Mirroring broken?

[maelle]

👋 @gmbecker! It seems this repo isn't at the same state as the one over at Bioconductor. I noticed that because the build of the docs website fails at https://dev.ropensci.org/job/genbankr/ because of a missing dependency (BiocStyle in Suggests).

Besides, there are two open issues with bug reports/feature requests, is this repo the place where you expect issues to be opened or is there a Bioconductor place for that?

Thank you!

[maelle]

Ping @gmbecker 😸

[gmbecker]

Sorry. I saw this but it sunk back under other things. I will look at this and hopefully get it fixed soon. Thanks for staying on me, and cool that you'er usign the package, I wasn't sure how much appeal/uptake it had.

…

On Wed, Nov 13, 2019 at 2:58 AM Maëlle Salmon ***@***.***> wrote: Ping @gmbecker <https://github.com/gmbecker> 😸 — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#5?email_source=notifications&email_token=AAG53MKBCXQ3DPXNF2L6E6DQTPMUZA5CNFSM4JGYMVIKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOED5XVXQ#issuecomment-553351902>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAG53MPEUDWEIRJ3Z2HGDJTQTPMUZANCNFSM4JGYMVIA> .

[maelle]

Hi! No problem, thanks for the answer!

Sorry for the lack of context, we saw a broken build on https://dev.ropensci.org/job/genbankr/ so I came here to see what was wrong. Once it's fixed there'll be a pretty pkgdown website for the package see https://ropensci.org/technotes/2019/06/07/ropensci-docs/ 🎇

[gmbecker]

So I've manually shoved the correct state into the github repo, but its worth noting that the github is not the repo of record for this package,w hich is at https://git.bioconductor.org/packages/genbankr I don't know that the mirroring is going to be working again (I'm not sure why it stopped), so this may come up again if the ropensci build system continues to go off the gh repo :( ~G

…

On Thu, Nov 14, 2019 at 12:09 AM Maëlle Salmon ***@***.***> wrote: Hi! No problem, thanks for the answer! Sorry for the lack of context, we saw a broken build on https://dev.ropensci.org/job/genbankr/ so I came here to see what was wrong. Once it's fixed there'll be a pretty pkgdown website for the package see https://ropensci.org/technotes/2019/06/07/ropensci-docs/ 🎇 — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#5?email_source=notifications&email_token=AAG53MNZXOQPUANX6NHWI5DQTUBTPA5CNFSM4JGYMVIKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEEA6XYI#issuecomment-553774049>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAG53MOG2IKMVVP36QY4K3TQTUBTPANCNFSM4JGYMVIA> .

[maelle]

Thanks, but that link doesn't work. Cc @jeroen

[FelixErnst]

@gmbecker

In my opinion, the best way to setup GitHub and Bioconductor is described in https://bioconductor.org/developers/how-to/git/

This assumes that GitHub is the main devel repo and is always ahead or level with the bioconductor one.

[gmbecker]

With respect, I do not agree with that assumption. Github is not the repository of record for the sources of biooconductor packages, and I don't think it should be treated as such, if anyone is going to install a bioc package from git it should be from the bioconductor git repository. Github is a convenience when it works, but should not be considered more than that by users. Certainly its not where you should look to get bleeding edge versions of things, cause its not what the Bioc build system pulls from, and I think those should be the same, always. That said I did sync github (I think) with these fixes so in praactice I suppose its largely moot

…

On Thu, Mar 26, 2020 at 11:49 PM Felix Ernst ***@***.***> wrote: @gmbecker <https://github.com/gmbecker> In my opinion, the best way to setup GitHub and Bioconductor is described in https://bioconductor.org/developers/how-to/git/ This assumes that GitHub is the main devel repo and is always ahead or level with the bioconductor one. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#5 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAG53MIAUSR5UXKQTALQ6DTRJRD6ZANCNFSM4JGYMVIA> .

[FelixErnst]

A totally valid point. However the convenience can be quite nice.

Thanks for the push to GitHub. Made testing the fix quite easy