Move to JuliaStats

[juliastats]

Move to JuliaStats

[dmbates]

"let someone else do it" sounds good to me.

You know that the #1 item on my personal bucket list is to, once in my life, have the task of tagging and publishing a release of a package go off without a hitch.

I hesitate to ask but how do I undo the change in my metadata. My reflog begins

6f9028a HEAD@{0}: commit: Tag Rmath v0.0.1
ed87094 HEAD@{1}: rebase finished: returning to refs/heads/metadata-v2
ed87094 HEAD@{2}: pull --rebase -q: checkout ed87094911ab43d2d78b417b666a955c97f39596
433e9a4 HEAD@{3}: rebase finished: returning to refs/heads/metadata-v2

I imagine I am supposed to type something starting with git rebase --force and if I get the syntax wrong I will be forced to become a theoretician and comfort myself with the fact that the biggest technological problem I will face will be resharpening my pencil.

[nalimilan]

Could be practical, but I worry about making it sound like a recommendation to use this package, which is just a stopgap AFAIK.

[tkelman]

We can move it elsewhere or just keep it here, as long as we give enough people commit access.

[simonbyrne]

I don't think having it under JuliaStats implies any recommendation. Even once it's replaced, we will probably want to keep it around for testing purposes.

[tkelman]

I think both of you guys are on vacation, but bump. We'll need this at least temporarily, and while Doug was kind enough to grant a few people commit access it may still help to have it in the organization.

[simonbyrne]

I think we should move it.

@dmbates I believe you are an owner on JuliaStats, so should be able to do it yourself: would you mind doing this?

[dmbates]

Will do.

[dmbates]

@simonbyrne Repository now moved. @tkelman Enabling Travis is one of those things that is sufficiently obscure and which I do sufficiently infrequently that I forget how to do it. Now that the repository is in the JuliaStats organization can other members of the organization do it?

[tkelman]

Yes, I think it'll need someone with admin access. Do you know the list of who that is? It's under https://travis-ci.org/profile/JuliaStats (you may need to click Sync Account at the top)

[dmbates]

@tkelman I believe I have Travis enabled now. Do you want to try a commit to check?

[tkelman]

Thanks! https://travis-ci.org/JuliaStats/Rmath.jl/builds/144524163 (can delete the testtravis branch when that finishes). Now could a JuliaStats member checkout master and do Pkg.tag("Rmath", :minor) ?

[dmbates]

I swear,, someday I am going to try to do a Pkg.publish() and have it succeed.

julia> Pkg.tag("Rmath")
INFO: Tagging Rmath v0.0.1
INFO: Committing METADATA for Rmath

julia> Pkg.publish()
INFO: Validating METADATA
INFO: Pushing Rmath permanent tags: v0.0.1
ERROR: Permission to tkelman/Rmath.jl.git denied to dmbates.
fatal: Could not read from remote repository.

Please make sure you have the correct access rights
and the repository exists.
ERROR: failed process: Process(`git --work-tree=/home/bates/.julia/v0.4/Rmath --git-dir=/home/bates/.julia/v0.4/Rmath/.git push -q origin refs/tags/v0.0.1:refs/tags/v0.0.1`, ProcessExited(128)) [128]
 in pipeline_error at process.jl:555
 in run at process.jl:531
 [inlined code] from pkg/entry.jl:389
 in anonymous at task.jl:447
 in sync_end at ./task.jl:413
 [inlined code] from task.jl:422
 in publish at pkg/entry.jl:374
 in anonymous at pkg/dir.jl:31
 in cd at file.jl:22
 in cd at pkg/dir.jl:31
 in publish at pkg.jl:61

[tkelman]

oh you may have run the code from my juliacon workshop that cloned my fork

[tkelman]

maybe undo that change in your metadata, delete Rmath from your .julia and be sure you clone it from JuliaStats (or let someone else do the deed)

[tkelman]

Wait, I think I still have commit access here so I can do it later. Sorry for the forgetfulness.

[tkelman]

I like doing git reset HEAD~ then checking git status and deleting the non tracked local Rmath/versions/0.0.1 files

[dmbates]

Thanks, Tony. It seems like it worked.