Tidy up some cf-convention repositories

[davidhassell]

Hello,

I would like to explore the possibility of tidying up the following repositories:

• https://github.com/cf-convention/repository-cf

  As far as I can tell, this contains a 9-year old XML version of the source for CF-1.6.

  Do we still need this? Can we delete it?

  Surely someone somewhere has the original source text for versions 1.0 to 1.6 (1.6 onwards are in asciidoc format in https://github.com/cf-convention/cf-conventions). Perhaps a new repository called archive could be made to store the source text for CF-1.0 to 1.5, and any other archivable artefacts

• https://github.com/cf-convention/Conformance

  This repo, with all of its commit history was transferred to https://github.com/cf-convention/cf-conventions. There are no issues (open nor closed). There are two open PRs, but these were replicated and implemented over at the new home.

  I suggest we can delete this repo.

• https://github.com/cf-convention/CF-2

  There is some really interesting discussion here, but I think that the repo name is perhaps not quite right anymore. Since these discussions took place (nearly entirely 7-8 years ago), some netCDF extended data model features mentioned therein have indeed found their way into CF-1.x (groups; unsigned bytes and integers; strings ...) and I get the impression that there is no longer such an impetus behind a CF-2.0 that is incompatible with CF-1.x. Not that it shouldn't be considered.

  I would like to propose renaming the repo to something like CF-futures-discussion.

Thanks,
David

[sadielbartholomew]

Hi @davidhassell, though I have only skim read your proposal it sounds good to me, the general idea certainly. So I agree with this, for one.

Just wondering if you are aware that repos can be put into a marked 'archived' read-only state:

You can archive a repository to make it read-only for all users and indicate that it's no longer actively maintained.

(one example is here) which might be an option we want to go for, perhaps for the first case you list?

[davidhassell]

Hi Sadie, thanks. These three repos already have "Public archive" GitHub status, and are so are indeed read-only. I seem to recall that this was considered a safe option at the time, that removed the need to think about whether or not we needed or wanted to keep the content. It just came up in a discussion with Jonathan that maybe we could have those thoughts now ...

[JonathanGregory]

Dear @davidhassell

Thanks for your sensible proposals.

https://github.com/cf-convention/repository-cf. As far as I can tell, this contains a 9-year old XML version of the source for CF-1.6. Do we still need this? Can we delete it? Surely someone somewhere has the original source text for versions 1.0 to 1.6.

I find that there are directories such as https://github.com/cf-convention/cf-convention.github.io/tree/main/Data/cf-conventions/cf-conventions-1.0/docbooksrc for each of the versions up to 1.6. The XML files are the DocBook source you mention, it looks like. If the files in repository-cf are different from those in https://github.com/cf-convention/cf-convention.github.io/tree/main/Data/cf-conventions/cf-conventions-1.6/docbooksrc, maybe they could have a permanent home as Data/cf-conventions/cf-conventions-1.6.subversion/docbooksrc. Then we could delete repository-cf as you say.

I agree we can delete Conformance. I think we kept it just for safety after having transferred it to cf-conventions. That was quite a while ago and no problems have emerged.

CF-2. Yes, it's a good idea to rename it. Could we get something like archive into the name? I think we should make clear, perhaps in its README as well as the name, that it's obsolete. Future discussions about the future would be fine and welcome, but they can take place in the discuss repo.

Best wishes

Jonathan

[JonathanGregory]

Can we do this now? The proposal is to delete two redundant repos, and rename another inactive one to be kept for future reference. No-one has objected since it was proposed in August, and enough support has been expressed.

[erget]

I think this is fine to complete as proposed - @davidhassell 's reasoning on deleting is sound, as they are redundant, and renaming CF-2 is probably also a good idea. I wouldn't put "archive" in the name, but simply mark it as archived as Sadie described.