I was talking mostly about side projects. I don’t have much time for them right now. Thanks to LLMs, I can spend those few hours a week on doing instead of reading what is the best way to do X in ever changing world of web front-end frameworks. I just sit down, ask: “how is it usually done?”, tweak it a bit and finish.

Example: I have published an app on flathub a while ago. Doing it from scratch is damn complicated. “Screw it” is what I would say in pre LLMs era after a few hours ;)

True and not true at the same time. Using agents indeed often don’t work, mostly when I’m trying to do the wrong thing. Because then, AI agent does not say “the way you do it is overly complicated, it does not make any sense”, but instead it says: “excellent idea, here are X steps I need to do to make it happen”. It wasted my time many times, but it also guided me quickly though some problems that would take hours to research. Some of my projects wouldn’t have been finished without AI.

Everything comes down to proper function naming. If it wasn’t clear what function should return, then it was not named properly.

That hits me like something a teacher tells you in a coding class that turns out to be nonsense when you get to the real world.

In a company I work in, we have “no comments policy” for at least ~10 years now and we are not planning to change that. It’s not just theory, we work like this in practice and purpose of each part of code is perfectly understandable just from variable names, file names, namespaces, function names.

The biggest problem with comments is that they can become outdated. If you change code but forget to change comment you introduce very dangerous situation where they become not only not useful, but also misleading.

If you rely on variable names, you’ve got a single source of truth, one thing to change at a time. Information updates itself.

To that end the company is developing a “Pay Per Crawl” system, which would give content creators the option to request payment from AI companies for utilising their original content.

So Cloudflare is not as much “saving the Internet”, as just becoming a middleman between LLM training companies and content creators. Which I believe has a potential of being a true goldmine in the future.

ofc I could even send raw api requests, but sometimes it’s good to have a nice GUI that “just works”.

Specifically I’m looking for something that could handle not only text responses, but also attachments, speech recognition and MCP support.

Taking advantage of the fact that this thread became popular, question to all of you guys: do you recommend some other open source LLM front ends?

I’m glad you are posting links you find interesting, that what I come here for, to see links that people find interesting.

What I don’t like is the form of it. You are pasting links to news aggregator inside news aggregator. This form kills discoverability: in other communities, you can scroll through list of topics and more or less learn just from headlines what is happening. In this community - you can’t. It also kills discussion - when I comment, I need to paste link to the article I am referring to, because otherwise it would be impossible to tell which story I am referring to.

I might of course ignore it, but I believe it is not neutral to the entire community. Many such posts in a row might suggest to new users that mostly bots are in here, in effect discouraging them to stay. That is why I think we should not allow such links in here, but I’ll leave this decision to moderators and I’ll accept it whatever they decide.

It looks like this when I enter this community:

As long as they are free and open source, I don’t care.

+1, I came here just to paste a link to it

Using Lemmy as a blog is an interesting idea, because it allows to distribute your posts to other users timelines easily, but does it do well in terms of cost of self hosting? How much more computational work does enabling federation require?

Honestly, I thought Meta talking about Fediverse integration was just marketing bullshit. Are they really doing it? 🤔

For start definitely Linux Mint. It is stable, has strong support, works out of the box.

I’m ok with auto generated content, but only if it is clearly separated from human generated content, can be disabled at any time and writing main articles with AI is forbidden

Wikipedia is not made to teach people how to read, it is meant to share knowledge. For me, they could even make Wikipedia version with hieroglyphics if that would make understanding content easier

I love being in EU

Object oriented languages can be either class based or prototype based. Java is class based, JS was back then prototype based

This clearly shows how chaotic development of javascript really was.

It has always been like:

1. Design a language for small browser scripts
2. People suddenly start using it somewhere else
3. It turns out it does not fit to where people try to use it
4. Hundreds of frameworks appear to fix this, milion standards appear, people fight for years to work out a single go-to solution
5. Return to point 2