[Bioc-devel] Solved Bug commiting problem

Martin Morgan mtmorg@n@b|oc @end|ng |rom gm@||@com
Thu Nov 26 21:39:20 CET 2020

Yes, the 'master' branch in git refers to the 'devel' branch of Bioconductor.

You should not aim to keep the RELEASE_3_12 branch the same as 'devel'.

The RELEASE_3_12 branch is meant to be a snapshot of the 'devel' branch at a particular point in time, so that users of the RELEASE_3_12 branch get a stable package for doing reproducible research. The one time when a change is introduced into the release branch is when it is a bug fix, and then the bug fix is usually also added to the devel branch.

The 'devel' branch is meant to allow you to introduce new features, and deprecate and remove old features, without disrupting the productive work being done by users in the release branch.

So generally there is much more activity in the devel branch, and almost all commits in the release branch are also present in the devel branch. Version numbers indicate the separate trajectories of the two branches. For version x.y.z, 'y' is even in the release branch, and odd (and equal to y + 1 of the current release branch) in the devel branch. 'z' is incremented separately in the release branch (rarely, bug fixes only) and the devel branch. As an example from one of my own packages, AnVIL is 1.2.0 (no bug fixes) in release, and 1.3.5 (five commits representing new features or non-portable bug fixes) in devel.  

The version numbering scheme is discussed at



On 11/26/20, 5:49 AM, "Bioc-devel on behalf of Selcen Arı" <bioc-devel-bounces using r-project.org on behalf of selcenarii using gmail.com> wrote:

      Hi everyone,

    I am developer of ceRNAnetsim package (
    https://git.bioconductor.org/packages/ceRNAnetsim). We recently saw that it
    gave errors in build report (
    http://bioconductor.org/checkResults/release/bioc-LATEST/ceRNAnetsim/ ),
    and we fixed the errors in the package, in this direction. After making the
    local changes, we committed to the master and RELEASE_3_12 branches on the
    bioconductor and github repos, following the instructions you can find in
    the links (1
    and 2
    The package version in the master branch was higher than RELEASE_3_12.

    Is it because the master branch refers to devel?Therefore, I wanted to sync
    all versions in these branches to the version in the master. However, in
    this case, I encountered the problem like "illegal version bump from 1.2.0
    to 1.3.2".

    As a result, while the package version is 1.2.1 in RELEASE_3_12 branch, it
    is 1.3.2 in the master branch. Will this cause any problems afterwards?


    Selcen ARI YUKA
    Yildiz Technical University
    Faculty of Chemistry and Metallurgical Engineering
    Department of Bioengineering
    Esenler, Istanbul, Turkey

    	[[alternative HTML version deleted]]

    Bioc-devel using r-project.org mailing list

More information about the Bioc-devel mailing list