zhaofengli
Project status?
First of all, I really like this project, I was looking for a self-hostable binary cache for quite a while now. According to my limited experience, attic seems to be a stable and very efficient tool. Although the README clearly states that it is a prototype, it feels like it is already more than that. To be honest, the number of unanswered issues and PRs a bit worries me as it seems like the project is unmaintained.
So my question is this: is there any plan to move this project forward and out of its prototype status?
Hi there, and thanks for the questions! I've been busy for the past months but I'm trying to find time to push my projects forward. Before moving onto Attic, my current focus is getting issues and PRs in Colmena resolved.
Regarding the prototype status, it's indeed the current characterization but I rely on Attic myself and it's meeting my own needs.
Would it be possible to move to a multi-maintainer setup? Right now people seem to want to contribute but prs remain open without any clear way forward.
Just an update: Still swamped with stuff but hopefully able to push things through soon. See also https://github.com/NixOS/nixpkgs/pull/274481
Would it be possible to move to a multi-maintainer setup?
Good point, and I'm considering adding others to help with maintenance.
Maybe this could be transferred to the nix-community org, while keeping you as the primary maintainer @zhaofengli ? Along with a couple more maintainers this could go a long way to helping secure longer term maintenance.
Still swamped with stuff
@zhaofengli Totally get that. I also don't want to add to your workload. Maybe a discord (or similar) could help? Like this there can be some level of self-organisation.
Maybe this could be transferred to the nix-community org, while keeping you as the primary maintainer @zhaofengli ? Along with a couple more maintainers this could go a long way to helping secure longer term maintenance.
I really like this idea. The project is really good and it would be fantastic if it would be an official part of the community. The official ways to setup a cache aren't that good and require a lot of disk space.
~Since @zhaofengli seems to be working for DeterminateSystems now.~ And DeterminateSystems seems to want to capitalize on binary caches. It feels likes it's time for a free fork. At least then we can do necessary maintenance and add community requested features.
Edit @zhaofengli: is a full-time student and is not working for DetSys. A closer read of this post makes that clear.
Well I’ve just thrown out my attic setup but I would be open to help with testing.
Since @zhaofengli seems to be working for DeterminateSystems now. And DeterminateSystems seems to want to capitalize on binary caches. It feels likes it's time for a free fork.
For the avoidance of doubt: I'm a full-time student and I'm not working for DetSys in any capacity. Over the past summer, there have been changes which are not ready to be merged.
@zhaofengli sorry for the misinterpretation of your job status.
The broader point still stands. This probably should be a community project. Like this, more people than just you, can do the work of maintaining and improving the project.
We are one year down the road now (since this issue was opened), It's time. If you don't have the mental-space or the interest to work on this, that's fine. At least open up the project to other people who are willing to do some of that work.
I know I'm being very harsh/direct here. I don't want to make you feel bad. This is a great project and it would be a shame to see it go into obscurity...
This is Zhaofeng's project. Zhaofeng has made amazing contributions to the community and ecosystem. If you think something should be different, feel free to exercise the freedoms of open source! It is fine!
This is Zhaofeng's project. Zhaofeng has made amazing contributions to the community and ecosystem.
I don't think anybody is debating that, do you?
If you think something should be different, feel free to exercise the freedoms of open source! It is fine!
I'd much rather like to work with @zhaofengli on this project than to have to make a fork. That's why this conversation exists.
The broader point still stands. This probably should be a community project. Like this, more people than just you, can do the work of maintaining and improving the project.
We are one year down the road now (since this issue was opened), It's time. If you don't have the mental-space or the interest to work on this, that's fine. At least open up the project to other people who are willing to do some of that work.
To start, I'm enlisting the help of @cole-h to help review and merge the PRs. He has contributed to the project and provided a lot of reviews.
I'd much rather like to work with @zhaofengli on this project than to have to make a fork. That's why this conversation exists.
Please do! It would really help to clarify what you want to see.
Over the past summer, there have been changes which are not ready to be merged.
For more clarity, I have private branches where I experiment and break things before cleaning them up. If you look at the recent PRs, I've been working to add in the integration bits little by little.
Edit @zhaofengli: is a full-time student and is not working for DetSys. A closer read of this post makes that clear.
There is no need to dig or read between the lines. I shouldn't have to clarify this, but I have not collaborated with DetSys since magic-nix-cache in 2023 (otherwise, I would have merged #99 ages ago). I'm acutely aware that there is a ton of drama and I don't have the spoons for it.