decade, judging from this comment. > Even if you buy the expansion just to get the "feel" on how it was, it's impossible. You don't buy previous expansions after a new one laun
9.
by sReinwald 5mo ago
Yeah, totally valid for things like D&D - I split that out too, for the same reason. Work and homelab overlap constantly for me though. I make heavy use of periodic notes, so even something like recipes would get linked in weekly logs f
10.
by sReinwald 5mo ago
I keep one vault for almost everything because the cross-domain links are where the interesting stuff happens. I studied sociology, political science, and media & communication before moving into IT, and I'm still actively interest
11.
by sReinwald 6mo ago
Third-party MCP servers create at least two different security problems. One is prompt/context injection through the tool output. The other is the much more conventional risk of executing untrusted code with transient dependencies on y
12.
by sReinwald 6mo ago
"It's not a complaint about water. It's a complaint about the wetness." If capitalism requires constant vigilant government intervention to prevent monopolistic practices, anti-competitive markets, and asymmetric bargain
13.
by sReinwald 6mo ago
Any single LLM-edited comment reads fine in isolation. The uncanny valley kicks in when you read thirty of them in a row and they all use the same "it's not X, it's Y" construction. The problem isn't that LLM prose
14.
by sReinwald 6mo ago
> Overall this was a worthwhile assist. I believe (totally understandable) anti-AI animus is coloring a lot of these replies. That, and hindsight bias. People know the second version came from an LLM, so it's automatically "fla
15.
by sReinwald 7mo ago
Interestingly, when I apply the "simply repeat the prompt" technique [1], Sonnet 4.6 on the website got it right every time, both with and without extended thinking. Not repeating the prompt got a mix of walk and drive answers. I
16.
by sReinwald 7mo ago
You are essentially amortizing a single dose over 6 months by micro-dosing and receiving a clearly sub-therapeutic dose. While that may work for your specific usage, it doesn't change the unit economics for someone who needs the standa
17.
by sReinwald 7mo ago
I suspect a big reason for why Mounjaro is still fairly expensive here in Germany (I pay nearly €400 for a 10mg Qwickpen - a 12.5mg Qwickpen is nearly €500) is due to health insurance not being allowed to cover them for anything but diabete
18.
by sReinwald 7mo ago
It would probably be interesting, but if you are not overweight, the appetite suppression will likely make this not a very healthy or very fun experiment. I started at 5mg in October, and even on that smaller dose I had to force myself to e
19.
by sReinwald 7mo ago
One thing I have noticed on Mounjaro is a (at least subjectively) significant decrease in impulse spending / buying random crap off Amazon. I have ADHD and that has been a real problem for quite a while - even with ADHD medication (Elv
20.
by sReinwald 7mo ago
Like I said, the first examples are fairly trivial, and you absolutely don't need an LLM for those. A good agent architecture lets the LLM orchestrate but the actual API calls are deterministic (through tool use / MCPs). My point
21.
by sReinwald 7mo ago
> But can't you do the same just using appropriate MCP servers with any of the LLM providers? Yeah, absolutely. And that was going to be my approach for a personal AI assistant side project. No need to reinvent the wheel writing a T
22.
by sReinwald 7mo ago
You could, but Claude Code's memory system works well for specialized tasks like coding - not so much for a general-purpose assistant. It stores everything in flat markdown files, which means you're pulling in the full file regard
23.
by sReinwald 7mo ago
Disclaimer: Haven't used any of these (was going to try OpenClaw but found too many issues). I think the biggest value-add is agency. Chat interfaces like Claude/ChatGPT are reactive, but agents can be proactive. They don't n
24.
by sReinwald 7mo ago
Not saying the frontier models aren't smarter than the ones I can run on my two 4090s (they absolutely are) but I feel like you're exaggerating the security implications a bit. We've seen some absolutely glaring security issu
25.
by sReinwald 8mo ago
IMO the value and differentiating factor is basically just the ability to organize them cleanly with accompanying scripts and references, which are only loaded on demand. But a skill just by itself (without scripts or references) is essent
26.
by sReinwald 8mo ago
I don't see how that makes it uniquely viable in France. Germany has something very much like this too. And we've had it for nearly 13 years. > Since 31 August 2013 companies which operate public petrol stations or have the pow
27.
by sReinwald 8mo ago
It's been a few days, but when I tried it, it just completely bricked itself because it tried to install a plugin (matrix) even though that was already installed. That wasn't some esoteric config or anything. It bricked itself rig
28.
by sReinwald 8mo ago
From my experience: TDD helps here - write (or have AI write) tests first, review them as the spec, then let it implement. But when I use Claude code, I also supervise it somewhat closely. I don't let it go wild, and if it starts to m
29.
by sReinwald 10mo ago
I think point a) is actually backwards and potentially counterproductive to the petition's stated goals. The petition explicitly highlights maintainer burnout and the "unausgewogene Verantwortungslast" (unbalanced responsibil
30.
by sReinwald 10mo ago
Strange framing, isn't it? Bariatric surgery shows 25-65% significant regain rates depending on definition and timeframe. And regular dieting is even worse. Nobody would frame that as a safety issue. That's... just how weight loss

More ›