@freakboy3742 @glyph @jacob @sgillies I won't really try to convince anyone of anything, I'll just share my most honest answer on how I think about this stuff right now. I don't want to charge people money to use our tools, and I don't want to create an incentive structure whereby our open source offerings are competing with any commercial offerings (which is what you see with a lost of hosted-open-source-SaaS business models).
@freakboy3742 @glyph @jacob @sgillies What I want to do is build software that vertically integrates with our open source tools, and sell that software to companies that are already using Ruff, uv, etc. Alternatives to things that companies already pay for today.
@freakboy3742 @glyph @jacob @sgillies An example of what this might look like (we may not do this, but it's helpful to have a concrete example of the strategy) would be something like an enterprise-focused private package registry. A lot of big companies use uv. We spend time talking to them. They all spend money on private package registries, and have issues with them. We could build a private registry that integrates well with uv, and sell it to those companies.
@freakboy3742 @glyph @jacob @sgillies I'm hesitant to share specific examples like that, because we may build something totally different! Or decide that the specific idea isn't good! But part of what we want to do is experiment. So while I worry about people anchoring on any specific idea we share, I also sense and understand a desire for those specific ideas / examples :)
@charliermarsh @freakboy3742 @jacob @sgillies I have heard you say as much on podcasts, and this *could* be quite a sustainable path forward, but the specifics of what and how are actually pretty important. you want to go up against Artifactory and friends, but it’s quite possible that in the course of doing so, you discover that users don’t care about Python so much as Docker, and that makes you heavily prioritize a linux-container workflow, which cuts off native app development at the knees
@charliermarsh +1 on this. The experience with private registries is a bit of a misery. I would be intrigued to see Astral build a "Vercel for Python".
@charliermarsh
I love what you're doing with uv - if you could replace devpi with something super simple and very fast, that supports all the relevant peps, you could eat the internal registry market from the bottom up and iterate from there. We would pay for that. You could call it ir - Internal Registry ;)
@freakboy3742 @glyph @jacob @sgillies
@charliermarsh @freakboy3742 @jacob @sgillies this is just a nightmare scenario that I have made up based on a niche edge case I happen to personally care about a lot, not particularly a likely one, but there are a million like it I could throw out. Until the community knows where this is all going, it’s going to be a point of concern. And I realize that trying to find that path really quickly is hard! I do not read any bad faith into what you’re doing.
@glyph asking Charlie to share specific business plans and ideas is super unfair. There’s real advantage to getting to test stuff out and iterate privately; developing a business model in the open is risky as fuck. It’s fine if you want to withhold judgement but I think right now at this early stage you’re being unfair. @charliermarsh @freakboy3742 @sgillies
@glyph would like to get your thoughts on caching proxies & private package hosting one day
@jacob @charliermarsh @freakboy3742 @sgillies later in the thread I did make it explicit that there’s no way to answer these concerns immediately. I will not feel comfortable until the business plan is proven out, that does not mean I expect that it should or even *could* be proven out immediately, and I am sorry if I implied otherwise!
@jacob @charliermarsh @freakboy3742 @sgillies I realize “the specifics are important” could imply the imperative “give me the specifics!” But that’s not what I was trying to say :).
@glyph @jacob @charliermarsh @freakboy3742 @sgillies I have the position that it does not matter and the PSF/core python should act as if astral turns into a burning pile of flames. (Which to be clear, I neither hope nor expects). A lot of open source software was build by failing companies and we still benefit from all the work that went into it.
@glyph yeah you did and I’m sorry for jumping on you before reading the whole thread. I’m just feeling kinda frustrated at the way this conversation is going and I took it out on you. Sorry again. @charliermarsh @freakboy3742 @sgillies
@jacob @charliermarsh @freakboy3742 @sgillies honestly if Charlie said “here is exactly what we will be doing for the next five years, I promise” I might be considerably *more* suspicious, there’s no way that could be committed to publicly right now
@glyph @freakboy3742 @jacob @sgillies Thanks, I genuinely appreciate that you're thinking about this seriously. To some degree what you're describing is just an unavoidable part of trying to build a business -- we might learn something new and our direction could change. It'd be wrong of me to try to convince you that it's impossible. The best I can do is just be honest about how I view the company and what we want to do vs. what we don't.
@glyph @freakboy3742 @jacob @sgillies For example... It's hard for me to imagine a future where we're not extremely focused on Python. It's the community and ecosystem that we know, it's huge and growing, it's where we we're known and where we know users / companies, etc. But again: that's just me being honest about how I view our roadmap / strategy today (and how I describe it to investors too). I think it's the best I can do right now, won't try to convince you of anything more.
@freakboy3742 @glyph @jacob @sgillies But the core of what I want to do is this: build great tools, hopefully people like them, hopefully they grow, hopefully companies adopt them; then sell software to those companies that represents the natural next thing they need when building with Python. Hopefully we can build something better than the alternatives by playing well with our OSS, and hopefully we are the natural choice if they're already using our OSS.
@charliermarsh @freakboy3742 @jacob @sgillies I certainly hope you succeed. I think there are ways that this could go bad, but I don’t think it *needs* to go bad. There are some significant challenges on the way there which need to be addressed as they come, there’s nothing to do or say right now, today, that can fully address those concerns
@charliermarsh Thanks for the response - and totally understand the limitations around sharing specific plans. The general direction definitely sounds promising; but there's a big gap between here and execution :-)
In terms of messaging - I've missed where you've said these things; the one place it isn't laid out (AFAICT) is the Astral blog/website (beyond high level “we believe in fast Python tools" stuff). A "Why should you trust uv/ruff/Astral?” post would go a long way.
@glyph @freakboy3742 @jacob @sgillies Thank you I appreciate this
@charliermarsh @freakboy3742 @jacob @sgillies in a way I feel bad being critical! You and the team are doing great work, I don’t begrudge you for taking money for it, and my concerns are more like “when the investors push Charlie out for poor EBITDA performance in a down market cycle, how is the new management going to be incentivized to maintain things in a way that serves the community’s interest”, which is obviously at least a few years away