Skip to content

[Bioc-devel] BiocStyle and R markdown

6 messages · Thomas Girke, Kasper Daniel Hansen, Jim Hester +1 more

#
Yes, they show under Bioc-devel, but not under Bioc-release. Building the
exact same *Rmd source from here
https://bioconductor.org/packages/devel/bioc/vignettes/BiocStyle/inst/doc/HtmlStyle.html
under Bioc-release looks quite different:
http://biocluster.ucr.edu/~tgirke/HtmlStyle.html

Thomas

On Wed, Jul 29, 2015 at 7:58 PM Kasper Daniel Hansen <
kasperdanielhansen at gmail.com> wrote:

            

  
  
#
Well, without really knowing details it could just be that more work has
been done on BiocStyle in devel.

One could backport the change to release, but that would mean that all
packages (which use BiocStyle) need to be version bumped and re-build,
purely for cosmetic issues in the vignette.  While I care about aesthetics,
it is not worth doing a massive update to our stable release branch for
this.

So lets try to hammer BiocStyle before the next release so we can iron out
any issues.

Biocore: could we get a Github version of BiocStyle so we could start
issues etc.  Easy way to keep a todo.

Best,
Kasper
On Thu, Jul 30, 2015 at 12:13 AM, Thomas Girke <thomas.girke at ucr.edu> wrote:

            

  
  
#
Kasper,

Easiest way to get a GitHub version of BiocStyle is to just fork
https://github.com/Bioconductor-mirror/BiocStyle/

Jim

On Thu, Jul 30, 2015 at 9:28 AM, Kasper Daniel Hansen <
kasperdanielhansen at gmail.com> wrote:

            

  
  
#
Thanks Jim, this prompted me to read the documentation one more time and I
understand it better now.  Since I have svn write permissions to BiocStyle
I can do what you're suggesting.

However, it makes sense to have one location on Github with which to track
issues and accept incoming pull requests - at least in my thinking.  Of
course, I could have a fork where I have the issues I am interested in and
will accept the pull requests that I am ready to support and someone else
can have a different fork.  But it seems to me that we may be loosing track
of a central list of issues etc. out by doing so.  Or is the thinking that
the read-only mirrors should contain issues and then fixes go through
individual people.  Here I am especially thinking about the packages
maintained by core.  What is your thinking on this situation?  I could see
myself making pull request to SummarizedExperiment in the next 6 months for
example, or documentation clarifications to all packages?

For my own packages, like minfi, it is pretty clear (I hope, from
Description and the fact that I maintain) which Github repos is the one
true hub for minfi development.

Best,
Kasper

On Thu, Jul 30, 2015 at 9:45 AM, Jim Hester <james.f.hester at gmail.com>
wrote:

  
  
#
----- Original Message -----
There is a read-only github mirror for every software package in Bioconductor. So it does not make sense to allow pull requests and issues on those mirrors because the package maintainer may not even be aware of the repository and in any case does not have the permissions to accept pull requests on it or to mark issues as resolved. 

Therefore these repos do not have issues enabled, and pull requests will be automatically rejected by a bot.

Users can find your canonical git repos if you follow the instructions at
http://bioconductor.org/developers/how-to/git-mirrors/#how-do-i-let-users-know-i-am-using-github-for-development-and-contributions

Dan
#
----- Original Message -----
After some discussion, I've forked the read-only mirror to

https://github.com/Bioconductor/BiocStyle

This can be considered the canonical github repos for BiocStyle, where issues and pull requests should be submitted.

The URL and BugReports fields in the DESCRIPTION file have been updated accordingly.

Thanks,
Dan