If you're a team developing full-time on a project, you should be releasing a stable increment every couple of weeks, at the latest. Gigantic merges are impossible for humans to review, and it means we have to beta test the entire thing, at once.
The users just mass-downloaded this, even though it's been off in a dev branch for months and is a major release (even if it's missing some of the changes I expected, like improved login).
I mean, everyone just trusted them and downloaded this monster. The whole point of open-source is that you don't need trust because the development is conducted in a maximally-transparent manner, with frequent releases facing scrutiny by the end users.
sometimes i wake up and decide to run some opensource. i would have run a primal, but i didnt even get to the clone.. because i realized, the code is gone.
agree with you and there is absolutely great value on releasing all backends (and shame for who keep it close). Still the drama about "users trust" is nonsense.
this is false, they are modern and auditable codebases, and the commits well organized and self-explicable.
All is open, accessible and understandable for everyone.
The point is that they follow a not so common dev practicing that could let new people on codebase a bit disoriented, but it doesnt impact the possibility of read the code and audit it by any means.
Every codebase is different as it could fulfill different porpouses, so this drama is totally nonsense.
i am interested in their backend, when i asked, they didnt have an answer. will they only open the client code and keep server private? imquiring minds.. want to know..
frankly dont know about backends codebases. Its super important to open source all and would be great to have responses about that.
Still, backend involves trust in who runs it and offer the services, opening the source doesnt make much difference in reducing the trust that users need to put as theres no way to audit the computation happening in the backed.
opening the backend source, means we can keep primal honest by running our own.. if we cant, thats not very nostr like. if clients all use a primal cache cause its so cool and fast and has all these features, but its closed source, then we didnt accomplish the mission to decentralize.
maybe the 3 months of no commits just means there arent any new commits because backend is the same as 1.0. i would highly doubt it. which is why i asked.
Oddbean new post about login | logout