push' might only push the current branch...
Take the git push out of the equation.
Is the local repository missing those commits?
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Software Architect - Intel Open Source Technology Center
ohn, there were two replies to your last two posts about the same core dump.
You have not replied to any of them.
Please reply and explain how you've addressed the suggestions or why they
don't apply.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Software
gt; with the following:
Is this an incremental import on top of what you had before? If so, please
start from the beginning of history.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Software Architect - Intel Open Source Technology Center
igned to allow resuming, but I recommend doing a full
conversion when you're going to do it for good. Just in case.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Software Architect - Intel Open Source Technology Center
tory
>
> match /Labs/ChattyApps/.*.project|.xml?$
> repository /Labs/ChattyApps
> branch master
> end match
>
> It complains about the path not being a root directory or something.
Can you get the exact error message?
--
Thiago Macieira - thiago (AT) macieira.info -
@calypso.com> 1321679712 +
> Could you please help me out how can I resolve
There looks like there's a problem in the way the tool encodes a user name
that contains an email address.
You have to fix the source code.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT)
eplaying the history is easily done, but creates so much trouble
afterwards...
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Software Architect - Intel Open Source Technology Center
PGP/GPG: 0x6EF45358; fingerprint:
E067 918B B660 DBD1 105C 966C 33F5 F005 6E
d also note that tagging has nothing to do with direct pushing.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Software Architect - Intel Open Source Technology Center
PGP/GPG: 0x6EF45358; fingerprint:
E067 918B B660 DBD1 105C 966C 33F5 F005 6EF4 5358
si
ted
> tags [1], so it is possible to push tags to Gerrit just fine.
His point was: think!
Qt has no direct pushes. Qt has tags.
Take your own conclusion.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Software Architect - Intel Open Source Technology Center
---cut here---start->8---
> match /(branches|tags)/KDE/[^/]+/$
> action recurse
> end match
> --8<---cut here---end--->8---
>
> in converted/KDE/kdelibs-rules from the KDE ruleset?
Yes. Note the $ at the en
As you can see, there's no file path changing that would match
/branches/KDE/(\d.\d)/kdelibs.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Software Architect - Intel Open Source Technology Center
PGP/GPG: 0x6EF45358; fingerprint:
E067 918B B660 D
s that it looks at the list of changed files from
SVN. In the svn cp that created the 3.5 branch like above, the repository root
is not listed, so it doesn't match any rules. You add a "recurse" rule to make
the tool list the repository at the specified revision and then apply the r
I spend one
minute logging in and cleaning up the queue.
So there's no need to add to your responsibilities.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Software Architect - Intel Open Source Technology Center
PGP/GPG: 0x6EF45358; fingerprint:
E067 918
Dec 1st. That gives us 45 days
to wrap up any subjects pending.
Any objections?
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Software Architect - Intel Open Source Technology Center
PGP/GPG: 0x6EF45358; fingerprint:
E067 918B B660 DBD1 105C 966C 33F5 F005 6
y and find which commits
are missing. Right now, it only needs to send the list of branch heads.
On the sender side, it needs to add that information to its tree scanning.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Software Architect - Intel Open Source Technology Cen
Isn't it?
I understand why you can never update it (fetch into it), since the remote may
conclude that it doesn't need to send some objects.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Software Architect - Intel Open Source Technology Center
PGP/GP
u're jumping from branch to branch.
Anyway, the point is made: since you are excluding less than half at each
iteration, the number of iterations to find the culprit is actually bigger than
a linear history.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Produc
to normal linear bisecting -- the merge commit
being the faulty one would be a worst-case scenario, but also an unlikely one)
The point is that whenever you have a fault between a branch and a merge, you
need to stop bisecting and instead exclude branches. That is not bisecting and
could, under
Em quarta-feira, 9 de fevereiro de 2011, às 04:21:03, Boyd Stephen Smith Jr.
escreveu:
> In <20110209100027.be9157...@nargothrond.macieira.info>, Thiago Macieira
>
> wrote:
> >You don't rebase someone else's commits. You only to do that to your own.
