JodieBenitez

> Stop Making TUIs

No, I don't think I will :)

It's great for admin tools, alongside CLIs.

It's portable, it's fast, it's discoverable, it has zero bullshit effects, it's as secure as my shell, it's efficient.

And what is this point about agents making it easy to generate GUIs ? They are equally good (if not better) at generating TUIs using any of the great frameworks for this.

Nice tips there ! :)

> taking nothing away from you

Hosting is not free.

Indeed ! So I don't see what the problem is here...

I'm pretty sure there is a silent majority of Go users who don't want/need/know about generics.

I don't think you thought this through. Learning to write is also very difficult and time consuming. Also you missed the point.

Here’s a little analogy: I’ve been composing electronic music for 35 years now, and some people still think it’s just a matter of pressing buttons, that it’s easy, that anyone can do it the same way, that I do this because I don’t have what it takes to learn to play a “real instrument” and that I’m not a “real musician”.

You get used to it, and eventually you stop caring. Too bad for them if they’re more interested in the process than in the work itself.

More than 25 years ago, the Web wiped out a lot of the things I loved. May it be devoured! But I have my doubts: there’s probably not much left of it anyway.

I like Go with agents too but:

> This is very different from Python, where training data is polluted (I presume) with tons of code written by non-software engineers and demonstrating many different ways of doing the same thing.

Counter-example: agents with Django-related stuff. Excellent output.

I don't understand the quality criticism. My 2026 professional code has much better quality than my 2025 code. Better algos, better performance, better comments, less bugs, better and more tests, better documentation, etc. What do people do with LLMs to get crap out of it ? Unless I let the agent run an entire monstruous plan and never review anything and never steer, I can never get output that could be described as slop. Bonus point: even after 25 years, I learn a lot with them, just because of their search capabilities applied to a context.

Damn, what a downer...

Hobby scale you said ? It doesn't have to be that complicated.

See sibling comment about fetch modes in coming 6.1 (https://docs.djangoproject.com/en/dev/topics/db/fetch-modes/), you can have a FieldFetchBlocked.

I have twenty years projects, no breakage when upgrading major versions. What a treat in the web app world !

> our queries ended up being several lines long

Which is... perfectly normal for non-trivial needs.

> I ended the project with having some proper aggregates, only for that to be thrown out of the window by the guy after me.

How is that a Django problem though ? Sounds like a skill issue on your successor.

I get what you say, there's plenty of debates about ActiveRecord vs. AnythingElse, but in the end this one has its use and obviously has served many of us just fine. Different strokes... you know the drill.

Having used Doctrine, hard disagree. Never again.

> BEGS you to cause n+1 problems

select_related, prefetch_related. n+1 problems be gone.

Most loved feature, for me: No dramatic changes, just sane and careful evolution.

"Googling" is pretty dead to me. Frontier model agents give much better, tailored to your case, answers. And it's faster. And there's no ads.

Yes, it just a way to remove responsibility in case of problem. Very bad attitude. Whether the code comes from an autocomplete, a LLM, the documentation, a Stack Overflow response or has been 100% typed with your fingers doesn't matter: you're the one in charge.

I thought the same... huge security risks (or huge security costs ?).

I propose we get rid of tiktok videos and let the kernel devs use AI. That should balance.

Ah ! So it was you... I have like 300 mp3s from your blog. Also, Fluokids.

> I think I'd remain happy if they just stayed at the level they are now, maybe bit faster and cheaper

Definitely ! It's so useful as is, even without spitting a single line of code, as an analysis and documentation tool, it's already great. Now a faster and/or cheaper version: perfect. Make it runnable on commodity hardware: icing on the cake.

> destroying books is in itself a morally repugnant act

It's a routine act in any publisher stock management.

> Shouldn't OSS be the first to benefit?

Yes, it should. And not just for code. Think better, more accurate documentation, quick start guides for both developers and users, reduce yak shaving effort, etc.

> should OSS be a bedrock of carefully curated code untouched by AI slop?

Slop is slop. There's plenty of human slop in OSS.

> I don't get the anti-AI sentiment.

There are valid social, ecological and maybe philosophical concerns.

But on the technical aspects... I don't get it either. It's a wonderful tool. even if it didn't write a single line, the fact that it can analyse a source and explain things to me in context is a game changer.

Different strokes for different folks, I guess. Where I live photos on a menu are mostly used for lower tier restaurants. It's not a good sign.

Apart from the AI/Not AI, I dislike photos on menus. It can be deceptive once the dishes are on the table or it can completely spoil the surprise.

Do it the EU way: ask the question, then ignore the answer.

The sovereignty thing is a theater. Many french unis use Google cloud because they're broke and can't maintain in-house services, and none gives a damn.