The "exactly" part is the problem.

Just last two weeks I had to slap Fable, three times, to stop writing 1000-2000 lines of defensive code... because of DB columns I just forgot should be NOT NULL. That was it. Nothing else. I told it that, boom, -4800 coding lines: gone.

LLMs defend the status quo and they regularly lose sight of everything bigger than the current PR they are working on.

I too am gradually making peace with the fact that LLM-maintained code does not have to be 100% readable for humans.

But this is not about readability. It's about the data model. So one concession I am willing to make is: don't care too much about the code _BUT_ manually curate the data model. So far: small wins on iteration turns and code volume producing. Too early to tell but for now I am happy with the results.

And? Why should they not use Cloudflare?

What do the EU investors do outside of demanding guaranteed profit with zero risk for them? How many are ready to absorb 10-30 years of new infrastructure investments and play the long game?

If the EU wants to compete, well, here's the global market, right that way. ==>

Get to it. Nobody is actually stopping you.

But if you have no appetite for risk, don't complain when one day you realize you don't have EU alternatives.

Huh? Can you show source on the "it will not compact"? Very interesting.

"With Astra, we’re introducing a new way for Codex to preserve and retrieve context when the context window fills. Historically, models have used compaction to summarize work during long sessions, such as when debugging complex issues or tackling large refactors. Each compaction can leave out details about why a fix failed or how a component behaves. In Codex, Astra can keep notes across context windows, preserving accumulated details without repeatedly compressing them into a single summary. Earlier context windows remain searchable, so Astra can find requirements or test results from previous messages and tool outputs—even if that information wasn’t captured in its notes. You can enable this experimental feature in your Codex config.toml, (opens in a new window) and it will become the default for Astra in the coming weeks."

https://openai.com/index/gpt-6-astra/

This clearly is Codex-specific, not as much a feature of the model (though obviously they probably have trained it to be great at working with their own tools). Sounds somewhat similar to pi-observational-memory I'm using with Pi.

> Or to search the contents of a JSON file, you would maybe use `grep`, or perhaps `jq` and a well crafted query

Try https://github.com/tomnomnom/gron.

How do you do the proxy thing?

Meridian[0] for Claude Code, CLIProxyAPI[1] for everything else

[0] https://github.com/rynfar/meridian

[1] https://github.com/router-for-me/CLIProxyAPI

I use OpenCode with my ChatGPT subscription, which is officially supported on both sides and a damn good deal (OpenCode Go is great too).

OpenCode has “providers” for many (many!) other services, but these are almost all unofficial and against ToS (Anthropic being famous for ban-hammering people).

But proxying via Claude CLI is OK? Or also not?

Sounds like their problem, not something a SaaS product should dance around.

Yet they kind of did. I've limited participation in my libraries with GitHub's setting that nobody who made an account in the last 6 months can do anything in my repos (after some misguided hustler thought they're an easy target and posted an ad).lp

Time's marching forward though. Wonder what will happen after a few more months. We'll have bot spam accounts that are no longer as fresh.

It's also a great opportunity for character development. Use it for that, and not for trying to overbook yourself to 300%.

You don't owe the world anything at all. If you're conscientious, then give a little -- here and there. Don't turn it into an unpaid job.

All being said, I have most sympathy towards engineers of all people. They don't get to call the shots, they do as they are told. 99% of the time by an extremely out-of-touch managers.

Managers and executives though? Those I do blame. Surely at this point it should be blindingly obvious their current strategy is not working.

Those engineers also chose to work in US big tech for the ridiculously high salaries that come with it. Having to deal with the problems resulting from that scale is just part of the deal.

Rust's disadvantage in adding time is mostly a one-time initial investment, in my experience.

After you get comfortable and learn the important idioms, you go through the development loop quite quickly.

Compilation/linking speed can still suck though.

What kind of workplaces you people have if you can code "for fun"?

I am tired of only ever being gauged by immediate usefulness. You hiring?

Also when you emphasize "exactly", you are not giving any actual reasons.

I code for fun and income. I don't stay at engagements where I can't invest personally in what I'm doing, and I don't hire people who don't enjoy what they're doing as they're doing it. Anything else feels punitive for everyone involved.