> >
> &g
tributed
> development, it is best to avoid rebase when possible. (It can be quite a
> good tool for rewriting bad commits.)
I don't agree either. Rebasing is a very powerful tool and should be used
wisely, when it makes sense. It does not impede collaboration.
--
Thiago Macie
ly, you should have git rerere
> > > enabled,
> > > so you won't have to repeat resolving any possible conflicts.
>
> Excuse my ignorance... But what is "git rerere"?
man git-rerere
NAME
git-rerere - Reuse recorded resolution of conflicted merges
--
Thiag
es, managed by the Release Manager only. He
cherry-picks necessary fixes and makes the packages. In Qt's case, the number
of changes can be in the order of hundreds.
Once a release is made, the RM tags the last commit in that branch, merges it
back to the mainline (using strategy "our
mmits)
> and then garbage-collecting the now-unreferenced redun-
> dant commits away.
Please remember to set the tagging date and author properly.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
PGP/GPG:
N tags to Git tags. If you committed
to the SVN path that was a tag, then it was really a branch. The proper way to
move a tag is to delete it and re-tag.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
PGP/GP
repack is that it actually increases the
> size of the repo. A repo went from 360 all the way to 600.
>
> But it'd be worth playing around with this sort of thing before Dec 20.
You did something wrong then.
git-fast-import produces very poorly-packed repositories. Repacking is
with those moved files. (I can check out a branch before the
> move for example and then I have all that history, of course. But since
> the files were just moved, I'd expect this history to show up for the
> current files, too).
git log -M
--
Thiago Macieira - thiago (AT) macieira.i
recurse' rule on
> /trunk/KDE/kdebindings/ruby/qtruby/ would fix this, but apparently it
> doesn't.
> Does anyone know what's wrong with the rules? (Rules attached).
> The relevant commit is r653117 [0].
There's nothing wrong.
--
Thiago Macieira - thiago (AT
, you don't delete or move at all when
> using the split approach as that makes no sense.
That doesn't include the case of a full-module delete.
So we'll still have a class of modules in the "unmaintained" state. But that's
something outside of the repository, n
On Saturday 25. September 2010 11.41.39 Niko Sams wrote:
> 2010/9/25 Thiago Macieira :
> > On Saturday 25. September 2010 10.32.31 Niko Sams wrote:
> >> Is such a change from a link to a file not supported by svn2git?
> >> I tried simply ignoring the error but that didn&
On Saturday 25. September 2010 10.32.31 Niko Sams wrote:
> Is such a change from a link to a file not supported by svn2git?
> I tried simply ignoring the error but that didn't work
Replacing a symlink with a file in one commit? Probably runs into a bug.
--
Thiago Macieira -
If you wish to move to the development
version, you have to cd into the subdirectory and issue the appropriate update
commands.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
PGP/GPG: 0x6EF45358; fingerprint:
E0
e 1.0.3.1 tag was created from branch 1.0.3.x but
> from a commit that only touches trunk.
There's nothing wrong with that. You can copy a path in any revision: it's
simply the state in which that path was in that revision, as affected by the
latest commit before that revision at
hat's all that the user has to do, that means the user has just downloaded
way more than was necessary -- about 100% more with SVN or something similar
with Git.
After unpacking, there has to be some operation to restore the checkout from
the repository data.
--
Thiago Macieira - thiago (AT)
On Friday 9. July 2010 20.34.28 you wrote:
> fatal: Empty path component found in input
>
> and at the end of "most recent commands before crash"
>
> * M 100644 :7143536 /lib/model/om/BaseReportsourcePeer.php
You have a match rule without the required end slash.
--
Em Domingo 27. Junho 2010, às 15.58.01, Johannes Obermayr escreveu:
> On Sunday 27 June 2010 15:35:35 Thiago Macieira wrote:
> > Em Domingo 27. Junho 2010, às 15.13.19, Johannes Obermayr escreveu:
> > > To keep complete history svn2git must offer pushing commits in "branch
put. This should be solvable with a keyword like
> "dir"
You don't need that. Import into separate branches, at the root, then use git
mv and git merge once the import is complete.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product M
Em Terça-feira 22. Junho 2010, às 09.20.53, Ben Cooksley escreveu:
> 2010/6/22 Thiago Macieira :
> > Em Terça-feira 22. Junho 2010, às 06.38.12, Ben Cooksley escreveu:
> >> Just a further update, that the file EOL style hook has now been
> >> ported to Git, and was pla
min
> Note that unlike the CIA and EMail hooks, which operate only on the
> main project repositories, the EOL hook runs on all repositories,
> including clones.
