Get publishing permissions on rpm-rs, or rename this crate

[cmeister2]

We need to get publishing permissions on the rpm-rs crate, or rename this to something else, otherwise further changes are moot.

[olivierlemasle]

Do you have any news on getting publishing permissions for rpm-rs?
If not, when do you plan to rename and publish the forked crate?

[dralley]

No news. The original contributor had some unrelated github activity 2 weeks ago but hasn't replied to emails or messages.

@cmeister2 I think, perhaps we should just continue publishing under the fez name for now.

[cmeister2]

Couldn't we just rename to something like rpm2 in the meantime, so we can actually get on with adding features? I'd be happy to use fez too, but there's a bit of hassle there with the feature I added.

[dralley]

We could, I just figured if we already have a crate name, there's no need to create yet another one, that may eventually be abandoned anyway.

Is the feature not one that makes sense to upstream?

[dralley]

@Richterrettich Trying again to reach you : /

@cmeister2 Do you feel it's a hard blocker?

[cmeister2]

@cmeister2 Do you feel it's a hard blocker?

I don't know if it's a hard blocker; the main changes in fez from rpm-rs:

• The overall LICENSE is MIT with a NOTICE of the portions of code which are Apache-2.0; AIUI the two licenses are compatible with doing this (https://opensource.stackexchange.com/questions/11640/can-mit-and-apache-licenses-be-used-together). There's also code adapted from debpkg which is MIT-licensed.
• I added a new struct RpmPkgReader which allowed for sequential parsing; I don't know whether it's the most architectural way of achieving the goal of being able to parse just the metadata from a package instead of having to load the whole RPM.

Generally I'd like to have this feature - it prevents having to download megabytes of RPM package data to just parse the metadata at the start. However my immediate use case has tailed off in urgency so I could live without it for a while.

[olivierlemasle]

Hello @cmeister2 @dralley @Shinyzenith,

Did you reach a decision regarding this crate? I'd be happy to contribute to this Richterrettich/rpm-rs fork, but I'd like to know if my pull requests are going to be merged some day, and if it a release of the forked crate will be made. If this fork is not more "maintained" that the original, that's a little bit sad...
Please understand that I feel entitled to nothing, that I do not blame you of anything and that I really want to help :-)

@cmeister2 Regarding fez, I'm not a lawyer, but I'm not sure you had the right to change the license as you did. The StackExchange post you shared tells that you can use dependencies under one of them in a project under the other. It does not say you can re-license the work made by other authors. I do not think the part of 4 of the Apache 2 license enables you to change the license, even when you add a NOTICE file with the old license. This other post from StackExchange tells that you cannot re-license a project from Apache-2.0 to Apache-or-MIT without asking all the authors, so I guess it's the same thing (or worse!) to re-license from Apache-2.0 to MIT.

[cmeister2]

@olivierlemasle looks like you're right; I'll have to fix that. Might be easiest to just blat the code here into that repo and forgo the change I made for now, or archive it.

Regarding this fork; I'm not an admin here. It does seem like we're not getting anywhere with the original rpm-rs package though. If I had admin permissions here I'd probably rename this repo to fez and update everything to reflect that name, and then publish it to the fez crate.

[dralley]

Yeah, I emailed him again as well as the owner of the https://crates.io/crates/rpm crate which has gone unused, but not really expecting anything.

@drahnr Since you created the org I assume you have admin rights? What are your thoughts on this.

IMO it would be OK to keep the repo named the way it is for now so long as we have it linked properly in Cargo.toml (which also shows up on crates.io)

[Shinyzenith]

Hello, I have been pretty invested with university lately so I apologize if I haven't carried out my duties properly. I'm still going through the thread and I'm slightly confused as to what we are trying to achieve, if anyone can summarize the issue that would be extremely helpful.

[drahnr]

@dralley I am happy to give you admin privileges.

I don't have a strong opinion, on naming, it'd be nice to continue with rpm-rs if possible, but my hopes are slim/low.

[dralley]

@drahnr I would appreciate that, thanks.

[dralley]

I have one more idea

[dralley]

@TommyLike Regarding the other thread, did you manage to reach out on Slack?

[TommyLike]

@TommyLike Regarding the other thread, did you manage to reach out on Slack?

No message replied on the slack till now, I am trying to reach him within other ways.

[drahnr]

@drahnr I would appreciate that, thanks.

Done

[cmeister2]

In lieu of a solution I've published current master under fez:0.8.1 (there were a couple of deprecations on the latest compiler that need looking into so I published manually).

[dralley]

@TommyLike I presume still no response from him?

That is disappointing but, so it goes. The author of the qubes-rpm-oxide (not published on crates.io) said they would be be open to collaborating on one library under the name rpm-oxide, so, if we hear nothing back, that is what we should probably plan to do.

Or at least start that discussion. QubesOS/qubes-rpm-oxide#12 (comment)

[TommyLike]

@dralley still need some time. hope I can contact him by this weekend.

[TommyLike]

@dralley hope it work now.

[dralley]

Indeed, we now have the rpm crate.

[drahnr]

I received the invite link, ty very much. I'll do an initial release, to be coordinated in another issue. Thank you!

[drahnr]

I am tempted to close this, since we have a canonical crate name under which we can publish. Collaboration imho is another concern that could be discussed separately.