Follow

📢 📢 📢 Today we publish an Open Letter from developers to the regarding shipping unfinished patches to users.

Shipping unfinished patches harms the user experience and slows down the review process in projects to improve the patches.

You can read our Open Letter at: do-not-ship.it

If you agree with this letter, you can support it by signing it, see: do-not-ship.it/sign

@dylanvanassche I wonder if someone from Manjaro is going to sign this ;-)

@linmob I can't comment on this, but developers who signed this, signed this from their personal position and not from the project they may be affiliated with. People who signed this Open Letter really care about users and other developers, they communicate with them on a regular basis to make sure that the experience improves and everyone gets the credit they deserve :)

@dylanvanassche I'll hire you as a diplomat if I ever need one ;-) I understood that, and I've seen the hard to solve issues on upstream bug trackers resulting from unnamed distributions shipping WIP software.

(I agree on the matter, but as long as I don't develop anything that's shippable in distributions (I have ideas, but lack expertise and more importantly time), I don't think it's helpful if I sign this letter.)

@linmob Hahahaha, I experienced this a few times myself that my patches were shipped way too early. I even had to write a blog post to correct some misconceptions because of it. I could have spent the writing time of the post of doing something else.

Regarding the signing, we have people on it that have the same feeling of 'not doing development' but agree with it. It is up to you :)

We don't want to point fingers or blame anyone, we just want to say that such practices are not appreciated.

@dylanvanassche That's a sane stance, finger pointing surely won't help.

@dylanvanassche I have signed this letter too. Shipping unfinished work gives upstream a bad reputation, makes support a big hassle and disincentivizes developers to work in public.
#dontshipwip

do-not-ship.it

@spaetz Exactly! It is discouraging that your patches are shipped before you have the chance to polish them. It looks bad for you as a developer and discourages yourself to work in the open. All these things are just the opposite of what stands for me IMHO.

@dylanvanassche Out of curiosity, what do you think about the Gentoo approach, where some packages have versions marked "99999" and these versions are sometimes compiled e.g. from the tip of the development/main branch?

@timorl Well, Gentoo users have deep knowledge about their system. They exactly know what may have broke their setup and are able to fix it if needed. I have not so much experience with Gentoo, but I'm sure they reach out to developers to help out and fix it. It is a whole different thing when users did not opt in for it, then users get disappointed while it looks bad for the developer, only because a distro decided to ship their patches without consulting the developer first.

@ariadne Alpine Linux for sure not ;) We have a lot of good distros out there who work with developers and even contribute . That's wonderful! However, I experienced it a few times myself, where this wasn't the case. I wrote a giant patch for for example, which was shipped way before it was properly reviewed. I mostly experienced it myself in the Linux Mobile community since it is growing so fast in terms of features. I don't want to do public finger pointing.

As a firm believer in the LTS model, I agree.

@dylanvanassche While working on @tenacity, we got burned out as people wanted it so badly we decided to package everything ASAP and then kept getting complaints as a result of lousy package maintenance work by many parties.

Which brings me to the question... who is this for?

@dylanvanassche @tenacity In retrospect, this makes me fully understand why people like the maintainer of MultiMC absolutely refuse to deal with any of that.

@n0toose @tenacity I'm trying to grasp what you mean, but I think I miss some context, could you ellaborate plz? Is this from an upstream or distro point of view?

@dylanvanassche @tenacity Upstream, having worked on packaging for the Haiku operating system before. Who is this letter aimed towards, is what I am trying to ask.

@n0toose @tenacity For me, the letter is mostly aimed at packagers picking recklessly patches and shipping them to end users which did not opt-in for unstable repositories. I wasn't aware of it which causes irrelevant issues to be created, introduces misconceptions, and miscreditting the author of the patches. Shipping fast and early is fine as long as the original author is informed and does not object in any way. This is especially true for large changes or features.

@dylanvanassche Why not just explicitly stop supporting distros that do this? Developers shouldn't have to worry about how their software is deployed in every single distro. And their bugs are not your responsibility.

To me, the site just sounds like "it's free software except when i don't like how you use it".

There's a license for that: all rights reserved.

@mavis Well if you don't provide support to users of those distros, it looks bad for you. You may explain to them because why, but in the end, you end up in a distro flame wars.

"it's free software except when i don't like how you use it".

Such a statement was never made, and will never be made. The whole reason why we have this letter in the first place is to keep working in the open without being afraid that your WIP patches are shipped without credit, unless there's a need for blame

@powersource It is an issue when you make an awesome thing, submit it upstream for feedback, and then notice your patches are shipped before you got feedback to end users, without their and your knowledge. The first time I noticed it, I thought, this is a one time thing, but no. I have seen it several times already so I had to put huge disclaimers in merge requests. Testing stuff is fine, but breaking stable is not fun for users and looks bad for everyone involved in the patches.

@dylanvanassche yeah i just had higher expectations on distro maintainers

@powersource To be clear, most distros and maintainers are super awesome and don't do this kind of things :)

Sign in to participate in the conversation
Fosstodon

Fosstodon is an English speaking Mastodon instance that is open to anyone who is interested in technology; particularly free & open source software.