Hacker Newsnew | past | comments | ask | show | jobs | submit | anthony_franco's commentslogin

Connect your coding tool to a browser (like with the playwright MCP) and just let it iterate on its own until it matches the design you want.


That’s a great idea. I haven’t tried connecting to a browser with something like Playwright before.


I agree with everything except for Kamal. I'm happy to have someone else handle the server side maintenance. Maybe once my service grows so huge that handling it myself makes sense moneywise, but starting off that way is overkill when there's such affordable alternatives.


Another thread just said basically the opposite, that it was easy to host multiple sites, and sold me on Kamal. I get not wanting to do it, but I just spent 10 minutes writing a basic HTML page and 90 minutes trying to get GitHub pages to do SSL and I’m still not sure I got it.

So if I have to do a little brain damage to configure Kamal but then can push sites to it easily? I’m in


I thought 2024 was very underwhelming for Rails - not really big advancements besides Kamal and Kamal should not have been part of rails to start with. If rails team wants to work on docker deployment tool, they can - just don’t call it rails feature


Genuinely, I don't understand this take. 2024 brought huge additions like Solid Cache, Solid Queue, and Solid Cable, the stable version of Strada (Hotwire Native) which completed the current vision for the Hotwire stack, in addition to Kamal (let's not sell that short, because whether you personally like it or not, it's a fairly robust tool for what it's meant for) and other niceties. Even if you don't like one/all of these, I don't understand seeing these as small advancements of the framework, particularly in a single year.


Lightweight umbrellas.


I find that if you check when the LLM's data was updated, and then choose a framework/language version release date that's before that, then it's pretty good.

In my case I use tailwind css a lot and found that sticking to v3.4.3 created the best output from LLMs.


You built a product first and then are searching for the market for it? That's tough. If you do it in reverse, find the market first and then build a product for them, it's 100x easier. Trust me.


The llm just returns a method name and arguments to pass it. Your code is in charge of actually executing it, and then replying with an answer.


There’s some YouTubers that release weekly videos about AI news. I suggest you search around for ones that suit your interests.


Friendster was one of the first big social networks but the amount of users slowed it down. That made everyone move over to MySpace/Facebook.

If I remember right, Friendster had a feature that told you how you were connected to every other person (friend of a friend of a friend, etc.). And it was this feature that made the pages slow to load.


So then that's not bad code, it's bad product leadership


The 4-Hour Workweek - motivated me to start my own business and now I have a lot of flexibility in my life.


How closely did you model the plan?


http://bookmooch.com/

It's a great way to trade your books with others. And you can pick and choose from others' collections. Plus it's a good way to get rid of any extra books you have lying around.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: