Version numbering in devel and release will ALWAYS be different as described


http://bioconductor.org/developers/how-to/version-numbering/


In a package version   package_x.y.z  -

Y in devel will always be odd and 1 greater than that in release

Y in release will always be even and 1 less than that in devel


When making bug corrections only bump the z -


The process you described is correct

your release version number should be 1.0.z

and the devel should be 1.1.z


The process you described, making changes, and then choosing the separate 
version number is the correct process, however, keep in mind the release branch 
should really only be updated when making larger bug corrections and not with 
development changes/enhancements that should only be in devel.



Lori Shepherd

Bioconductor Core Team

Roswell Park Cancer Institute

Department of Biostatistics & Bioinformatics

Elm & Carlton Streets

Buffalo, New York 14263

________________________________
From: Bioc-devel <bioc-devel-boun...@r-project.org> on behalf of Tang, Wenhao 
<wenhao.tan...@imperial.ac.uk>
Sent: Tuesday, November 13, 2018 6:55:09 AM
To: bioc-devel@r-project.org
Subject: [Bioc-devel] 答复: 答复: A little confuse about how to update accepted 
package

Dear all,


Now I understand how to sync my local master, local RELEASE_3_8, remote master 
and remote RELEASE_3_8.


All of them have version 1.0.5. I also want to update the DEVEL branch, but 
when I did (the step 9 in 
http://bioconductor.org/developers/how-to/git/sync-existing-repositories/)


git push upstream master

I received the following message:

Wenhao@TANG MINGW64 /e/RNAseqProject/bayNorm_dev/bayNorm (master)
$ git push upstream master
Total 0 (delta 0), reused 0 (delta 0)
remote: Error: Illegal version bump from '1.1.2' to '1.0.5'. Check
remote: http://bioconductor.org/developers/how-to/version-numbering/
remote: for details
To g...@git.bioconductor.org:packages/bayNorm.git
 ! [remote rejected] master -> master (pre-receive hook declined)
error: failed to push some refs to 
'g...@git.bioconductor.org:packages/bayNorm.git'


What I have tried is to replace 1.0.5 with 1.1.3 and then push it. After that I 
replaced 1.1.3 with 1.0.5. Obviously this way is not efficient. Since RELEASE 
nad devel branches have different pattern of version numbers, how could I 
update them in a consistent way?

Thank you in advance!

Best wishes,
Wenhao






________________________________
发件人: Martin Morgan <mtmorgan.b...@gmail.com>
发送时间: 2018年11月12日 1:15:42
收件人: Tang, Wenhao; bioc-devel@r-project.org
主题: Re: [Bioc-devel] 答复: A little confuse about how to update accepted package

The release branch should be numbered 1.0.1, 1.0.2, 1.0.3, ...

The devel branch should be numbered 1.1.1, 1.1.2, 1.1.3, ...

It does not matter, from the build and distribution perspective, what your 
github repository contains. What matters is the content of git.bioconductor.org 
repository, the 'devel' / master branchs

Packages build and check nightly, so please look at the nightly build reports at

http://bioconductor.org/checkResults/

The release build report for your package is at

http://bioconductor.org/checkResults/3.8/bioc-LATEST/bayNorm/malbec1-install.html

You can see that the most recent build report (at the time I write this) 'was 
generated on 2018-11-11' and for your package

Snapshot Date: 2018-11-10 16:45:10 -0500 (Sat, 10 Nov 2018)
URL: https://git.bioconductor.org/packages/bayNorm
Branch: RELEASE_3_8
Last Commit: d371781
Last Changed Date: 2018-10-30 11:42:10 -0500 (Tue, 30 Oct 2018)

This says that the most recent builds completed on 2018-11-11, but started on 
the Snapshot Date 2018-11-10 16:45:10 -0500.

Your commit was

commit ead59d29ff65eb9eca7b1aa4dc49c12db1f60190 (HEAD -> RELEASE_3_8, 
origin/RELEASE_3_8)
Author: WENHAO TANG <wt...@ic.ac.uk>
Date:   Sun Nov 11 06:21:06 2018 +0800

    1.0.2

which is AFTER the build started, hence NOT INCLUDED in last night's build and 
NOT YET AVAILABLE to the end user.

The version of your package on git.bioconductor.org has the following 
DESCRIPTION file

git/bayNorm RELEASE_3_8$ head DESCRIPTION
Package: bayNorm
Type: Package
Title: Single-cell RNA sequencing data normalization
<<<<<<< HEAD
Version: 1.0.0
=======
Version: 1.0.2
>>>>>>> 55942b4... 1.0.2
Authors@R: c(person("Wenhao", "Tang", role=c("aut", "cre"),
        email="wt...@ic.ac.uk"),

So even though your commit will be included in the Sunday night build, it will 
not complete successfully. You'll need to update the DESCRIPTION file so that 
it is formatted correctly, and push that to git.bioconductor.org before 
tonight's builds start.

For your specific questions

1. Yes you have to wait for the build system to work on a daily basis. But best 
practice is to build and check your package before committing. Personally I 
would do this without third-party tools (e.g., devtools) since this is how the 
nightly builder works.

So for instance using R and Bioconductor version 3.8 I have

~/b/git/bayNorm RELEASE_3_8$ cd ..
~/b/git$ bioc-3.8 CMD build bayNorm
* checking for file ‘bayNorm/DESCRIPTION’ ...Error : file 
'/Users/ma38727/b/git/bayNorm/DESCRIPTION' is not in valid DCF format
 EXISTS but not correct format

and you therefore know immediately that something needs to be fixed without 
waiting 24 hours.

A good practice might be

  R CMD build bayNorm
  R CMD check bayNorm_1.0.2.tar.gz

2. No, work flows do not require two local repositories. For your scenario 
there is a single repository, with different 'branches' (for devel and 
RELEASE_3_8) and different 'remotes' (for git.bioconductor.org and github.com)

Item 10 of

  http://bioconductor.org/developers/how-to/git/sync-existing-repositories/

tries to sketch the steps required to maintain two remotes and two branches, 
which is what you art trying to do (it seems like you have the various steps in 
place, perhaps not coordinated a cleanly as they could be).

Martin


On 11/11/18, 8:11 AM, "Bioc-devel on behalf of Tang, Wenhao" 
<bioc-devel-boun...@r-project.org on behalf of wenhao.tan...@imperial.ac.uk> 
wrote:

    Dear all,


    Currently, for both branches: master and RELEASE_3_8 in my github 
repository:

    https://github.com/WT215/bayNorm/tree/master


    have been updated with version number 1.0.2.


    However, in the latest build report:

    http://bioconductor.org/checkResults/3.8/bioc-LATEST/bayNorm/


    the version is still 1.0.0.


    I have the following two questions:


      1.  Do I have to wait 24 hours to see whether the update in the release 
version success or not?
      2.  Do I need to have two local repositories for Release and Developed 
versions bayNorm respectively? In the tutorial: 
http://bioconductor.org/developers/how-to/git/sync-existing-repositories/, it 
seems that only local (Release), Bioconductor (Release), and GitHub 
repositories (Release) can be synced.

    Thank you very much!

    Best wishes,
    Wenhao








    Thank you very much!


    Best wishes,

    Wenhao

    ________________________________
    发件人: Martin Morgan <mtmorgan.b...@gmail.com>
    发送时间: 2018年11月10日 19:10:13
    收件人: Tang, Wenhao; bioc-devel@r-project.org
    主题: Re: [Bioc-devel] A little confuse about how to update accepted package

    The steps below are correct, but perhaps not sufficient for what you want 
to do.

    Bioconductor packages consist of release branches, which are meant to be 
stable for the end user, and the 'devel' branch where new features can be 
introduced.

    There is a version numbering scheme, outlined at

      http://bioconductor.org/developers/how-to/version-numbering/

    to ensure that each published version of the package has a different 
version, that versions in release always increase, and that the devel version 
is always ahead of the release version. For your package, the release version 
is 1.0.0, and the devel version is 1.1.0.

    If you intended to make a change to the devel version of your package, 
because you were introducing a new feature or perhaps testing a bug fix, then 
you have been successful; you have pushed your changes to the devel ('master') 
branch, and incremented the version to 1.1.1.

    If you intend to make this change also in the release, then the extra steps 
in

      
http://bioconductor.org/developers/how-to/git/bug-fix-in-release-and-devel/

    are required -- check out the RELEASE_3_8 branch, port your changes from 
master, increment the release version to 1.0.1, and push.

    During the submission process packages are built immediately, but in the 
devel and release branches they are built nightly, approximately (I think that 
last night in particular there were issues with network connectivity that might 
severely limit the success of the build). The build reports are available from

      http://bioconductor.org/checkResults/

    note the 'snapshot' time stamps and commit information for the overall 
build and for your individual package to see whether your commit was included 
in the nightly build.

    Martin

    On 11/10/18, 5:53 AM, "Bioc-devel on behalf of Tang, Wenhao" 
<bioc-devel-boun...@r-project.org on behalf of wenhao.tan...@imperial.ac.uk> 
wrote:

        Dear Bioconductor team,


        My package bayNorm has recently been accepted, it now has version 
1.0.0. Currently I would like to make some changes to it. I look through the 
tutorials online but I am still very confused. Below is what I have done:


        My local repository has version 0.99.19. So I  did the following in my 
local repository:

          1.  git remote add upstream g...@git.bioconductor.org:packages/bayNorm
          2.  git fetch --all
          3.  git checkout master
          4.  git merge origin/master
          5.  git merge upstream/master

        Now the local version is 1.1.0 rather than 0.99.19.


        Then I just did a small change (in addition, I bump the version to 
1.1.1) and then run the following command under the local repository:

          1.  git push upstream master
          2.  git push origin master

        Is the above procedure correct? Will I see the package building report 
like what I have received during the submission process?

        Your tutorials are very clear, and I am sorry if my question is too 
trivial.

        Thank you very much!

        Best wishes,
        Wenhao

             [[alternative HTML version deleted]]

        _______________________________________________
        Bioc-devel@r-project.org mailing list
        https://stat.ethz.ch/mailman/listinfo/bioc-devel


         [[alternative HTML version deleted]]

    _______________________________________________
    Bioc-devel@r-project.org mailing list
    https://stat.ethz.ch/mailman/listinfo/bioc-devel


        [[alternative HTML version deleted]]

_______________________________________________
Bioc-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/bioc-devel


This email message may contain legally privileged and/or confidential 
information.  If you are not the intended recipient(s), or the employee or 
agent responsible for the delivery of this message to the intended 
recipient(s), you are hereby notified that any disclosure, copying, 
distribution, or use of this email message is prohibited.  If you have received 
this message in error, please notify the sender immediately by e-mail and 
delete this email message from your computer. Thank you.
        [[alternative HTML version deleted]]

_______________________________________________
Bioc-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/bioc-devel

Reply via email to