2017-08-05 1:13 GMT+02:00 Mark Wielaard <m...@klomp.org>:
> On Fri, 2017-08-04 at 21:58 +0200, Mark Wielaard wrote:
>> > How the test migration was performed:
>> > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>> > See recipes at https://github.com/ivosh/valgrind-git-migration
>> I'll follow that to recreate the git repository with all branches
>> and put it back on sourceware.
> That are some very nice instructions. Thanks!
> It does take a while though and eats up lots of disk space.
> I didn't put the result on sourceware. Since I am not sure that wouldn't
> screw up something. I placed a copy at:
> https://code.wildebeest.org/git/user/mjw/valgrind/
> Looking at the result it also doesn't have the svn/VEX_x_YY_BRANCHes I
> assumed would be there. Looking at the migration steps I think they
> weren't supposed to be?
> If so, then I didn't actually mess up and you might want to change Step
> 16 to do a git gc --prune=now --aggressive. It really saves a lot of
> space (and it gives everybody that git clones a much more compact
> repository).
> You also might want to consider moving Step 15. Add SVN->GIT patches
> after Step 17 verification. It causes the branches.diff to look scary.
> The diffs also get slightly confused by the empty VEX/docs directory. We
> probably should remove that directory before/after migration.
> The tags.diff show a couple of "off by ones" (if you could call them
> that). Which look mostly harmless (configure.ac version
> with/without .SVN postfix and some copyright year updates), but for
> 3.0.1, 3.2.3 and 3.4.1 they look slightly bigger. I haven't investigated
> why yet. I am also not sure it really matters too much given how old
> these tags are. The last 5 tags/releases look spot on perfect.

Hi Mark,
I incorporated all your suggestions to the latest migration recipe
and refreshed https://sourceware.org/git/?p=valgrind.git.
Thank you for going through this with me!

Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
Valgrind-users mailing list

Reply via email to