On 06/23/2016 03:31 AM, Steffen Neumann wrote:
Hi,

are there any discussions to support github
as first class citizen source repository for github ?

I remember that some projects have a setup
where pull-requests to the main repository
are accepted automatically if they can be merged
cleanly. As improvement over the normal SVN workflow,
the pull request could trigger a single package build,
and be accepted only if that build passes without error.

I also experience the pain Kasper described in packages
I am involved in, and feel that this is an unnecessary burden.

Yes we do plan to support git repositories in a 'native' way, but obviously we are not there yet. Certainly our previous attempts at supporting git can put a lot of burden on those wishing to use it. It's clear that combining git and svn require 'expert' understanding of both version control systems, as well as deep understanding of the Bioconductor project development and build model.

The idea of build-on-commit is not in the immediate future; partly commits influence reverse dependencies and a successful build would involve building all of those; this becomes very expensive. Conversely some form of nightly build seem necessary anyway, to account for changes in packages outside Bioconductor control.

Martin Morgan


Yours,
Steffen

On Mi, 2016-06-22 at 22:46 -0400, Kasper Daniel Hansen wrote:
That's crazy.  Not so much that I have to do it once, but that I
apparently
will have to do this going forward forever.  I can vaguely understand
that
this is caused by the incompatibility between a linear only history
and a
branching history.

It also doesn't seem to work; when I try to cherry pick commits I get
stuff
like

$ git cherry-pick 94ef58b7ff5e24a9f36844f084ea3cebc89e846b
error: Commit 94ef58b7ff5e24a9f36844f084ea3cebc89e846b is a merge but
no -m
option was given.
fatal: cherry-pick failed

One workable, but unsatisfactory, solution is to just checkout the
master
branch onto the devel branch and then commit.  This will make svn
loose my
commit history, but at this points its not clear that it is worth the
time
to get my commit history into svn.

Best,
Kasper

On Wed, Jun 22, 2016 at 9:35 PM, Dan Tenenbaum <dtene...@fredhutch.or
g>
wrote:


My understanding is that this happens if you ever had non-linear
commit
history in your github repository (i.e. more than one branch
changed and
then were merged together), or if you used the old git/svn bridge.
As far
as I understand it, the only solution (in the near term) is to
cherry pick
the commits you want, a la


https://bioconductor.org/developers/how-to/git-mirrors/#dealing-wit
h-prior-history--merge-conflicts

And this has to be done every time you want to sync git to svn.

Dan


----- Original Message -----

From: "Kasper Daniel Hansen" <kasperdanielhan...@gmail.com>
To: "bioc-devel" <bioc-devel@r-project.org>
Sent: Wednesday, June 22, 2016 6:11:24 PM
Subject: [Bioc-devel] git svn dcommit

So I have made some updates to minfi and they are "safe" in
Github.  I
now

want to commit them to Bioconductor.  I get pretty strange
results,
although I never really understand what is going on.

1. I clone a fresh copy of minfi from Github.
2. I do
# bash update_remotes.sh
(using proper path of course)
3. I do
# git checkout devel
# git svn rebase
# git merge master --log
without any hiccups, and all seems to work so far.

Then I do
  # git svn dcommit --add-author-from

and hell starts (see output below).  What is kind of weird is
that now it
wants me to do 241 (!!!!!) rebase.  And it starts to involve
commits or
changes in files that haven't been touch for months.  This is
really
pretty

weird, given that I have completed the previous steps fine.

Any idea?  Otherwise I hope I can get special help at Stanford in
a few
days ...

Best,
Kasper






$ git svn dcommit --add-author-from
Committing to
https://hedgehog.fhcrc.org/bioconductor/trunk/madman/Rpacks/minfi
  ...
M DESCRIPTION

ERROR from SVN:
Item is out of date: Item
'/trunk/madman/Rpacks/minfi/DESCRIPTION' is out
of date
W: 85e070dd7d7825132ca3e6399f2e4dc19195b627 and
refs/remotes/git-svn-devel

differ, using rebase:
:100644 100644 0025283098af2a8f4376b1300533a4c5aabaea7b
8ac95ea78753b0a25d97c5debecbc6b2ead04ac0 M DESCRIPTION
:100644 100644 64a40f0b2655c7953ff154b42d6fa849e14462d3
92430c671d00d2131f5dc59cd3dbaa54679b585a M NAMESPACE
:040000 040000 732260942165586cd80d15ad9aa923e0969f604f
5098b8e0b0cabed337857680555df1ea6dc853f8 M R
:040000 040000 655f45f8ddd850a62a643151b73711c12ca0c5bb
9e9250da0089f89b9fb9ab986c617dbe66a3aab2 M inst
:040000 040000 3b781c09f0472bdff07ee734418b1bc263a31bd9
0f59f20b6cc89eaeeb94626a4445e3da26cef08f M man
:040000 040000 dd0723ff2712c28820000b81fafc1eb2de441e86
83da824cbedf66e528e93ba03abf4adb7e243669 M vignettes
First, rewinding head to replay your work on top of it...
Applying: updated citation information
Using index info to reconstruct a base tree...
.git/rebase-apply/patch:34: trailing whitespace.
beanplot,
.git/rebase-apply/patch:35: trailing whitespace.
RColorBrewer,
.git/rebase-apply/patch:36: trailing whitespace.
nor1mix,
.git/rebase-apply/patch:37: trailing whitespace.
siggenes,
.git/rebase-apply/patch:38: trailing whitespace.
limma,
warning: squelched 287 whitespace errors
warning: 292 lines add whitespace errors.
Falling back to patching base and 3-way merge...
Auto-merging vignettes/minfi.bib
CONFLICT (add/add): Merge conflict in vignettes/minfi.bib
Auto-merging man/utils.Rd
CONFLICT (add/add): Merge conflict in man/utils.Rd
Auto-merging man/read.450k.sheet.Rd
CONFLICT (add/add): Merge conflict in man/read.450k.sheet.Rd
Auto-merging man/read.450k.exp.Rd
CONFLICT (add/add): Merge conflict in man/read.450k.exp.Rd
Auto-merging man/read.450k.Rd
CONFLICT (add/add): Merge conflict in man/read.450k.Rd
Auto-merging man/preprocessQuantile.Rd
CONFLICT (add/add): Merge conflict in man/preprocessQuantile.Rd
Auto-merging man/mdsPlot.Rd
CONFLICT (add/add): Merge conflict in man/mdsPlot.Rd
Auto-merging man/getAnnotation.Rd
CONFLICT (add/add): Merge conflict in man/getAnnotation.Rd
Auto-merging man/estimateCellCounts.Rd
CONFLICT (add/add): Merge conflict in man/estimateCellCounts.Rd
Auto-merging man/densityBeanPlot.Rd
CONFLICT (add/add): Merge conflict in man/densityBeanPlot.Rd
Auto-merging man/cpgCollapse.Rd
CONFLICT (add/add): Merge conflict in man/cpgCollapse.Rd
Auto-merging man/bumphunter.Rd
CONFLICT (add/add): Merge conflict in man/bumphunter.Rd
Auto-merging man/blockFinder.Rd
CONFLICT (add/add): Merge conflict in man/blockFinder.Rd
Auto-merging man/RGChannelSet-class.Rd
CONFLICT (add/add): Merge conflict in man/RGChannelSet-class.Rd
Auto-merging man/GenomicRatioSet-class.Rd
CONFLICT (add/add): Merge conflict in man/GenomicRatioSet-
class.Rd
Auto-merging man/GenomicMethylSet-class.Rd
CONFLICT (add/add): Merge conflict in man/GenomicMethylSet-
class.Rd
Auto-merging inst/unitTests/test_preprocess.R
CONFLICT (add/add): Merge conflict in
inst/unitTests/test_preprocess.R
Auto-merging inst/NEWS.Rd
CONFLICT (add/add): Merge conflict in inst/NEWS.Rd
Auto-merging R/utils.R
CONFLICT (add/add): Merge conflict in R/utils.R
Auto-merging R/rset.R
CONFLICT (add/add): Merge conflict in R/rset.R
Auto-merging R/rgset.R
CONFLICT (add/add): Merge conflict in R/rgset.R
Auto-merging R/read.manifest.R
CONFLICT (add/add): Merge conflict in R/read.manifest.R
Auto-merging R/read.450k.R
CONFLICT (add/add): Merge conflict in R/read.450k.R
Auto-merging R/preprocessSwan.R
CONFLICT (add/add): Merge conflict in R/preprocessSwan.R
Auto-merging R/preprocess.R
CONFLICT (add/add): Merge conflict in R/preprocess.R
Auto-merging R/plot.R
CONFLICT (add/add): Merge conflict in R/plot.R
Auto-merging R/mset.R
CONFLICT (add/add): Merge conflict in R/mset.R
Auto-merging R/manifest.R
CONFLICT (add/add): Merge conflict in R/manifest.R
Auto-merging R/grset.R
CONFLICT (add/add): Merge conflict in R/grset.R
Auto-merging R/gmset.R
CONFLICT (add/add): Merge conflict in R/gmset.R
Auto-merging R/getSex.R
CONFLICT (add/add): Merge conflict in R/getSex.R
Auto-merging R/estimateCellCounts.R
CONFLICT (add/add): Merge conflict in R/estimateCellCounts.R
Auto-merging R/bumphunter.R
CONFLICT (add/add): Merge conflict in R/bumphunter.R
Auto-merging R/blocks.R
CONFLICT (add/add): Merge conflict in R/blocks.R
Auto-merging R/anno.R
CONFLICT (add/add): Merge conflict in R/anno.R
Auto-merging NAMESPACE
CONFLICT (add/add): Merge conflict in NAMESPACE
Auto-merging DESCRIPTION
CONFLICT (add/add): Merge conflict in DESCRIPTION
error: Failed to merge in the changes.
Patch failed at 0001 updated citation information
The copy of the patch that failed is found in: .git/rebase-
apply/patch

When you have resolved this problem, run "git rebase --continue".
If you prefer to skip this patch, run "git rebase --skip"
instead.
To check out the original branch and stop rebasing, run "git
rebase
--abort".

rebase refs/remotes/git-svn-devel: command returned error: 128

kaspers-air:~/Work/packages_github/minfi (devel *+|REBASE 1/241)
$

       [[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...{{dropped:2}}

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

Reply via email to