I have only had positive experiences using pydantic tools. They keep iterating on the pydantic validation tool and are planning on a v3 at some point, and pydantic-ai has been a superb addition to the AI engineering toolkit; which makes me hopeful and confident about the work they will do with httpx2. I look forward to explore their eval and observability tools.

I do wonder about pydantic being a VC-backed company and how much of a load-bearing column for the Python ecosystem they are. HN has similarly discussed uv at great length. For what it's worth, Mitchell Hashimoto has an interesting point of view [1] on this matter: "I think VC is unfortunately (for the cons) the best tool we have as an industry right now to produce the highest amount of accelerated, public good R&D right now".

May we only reap the benefits of the situation.

[1] https://lobste.rs/s/oznirn/redis_cost_ambition#c_66fiqh:~:te...