No kidding. I decided to check how it’s going. Apparently they have raised 192k which means that with 5 weeks left they have to raise 800k.
Another way of looking at it:
They have raised about 200k in 75 days.
200000/75 = 2666 per day on average
Between today and July 8th there are 37 days
2666*37 = 98.6k
So unless something changes they will be about 700k short.
They are fucked.
Session will keep going, see update video and donation page.
A 1-3 person team lead by Jason Rhinelander[1] is to carry on the development of Session, with even some STF members stepping back or only serving in a volunteer capacity. They say that most major features, including Session Pro, are very close to being released. While being regretful about the layoffs, Jason implies that the small team will use the limited funds to continue with LLM-assisted development to ensure that Session is worth the Session Pro subscription. He goes on to say that support is still appreciated to achieve these goals faster. Jason also expects that development will open up more to the community in some respects, especially once the multi-platform codebase restructuring is fully complete.
My opinion:
Given the limited funds, I think this is about as well as it could have gone (for Session).
I respect Jason’s mellowness, honesty and realism and I trust him to do what’s good for Session. He’s been active and approachable on Session, and the STF pretty much agrees he’s the right guy for the job. I’m willing to look past the STF’s history of leadership and throw my weight behind Jason on his work.
I know I’ve been quite critical of project’s direction. I think now’s the time for me to put that aside and focus on the survival of this privacy project. It undoubtedly has received much of the STF’s time and effort.
There have been some strong reactions to Jason’s statement that PFS and post-quantum crypto was added using AI. My interpretation is that no actual cryptography code was written using AI, just that AI was used to integrate a crypto library into Session. I think the statement deserves some clarification by Jason; see also my thread on the (very intentional) deterministic encryption used for Session profile pictures.
Chief Software Architect, Session Technology Foundation ↩︎
I know the remaining Session team may be confident in this approach but from an outsiders perspective it still doesn’t look great.
Similarly while they may not have written any cryptography code using “AI”, it is my opnion that proper integration of cryptography code is just as important as ensuring it is " written" properly, so handing that off to an AI/LLM still concerns me.
I would also be interested just what this statement entails. How much development will move over to the community and will there be any special/restrictive rules community members would need to follow in order to contribute development effort to Session?
Not a good look. But the alternative is untenable. I won’t lose sleep over AI being used for a popular GPL’d privacy project in the hands of someone who knows what they’re doing.
I’ve spoken to Jason about this. Since he is a senior developer who’s been involved in the project’s core since the start, I don’t expect him to lower his standards on cryptography. That’s not to say that Session’s cryptography isn’t hiding some strange decisions. But that’s the more reason for a rewrite to achieve PFS and PQ.
Development “opening up” entails a lowering of the barrier for outside contributions. The requirement for cross-platform feature parity has been a blocker for anyone trying to submit MRs, and there’s certainly been an aspect of pride & tunnel vision stopping outside contributions up till now (see my previous posts in this thread).
With how the code will be structured going forward, it should allow others to build bots, custom clients, change the UI across platforms more easily than ever before… That’s not a guarantee features will be accepted, but I trust Jason will want to put some time aside to review new contributors’ work within in a year or so. I’m hoping that this change in Session’s stewardship will be supported by the rest of the STF. (No clue if I’ll still be around by then!)
I know I’m praising Jason a lot here. But even the author of this scathing 2024 blog post (who happens to be the ex-contributor mentioned in the update video) agreed that Jason was THE person who could merge outside code. My experience is similar, even if a bit less fortunate; I just maintain some forks and auxiliary projects.