manmal

24 ms·

511.
by manmal 1y ago
That sounds a bit like using a jackhammer to drive in a nail. Wouldn’t it be smarter to enumerate edge cases and test all permutations of those?
512.
by manmal 1y ago
That sounds worse than letting an LLM dream up tests tbh. I wouldn’t consider grooming a huge number of tests for their usefulness after they‘ve been generated randomly. And just keeping all of them will just lock the implementation in plac
513.
by manmal 1y ago
What kind of bugs do you find this way, besides missing sanitization?
514.
by manmal 1y ago
> Tests are the source of truth more so than your code Tests poke and prod with a stick at the SUT, and the SUT's behaviour is observed. The truth lives in the code, the documentation, and, unfortunately, in the heads of the dev tea
515.
by manmal 1y ago
It’s definitely geriatric.
516.
by manmal 1y ago
ChatGPT lists clickable sources in a lot of nontrivial queries. Those sites don’t even need to pay OpenAI for the traffic (yet). If you ask „what’s happening in the world today“, you might get 20 links. How is this worse, exactly?
517.
by manmal 1y ago
Maybe Grok was just pushed by their political influencers. It’s a republican, anti-woke LLM after all.
518.
by manmal 1y ago
It can still give you an overview with a few choices and a link to the prepared checkout page, and you enter your CC details yourself.
519.
by manmal 1y ago
Temu ad just below the fold, no thank you.
520.
by manmal 1y ago
How good is it at editing files? Many write/replace errors?
521.
by manmal 1y ago
Nothing's wrong with using autocomplete in addition to agents.
522.
by manmal 1y ago
It would be interesting what this functional reserve is, right? The microbiome perhaps, or intracellular minerals? Some other thing we haven’t even identified?
523.
by manmal 1y ago
I thought FSD has to be at least level 4 to be called that.
524.
by manmal 1y ago
Try giving it a repomap, eg by including it in CLAUDE.md. It should pull in less files (context) that way. Exactly telling it which files you suspect need editing also helps. If you let it run scripts, make sure to tell it to grep out only
525.
by manmal 1y ago
Oh I meant pay money as in, pay per view or cinema.
526.
by manmal 1y ago
Vibing a movie will be possible, but people also won’t pay money for watching that. It will be great for script writers to be able to make an MVP of their movie first, and someone will still need to produce this into an actual movie.
527.
by manmal 1y ago
Doesn’t SQLite have its own in-memory cache? Is this about having more control re cache size?
528.
by manmal 1y ago
Where I live we have this saying about building houses: Build one for your enemy, then build one for your friend, and then build your own one.
529.
by manmal 1y ago
Containers bring their own set of problems, there are some examples brought up in this thread, mainly around communication with the host OS. I‘d argue the reproducible parts of vibe coding (agentic engineering) setups are just text files. M
530.
by manmal 1y ago
Aren’t those _way_ more expensive per token?
531.
by manmal 1y ago
Have you tried setting up multiple git worktrees as well? I wonder if that doesn’t solve most of the issues.
532.
by manmal 1y ago
Also check out claude-trace, which injects fetch hooks to get at the data: https://github.com/badlogic/lemmy/tree/main/apps/claude-trac...
533.
by manmal 1y ago
vibetunnel.sh perhaps
534.
by manmal 1y ago
The SOTA models will always run in data centers, because they have 5x or more VRAM and 10-100x the compute allowance. Plus, they can make good use of scaling w/ batch inference which is a huge power savings, and which a single develope
535.
by manmal 1y ago
Indeed, the town I‘m living in was 80% destroyed during WWII and that still shows in its finances. It’s amazing how long major disaster affects a region. Big drug issues, highest cancer rate in the country etc
536.
by manmal 1y ago
Yes, and cities with lots of trees are way more livable due to this. Planners in our town seem to hate trees with a passion, thank god we‘re moving away from this concrete desert.
537.
by manmal 1y ago
What do you tell Claude to do with repomix? Get an overview into the context?
538.
by manmal 1y ago
LLMs have no idea what’s really interesting to you as a person. You can try to include all the topics of interest in the prompt, but the result won’t be desirable.
539.
by manmal 1y ago
If it’s not weighing you down that you have to do put in this effort later, then sure.
540.
by manmal 1y ago
Tools like honeycomb (and observability 2.0 in general) do so well exactly because they solve the very real problem of drowning in log data.

More ›