Oh, the git-svn-mirroring and git-svn-bridge projects are not really
an option for relax. One day if the Gna! infrastructure supports git
repositories, then maybe a migration from subversion to git would be
considered. But even then, such a disruptive change would not be
taken lightly. The git-svn option is a useful stop-gap, but branch
creation and management would need to remain as subversion operations
with the help of the svnmerge.py script.
Regards,
Edward
On 23 June 2013 10:31, Edward d'Auvergne <edward@xxxxxxxxxxxxx> wrote:
I've always wanted to take a look at this. Maybe I'll give it a go
now and see how it runs. I don't know how compatible this will be
with the svnmerge.py script used to manage the branch merging.
Regards,
Edward
On 23 June 2013 09:32, Troels Emtekær Linnet <tlinnet@xxxxxxxxx> wrote:
Maybe one should look at a git-svn solution?
http://learn.github.com/p/git-svn.html
https://github.com/soxofaan/git-svn-mirroring
https://github.com/mrts/git-svn-bridge
Best
Troels Emtekær Linnet
2013/6/19 Edward d'Auvergne <edward@xxxxxxxxxxxxx>
That is a different story, and I have considered it. But the
constraint is again the Gna! open source infrastructure - it does not
support git. Most of the time SVN works beautifully. But there are
cases where it isn't so great. For example I sometimes code without
internet access when on a train. For that I simply make lots of
copies of the checked out code and make separate modifications to each
:S Really not idea but you have to work within such constraints.
Regards,
Edward
On 19 June 2013 19:32, Troels Emtekær Linnet <tlinnet@xxxxxxxxx> wrote:
Allright.
But being at patch level, it is very hard to brake
a commit into several pieces.
Since there is no local repo, each commit has to be send
to online repo.
This could be solved by using git. :-)
But that would be a different story!
Best
Troels
Troels Emtekær Linnet
2013/6/19 Edward d'Auvergne <edward@xxxxxxxxxxxxx>
Hi,
Here is how I have broken the patch up into 3 separate parts:
The file permission problem (r20217):
http://article.gmane.org/gmane.science.nmr.relax.scm/17973
The switch from argparse to optparse (r20218):
http://article.gmane.org/gmane.science.nmr.relax.scm/17974
The grace2images.py script improvements (r20219):
http://article.gmane.org/gmane.science.nmr.relax.scm/17975
This allows us to revert one of these in the future. For example if
it is decided in the distant future that everyone has access to the
optparse module (in 5-10 years maybe), then we can revert revision
r20218. The file permission problem fix and script improvements would
then remain.
The svn repository allows us to play with the code like it is lego -
each commit is like a block. We can take out pieces when we like and
replace them with something else. Or we can remove a commit from a
branch and shift it into the trunk, if that is were the code belongs
(it then comes back to the branch using svnmerge.py but, importantly,
it must be removed from the branch first). Therefore it is very
important that each commit only consists of one idea. Otherwise such
shifting around of the commit code blocks will end up shifting too
much. I hope this is now clearer.
Regards,
Edward
On 19 June 2013 16:03, Edward d'Auvergne <edward@xxxxxxxxxxxxx> wrote:
Ok, I'll make the commits then post again to show how it was done.
Regards,
Edward
On 19 June 2013 16:00, Troels Emtekær Linnet <tlinnet@xxxxxxxxx>
wrote:
Hi Edward.
I am so confused now, what you want.
Can you fix it from here?
I will try next time to follow, but this is killing me.
Best
Troels
Troels Emtekær Linnet
2013/6/19 Edward d'Auvergne <edward@xxxxxxxxxxxxx>
Hi,
Don't worry about all the bouncing back and forth at the start.
This
process will become quicker with time as I have less to comment on.
It's just part of the initial learning curve for building up enough
knowledge as to how a large software project operates, to then
become
accepted as a developer with full commit access. Learning about
the
commit process - what should constitute a commit and how to write
the
commit message - is quite important for this.
Regards,
Edward
On 19 June 2013 15:43, Edward d'Auvergne <edward@xxxxxxxxxxxxx>
wrote:
Hi,
The fix for the pipe_control.palmer module is not necessary.
This
was
fixed in trunk - where I made the mistake - and is now merged
into
the
relax_disp branch. I was just a bit too slow for you ;) Also,
where
you make changes such as:
- file.write(" if \"PNG\" in types:\n")
+ file.write(" if (\"PNG\" in types or \".PNG\" in types or
\"png\" in types or \".png\" in types):\n")
this should be in separate patch as it does something different.
Ideally the switch to optparse and the chmod() call should also
be
separate as they are different, and they apply to different parts
of
the codebase.
Cheers,
Edward
On 19 June 2013 15:19, Troels E. Linnet
<NO-REPLY.INVALID-ADDRESS@xxxxxxx> wrote:
Follow-up Comment #8, bug #20916 (project relax):
Fix for bug #20916 (https://gna.org/bugs/?20916) Suggestion for
python
script
for PNG/EPS/SVG conversion of grace files
Optimized according to
(http://thread.gmane.org/gmane.science.nmr.relax.devel/3953)
Also a small fix to the pipe_control/palmer.py
since there where a wrong call to the import function.
Patch6 added
(file #18119)
_______________________________________________________
Additional Item Attachment:
File name: patch6 Size:8 KB
_______________________________________________________
Reply to this item at:
<http://gna.org/bugs/?20916>
_______________________________________________
Message sent via/by Gna!
http://gna.org/
_______________________________________________
relax (http://www.nmr-relax.com)
This is the relax-devel mailing list
relax-devel@xxxxxxx
To unsubscribe from this list, get a password
reminder, or change your subscription options,
visit the list information page at
https://mail.gna.org/listinfo/relax-devel
Re: [bug #20916] Suggestion for python script for PNG/EPS/SVG conversion of grace files