2
points
tommy_axle
Joined 78 karma
- tommy_axle parentI see a service like this as being in the ip lookup API category (like ipinfo.io) but I wanted to mention that for this (and IP lookup, captcha etc) I would expect that if the service is down then you allow the registrations then review later, and not simply prevent all registrations.
- Ok so taylorswift is reserved but taylor_swift and realtaylorswift can be used? It seems like impersonation would still be a problem.
- An aside: it looks like there is a certificate error for https://certkit.com/ as it's for *.mscertkit.com (this was on Chromium + Linux)
- You want it to be as close to deterministic as possible to reduce the risk of the LLM doing something crazy like deleting a feature or functionality. Sure, the idea is for reviews to catch it but it's easier to miss there when there is a lot of noise. I agree that it's very similar to an offshore team that's just focused on cranking out code versus caring about what it does.
- Technically writing calls is also taking the downside.
- Not the OP but yes you can definitely get a bigger quant like Q6 if it makes a difference but you also can go with a bigger param model like gpt oss 120B. A 70B would probably be great for a 128GB machine, which I don't think qwen has. You can search for the model you're interested in on hugging face often with "gguf" to get it ready to go (e.g. https://huggingface.co/ggml-org/gpt-oss-120b-GGUF/tree/main). Otherwise it's not a big deal to quantize yourself using llama.cpp.
- There was more also going on in that time-frame: several interest rate hikes, no fix for section 174 changes by the end of 2022. Maybe someone will pinpoint whatever had the largest impact in a detailed study.
- Not sure, most were probably created just for this hackathon so I'd expect few if any. It's just a good way to see how far one can take it with vibe coding. It's easy to crank out smaller apps these days so marketing and distribution will be more important going forward.
- One way to gauge is by taking a look at some of the projects submitted for Bolt's contest at https://worldslargesthackathon.devpost.com/project-gallery
- This piece is also covered by a bunch of other cli/tui agents (like codex-cli and opencode) allowing you to switch between Claude and other models (comes in handy depending on the task) so it really all depends on the setup you like. As mentioned in the sibling comment there are ways to get it to work with Claude Code too.
- Most likely "In my experience" and "business analyst"
- To a non-developer, or no code review, couldn't the AI model also generate buggy code that then made it's way to production and deleted data just the same?
- npm package registry definitely down. Happened to be doing a package install and it kept 504'ing
- Wasn't there something when this went into effect about the mid-year being the start so it is 10% in years 1 and 6?
- Neat. Did you have any thoughts on json5 or hjson?
- Working on some starter templates for going from plain English requirements to a ready-to-use applications in a few minutes. Are you a small team looking to go from Excel to something a bit more streamlined but would like to avoid a time-consuming and costly process? Schedule a live demo today. https://genatron.ai
- That raises an interesting question. Are all of these agencies using the build from the Play/iOS stores or is there a build based on the audited public repo?
- For using requirements, much sooner. See Genatron.
- Yes it was only those two missing but it might have been specific to Linux (desktop). When on my phone it renders properly.
- Is there supposed to be an icon next to the "Featured Products" and Leaderboard headings? Just pointing out that it seems to be a missing font/icon box for me.
- Yes to gathering requirements but for everything else it might be possible. See Genatron AI
- As a PM if you're gathering requirements and building too have a look at genatron.ai
- This is one that can sneak up on you even when you're not intentionally exposing a port to the internet. Docker manages iptables directly by default (you can disable it but the networking between compose services will be messed up). Another common case this can bite you is if using an iptables front-end like ufw and thinking you're exposing just the application. Then unless you bind to localhost then Posgres in this case will be exposed. My recommendation is to review iptables -L directly and where possible use firewalls closer to the perimeter (e.g. the one from your vps provider) instead of solely relying on iptables on the same node
- It's not only a few mins when deploying. Often you want to test the production build locally and it feels like waiting forever. Especially since clicking around in an app for the dev build is slow during cold use. I think it has gotten worse as I don't remember build times taking so long in next versions 5 thru 9 or so.
- With sqlite3 -backup There's an online backup api
- Good to hear. Even with edge you can have a request that blocks or does something silly so IMO it should have had node from the beginning. A common use case is verifying a session or authorization for an entire area.
- The thing is that even on a simple site with html and css while creating it there might be issues that you can quickly catch something that assists you... tools. The page looks broken? Oh, being able to lint that html or css will quickly find any issues stemming from that but hey those are tools to introduce. Not a build tool as in generating an artifact in this case but often checking is part of that pipeline to avoid errors.
I agree that it's gotten pretty crazy for certain things over the years but there's a good balance in there somewhere and it's only natural to advance technologically speaking with the addition of tools.
- Not a standard per se but nanoid seems to fit the bill. Widely implemented.
- Yes, but you would also need to guarantee that content that does 3rd party stuff can't get injected.