[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#617296: Any Progress with RStudio?

[Bug CCed since the discussion was started there - should be moved
 to Debian Med devel list, but I personally do not have to say much
 more probably.]

Hi Charles,

On Thu, May 15, 2014 at 04:24:06PM +0900, Charles Plessy wrote:
> Part of the bitrot might be caused by the fact that on my side, I have stopped
> working on R packages because the FTP team made quite clear that my
> ...
> The main
> problem with this rule is that it makes it too hard to introduce new packages.

While I agree that geting new R packages accepted has become harder by
the new rule I think this is totally different from the issue Dirk was
addressing.  He is claiming since a long time that we are not coping
with upstream in our packaging and we should rather add less packages.
Even if you named a counter example it is usually not needed in most
cases to add new dependencies - we are simply lagging behind upstream
despite the fact that usually upgrading a R package is brain dead
simple.  I agree that the pure fact Dirk is stating is true:  Yes, we
are lagging behind upstream.

My point is that we are delivering Debian stable *releases* to
scientists and we are offering a good amount of quite up to date
software in testing/unstable.  I think there was not a single request
for a "new upstream version available" from any user who was not
answered right in time (= less than one week) accept in the case of
severe technical difficiulties (for instance r-bioc-cummerbund took
quite long for a long chain of new dependencies in times of slow
acceptance of packages).  But these difficulties are not the point here.

I personally do not see my time well spent to hunt behind any minor CRAN
package update which are quite frequent.  I rather do regular overhaul
of the R packages at certain points in time (preferably) in time before
we freeze.  I also plan such an overhaul in June to get all the
BioConductor packages refreshed.  Everybody else is really welcome to
push needed changes to SVN / Git and ping me for an upload.  However,
the simple fact that some R package has a new upstream version does not
pull a trigger on my side.  Dirk has simply a different opinion which is
fine for me but I'm bored to discuss this over and over.

Kind regards



Reply to: