News for hackers. Stuff that matters.
Re:GPT‑6.2 Astra and Sol by Someone1234 on Tuesday November 17, 2026 @01:20PM (#50535803)
My big hope for this release is - Have they got their massive over-engineering under control?

Say what you will about Opus 5.5, but it has a good "gut instinct" for how large/complex to design the solution. The previous Astra just absolutely loved pointless abstractions, excessive numbers of files, and just over-complicating very simple things within an inch of their life.

And, yes, you can prompt your way out of anything. But needing a default prompt to tell OAIs models to work better isn't preferable, particularly when that sometimes gets lost in sub-agents and the over-engineering returns into parts of the work.

Re:GPT‑6.2 Astra and Sol by holzhaker1 on Tuesday November 17, 2026 @01:27PM (#50535916)
yeah i feel like with astra they went all in on 3D modelling which encourages "just one tweak, just one more tweak" which makes it detailed enough, but in coding that same behaviour just sucks.
Re:GPT‑6.2 Astra and Sol by inexcf on Tuesday November 17, 2026 @01:58PM (#50536430)
And it is absolutely crazy then that Astra sucks at 3D modelling compared to Opus 5.5. It is not even comparable and Astra needs 10x the usage to do it.
Re:GPT‑6.2 Astra and Sol by scandinavian on Tuesday November 17, 2026 @01:31PM (#50535988)
I've been using both sol and astra to write some python and rust, and they generate the sloppiest code I've seen in a long time. I would be embarrassed to commit anything they write. The rust code especially was terrible. Had it implement a reader/writer for a binary format and it was just littered with magic numbers, terribly named variables, very compact code visually etc. It produces working code, so I guess the GPT models are mostly for people who let AI take the wheel 100%.
Re:GPT‑6.2 Astra and Sol by hermitShell on Tuesday November 17, 2026 @02:00PM (#50536453)
Do we want LLM generated and maintained code to look similar to human generated and maintained code?

I've always used AI tools to write very constrained and small functions and modules, and then stitch it all together myself. So far. Over the last year it's become evident that many people let the AI drive 100% and there are proponents (DHH notably) that this is the way.

I'll be interested to see how compilers and LLM's compare at generating assembly over the next decade. Assembly offers enough abstraction, why do we need C? \s

Re:GPT‑6.2 Astra and Sol by andoando on Tuesday November 17, 2026 @02:18PM (#50536709)
I think even for LLMs it's going to be much easier to analyze code, and especially its intent with proper naming vs trying to figure out what a bunch of assembly is doing. A function called binarySort is alot less tokens and communicates what the function is attempting to do vs 100 lines of assembly
Re:GPT‑6.2 Astra and Sol by logicchains on Tuesday November 17, 2026 @02:08PM (#50536561)
You're using it wrong. If you want it producing clean code, you should be telling it to clean up and refactor the code, not just treating the first thing it spits out as the final product.
Re:GPT‑6.2 Astra and Sol by SaltyBackendGuy on Tuesday November 17, 2026 @02:27PM (#50536842)
Just like humans honestly.
Re:GPT‑6.2 Astra and Sol by water-drummer on Tuesday November 17, 2026 @02:49PM (#50537146)
Make no mistakes
Re:GPT‑6.2 Astra and Sol by seizethecheese on Tuesday November 17, 2026 @02:26PM (#50536821)
I looked into Sol 6.1's performance by language recently (data from a head to head battle system built on OpenCode) and Sol 6.1 is actually has a decent win rate when it uses Rust (https://micjm.com/blog/gpt-6.1-sol-long-tail-languages ) so it probably has the same problem with any language.
Re:GPT‑6.2 Astra and Sol by vbezhenar on Tuesday November 17, 2026 @02:50PM (#50537149)
> magic numbers, terribly named variables, very compact code

Sounds like I'd like it. I'm really tired of claude introducing pointless constants everywhere.

