[Bioc-devel] svn repository for data experiment packages now having branches too

[Hervé Pagès]

Hi developers,

Taking advantage of this particularly calm week on Bioc-devel,
I'd like to formally announce that we're now using branches
in the svn repo for data experiment packages.

The trunk/branches layout is similar to what we already do in
the software repo. That is:

  - URL for packages in trunk:

      https://hedgehog.fhcrc.org/bioc-data/trunk/experiment/pkgs/

  - URLs for packages in the branches:

 
https://hedgehog.fhcrc.org/bioc-data/branches/RELEASE_2_13/experiment/pkgs

 
https://hedgehog.fhcrc.org/bioc-data/branches/RELEASE_2_14/experiment/pkgs

 
https://hedgehog.fhcrc.org/bioc-data/branches/RELEASE_2_15/experiment/pkgs

Only the RELEASE_2_13 branch exists right now. This is where the
packages that are in the current release version of BioC (2.13)
are living. Like for software packages, a new branch will be created
for each new BioC release. For example the RELEASE_2_14 branch will
be created in April one day or so before the BioC 2.14 release,
the RELEASE_2_15 branch in October, etc...

More precisely, at any given time, you only have access to 2
versions of your data experiment package: the "release" and
"devel" versions.

The "release" version of your package is always in the latest
branch (RELEASE_2_13 at the moment) and accessible at:

 
https://hedgehog.fhcrc.org/bioc-data/branches/RELEASE_2_13/experiment/pkgs/<PKGNAME>

Only bug fixes and documentation improvements should go here.

The "devel" version of your package is always available at:

   https://hedgehog.fhcrc.org/bioc-data/trunk/experiment/pkgs/<PKGNAME>

One important difference with the svn software repository still
remains though. For data experiment, the recommended way to checkout
a package is:

   1. Checkout a "shallow" working copy of the pkgs/ folder:

        svn co 
https://hedgehog.fhcrc.org/bioc-data/trunk/experiment/pkgs --depth=files

      Use the URL to the current branch if you need to checkout a
      "release" package.

   2. cd to the pkg/ folder

   3. Fetch the working copy of a particular package:

        svn up <PKGNAME>

      (You can list all the folders in pkgs/ with 'svn ls')

      IMPORTANT: This doesn't fetch the data. But most of the
      time you don't need them e.g. if you only need to make a
      change to a man page, the vignette, the DESCRIPTION file,
      the NAMESPACE etc... you can do so and commit.
      But of course, if you didn't fetch the data, it's very
      likely that you won't be able to test your changes with
      'R CMD check'.

   4. If you need to fetch the data:

        (from the pkgs/ folder)

        ./add_data.py <PKGNAME>

      This will populate the <PKGNAME> folder with the data.
      Note that you need to be in a data folder in order to commit
      any change to the data. Any commit you do from the top level
      folder of the package won't include any change you could have
      made to the data.

We're also wondering whether we should start to use the same x.y.z
versioning scheme as for software packages, that is, an even y in
release, an odd y in devel, with y bump (and z reset to 0) at each
new release. Feedback welcome.

Please let me know if you have questions or concerns about this.

Cheers,
H.

-- 
Hervé Pagès

Program in Computational Biology
Division of Public Health Sciences
Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N, M1-B514
P.O. Box 19024
Seattle, WA 98109-1024

E-mail: hpages at fhcrc.org
Phone:  (206) 667-5791
Fax:    (206) 667-1319