Maybe it's time to get out my Cowlishaw Rexx book again…
Astral to Join OpenAI
511–520 of 951 posts
Re: Astral to Join OpenAI
#512What strikes me most about this acquisition isn't the AI angle. It's the question of why so many open source tools get built by startup teams in the first place. I maintain an open source project funded by the Sovereign Tech Fund. Getting there wasn't easy: the application process is long, the amounts are modest compared to a VC round, and you have to build community trust before any of that becomes possible. But the…
Most likely, because it is less money :-p. But also because it is less known and harder, as you already mentioned. Personally, I'm based in Mexico, and I would never have thought about trying to get nonprofit funding for a community project, nor would I know where to start to get that.
Re: Astral to Join OpenAI
#513Time for https://zubanls.com/
Re: Astral to Join OpenAI
#514Earlier quoted context omitted.
They are not. A very large proportion of their revenue comes from developers. A large proportion of their marketing and product work is aimed at developers. You have to work really hard to not see this. Just look at what Altman and Brockman tweet about. https://xcancel.com/gdb https://xcancel.com/sama/
All the various APM companies are implementing "Assign to agent" flows. The various foundation model providers will be satisfied getting a subscription for 10% of total comp of a developer, instead of pocketing 60% of the total comp completely replacing them? The only thing that could prevent this is lack of ability to execute, like how Uber wanted to replace drivers with FSD vehicles.
It's not about what they wish would happen, it's about what they think will happen. In my view they are acting precisely like they believe they will be making a proportion of developer pay by making them more productive rather than replacing developers. I think they understand that the alternative doesn't really work out for them or anyone.
Even if they believe that their systems will eventually tank employment and replace developers rather than augment meant, the fate of Astral doesn't matter at all in that scenario because a) nobody has a job, and b) you can build your own uv replacement for $20.
Re: Astral to Join OpenAI
#515Earlier quoted context omitted.
Just a tiny project with over 100 million downloads every month, over 4 million every day. No big deal. Just a small shop, don't overstate its importance. https://pypistats.org/packages/uv
Sure, but if tomorrow uv and ruff ceased to exist, we could all go back to any number of other solutions.
I see Apache and MIT license files in their GitHub. What's to prevent the community from forking and continuing development if the licenses change?
Re: Astral to Join OpenAI
#516Re: Astral to Join OpenAI
#517Ah shit, back to poetry
Re: Astral to Join OpenAI
#518I (along with many others) always thought that Astral being VC backed is going to lead to a future disappointment for the community.
I at least thought we'd get a few more good years before the rug pull.
Re: Astral to Join OpenAI
#519Earlier quoted context omitted.
pixi? https://pixi.prefix.dev/latest/ Have not tried it too much yet because I was pretty content with `uv`, but I've heard lots of good things about it
It literally says on the homepage/the page you linked, that pixi is just a frontend for uv in the background to interface with PyPI and the project TOML files.
that isn't quite true, it defaults to conda packages (and so supports non-pypi things, its main advantage), I believe only when you are mixing these with pypi packages does it then also use the same resolution library backend as uv (and indeed directly at the rust level) https://pixi.prefix.dev/latest/concepts/conda_pypi/
Re: Astral to Join OpenAI
#520Earlier quoted context omitted.
Why would they diffuse their attention when competition is not sleeping?
Their attention apears to already be quite diffuse if buying a python package manager is an item anywhere on their agenda. Also, once again, if the tool did what it was promised -- it wouldn't even be a diffusion of attention. The entire schtick is that software engineers are being replaced and that you can just run a model to create the product for you. Unless, of course, the thing does not do what is promised.
> Their attention apears to already be quite diffuse if buying a python package manager is an item anywhere on their agenda.
According to the blog [0], their whole monorepo is in Python, their models are obviously trained using Python, their experiments are written using Python and core and CLI of their Codex is written using Rust. Uv brings both Python and Rust expertise. You’re talking nonsense because of your blind hate of LLMs. Even though I agree that they’re capitalizing on the fear of SWE being redundant.