Re:GPT‑6.2 Astra and Sol by uw_rob on Tuesday November 17, 2026 @01:31PM (#50535999)
Honestly I don't even know if you can prompt your way out of these types of behaviors - stylistic tics and judgement seem to be extremely difficult for the models to overcome. It seems we are tied to what the labs do unless we want to dive into and get dirty with models yourself. In retrospect with time it becomes obvious that the previous generation of models couldn't be prompt engineered to do what the current generation of models do. There's just no way you will prompt engineer Opus 4.8 to write as well as Opus 5.5. I'm guessing with time we'll see that there was no way you could prompt engineer Sol 5.1 to be as token efficient as Sol 5.2
Re:GPT‑6.2 Astra and Sol by contron on Tuesday November 17, 2026 @01:39PM (#50536111)
I have found no way to overcome these kinds of things myself, even when showing the model examples of what not to do. Theres not really a good incentive for the labs to train them not to. If the model can overengineer, you spend more tokens. If you then prompt for the model to slim things down, thats more spend. If anything should come from the next generation of models, I hope its a sense of brevity and TIGHTNESS.
Re:GPT‑6.2 Astra and Sol by pants2 on Tuesday November 17, 2026 @01:35PM (#50536052)
Lol I made the mistake of having Astra design and build a complex system without any input (experimentally). Seemed functional at first and shipped it internally and it had dozens of complex bugs, so I started digging into the architecture and it was a monstrosity! Definitely 10X more complex than it needed to be.
Re:GPT‑6.2 Astra and Sol by myworkaccount2 on Tuesday November 17, 2026 @02:01PM (#50536456)
I was about to fall into the same trap. I wanted to try the "smart" model (astra), and made the mistake of thinking benchmark smartness is the same as designing sensible systems.

After having it write around 13k loc i had to tell it to remove most of the complicated bs, we went down to 3k loc.

Re:GPT‑6.2 Astra and Sol by incrudible on Tuesday November 17, 2026 @01:45PM (#50536204)
I have limited experience with OpenAI models, but I have to say that Opus 5.5 has not improved on the terrible writing style, and I am struggling to prompt my way out of it. It always needs an extra pass dedicated just to rewrite all the prose.
Re:GPT‑6.2 Astra and Sol by teach on Tuesday November 17, 2026 @01:53PM (#50536346)
I've also given up on allowing Opus 5.5 to write basically any prose or technical documentation.

My favorite model for this at the moment is Gemini 4 Argon, which writes very naturally and isn't too expensive. If you're not using Cursor or some other harness that allows you to choose non-Anthropic models, then I don't know what to suggest.

Re:GPT‑6.2 Astra and Sol by jpardoe on Tuesday November 17, 2026 @02:05PM (#50536521)
I’m sorry but this is crazy, I think you’ve forgotten just how bad 5 was. It’s not perfect by any means but it’s a clear improvement.
Re:GPT‑6.2 Astra and Sol by afro88 on Tuesday November 17, 2026 @02:00PM (#50536444)
I have been bitten by this 3 times and each time Opus 5.5 brought it back to something sane. The last time I left Opus in a /goal to get behaviour matching the docs, and it decided on it's own that a rewrite was the best path (and it was, and it worked)
Re:GPT‑6.2 Astra and Sol by pettijohn on Tuesday November 17, 2026 @02:02PM (#50536481)
Oh my god yes. 6.x has been terrible at overcomplicating things. I have to prompt it to find the simplest solution that meets the requirements and then ask it to simplify and tell it to STICK TO THE PLAN. In addition it feels like they overfit for persistence. I created a new work tree and forgot to copy .env files. I caught codex hammering away and docker and psql commands and stopped it and asked it to explain what it was doing. So now my agents.md says "if documented tooling fails STOP and seek human assistance."
Re:GPT‑6.2 Astra and Sol by pennomi on Tuesday November 17, 2026 @02:40PM (#50537041)
Until benchmarks punish agents for writing dirty code, it will continue. Right now “completing the task” is more important than “don’t damage the codebase”.
Re:GPT‑6.2 Astra and Sol by seizethecheese on Tuesday November 17, 2026 @02:09PM (#50536582)
[delayed]
Re:GPT‑6.2 Astra and Sol by seizethecheese on Tuesday November 17, 2026 @02:24PM (#50536806)
[delayed]
Re:GPT‑6.2 Astra and Sol by ibejoeb on Tuesday November 17, 2026 @02:51PM (#50537162)
I have experienced Sol 6.1 writing tests to check values of environment variables. Not the structure, or range of values, or anything useful like that. Like `assert(DATABASE_HOST == 'localhost')`. Gotta wonder how something like that was learned.
Re:GPT‑6.2 Astra and Sol by contextfree on Tuesday November 17, 2026 @02:25PM (#50536811)
Man, I haven't tried any Astra yet but I find Opus 5.5 rather bad about this already.
Re:GPT‑6.2 Astra and Sol by jeffybefffy519 on Tuesday November 17, 2026 @02:49PM (#50537143)
This is why i stopped using GPT 5.6 onwards... in my view GPT 5.3 Codex is when they peaked for coding.
Re:GPT‑6.2 Astra and Sol by riazrizvi on Tuesday November 17, 2026 @02:58PM (#50537253)
I don't mind it too much. Refactoring the bloated poorly integrated code is one way I stop my skills atrophying.

It also reassures me about the longevity of my career. I have sustained a theory that there simply isn't the data to teach these things to be really good, so I love to get the evidence to back it up.