Does it respect .gitattributes settings? E.g., how do I add a CRLF file for,
say, testing?
--
Thiago Macieira - thiago (AT) ma
the first time is not straight-forward, though:
>
>git push origin my-topic:refs/personal/hausmann/my-topic
>
> and I don't know whether cgit and redmine can show or work with such refs,
> and any cleanup is still not automatic ;)
That's no different than a separate repo.
. Thats what
> Gitolite does.
We also need repositories connected to wikis. I don't like that the playground
today is a mess of stuff no one knows about.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
n branches of the FreeBSD svn repo converted
> correctly.
I don't have time to look into this.
Can someone else in the mailing list review Uli's patch and apply to the
mainline svn2git?
Also, the code I suggested isn't hard to implement. Someone please look into
it.
--
Thiago Maci
f + "\nfrom " + branchFromRef +
"\n\n");
It needs to replace
"\nfrom " + branchFromRef
with
"\nfrom :" + QByteArray::number(mark)
Note, however, that this list of commits would not be saved, which means this
process would work only for a f
at svn2git cannot handle, or am I missing
> something?
svn2git cannot handle it.
When you cp a branch, it takes the latest revision of the branch. It doesn't
know which Git commit matches the SVN commit you copied.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.or
d the repository first to your machine.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
PGP/GPG: 0x6EF45358; fingerprint:
E067 918B B660 DBD1 105C 966C 33F5 F005 6EF4 5358
signature.asc
Description: Thi
an atomic move.
> >
> > No. Then you no longer have a correct import of the history.
>
> does that worry you? i find a useful history more important than an
> accurate one. that's the whole idea behind rebase -i, after all.
Then fixup the history with git filter-branc
Em Domingo 28. Março 2010, às 16.54.47, Thomas McGuire escreveu:
> Hi,
>
> On Sunday 28 March 2010 16:06:59 Thiago Macieira wrote:
> > Em Domingo 28. Março 2010, às 15.29.20, Thomas McGuire escreveu:
> > > > That's how Git works. Never mind the issue.
> &g
th svn2git?
No, the history isn't gone if the older files are listed as deleted any time
before. Your email confirms that.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
PGP/GPG: 0x6EF45358; fingerprin
adi/libakonadi
> -> kdepimlibs/akonadi at r788342. The history is there but git cannot
> follow the current HEAD backwards across the border.
That's how Git works. Never mind the issue.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Deve
Em Sábado 13. Março 2010, às 08.29.51, você escreveu:
> On Sat, Mar 13, 2010 at 15:08, Thiago Macieira wrote:
> > Em Sábado 13. Março 2010, às 00.57.57, Niko Sams escreveu:
> >> Hi,
> >>
> >> I still working on quanta conversion and need some advice (Thi
Em Sábado 13. Março 2010, às 06.39.52, Niko Sams escreveu:
> On Sat, Mar 13, 2010 at 15:08, Thiago Macieira wrote:
> > Em Sábado 13. Março 2010, às 00.57.57, Niko Sams escreveu:
> >> Hi,
> >>
> >> I still working on quanta conversion and need some advice (Thi
of doing that in svn2git because the history is actually
wrong in SVN, from the way that svn2git looks at it.
I'd recommend that you do the import as-is and fix it up in Git later.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Developm
Em Quinta-feira 11. Março 2010, às 11.30.59, Andreas Pakulat escreveu:
> On 11.03.10 07:36:41, Thiago Macieira wrote:
> > Em Quinta-feira 11. Março 2010, às 07.23.53, Torgny Nyblom escreveu:
> > > Well I wan't to import the files into the same directory as they exit
> &g
Em Quinta-feira 11. Março 2010, às 07.23.53, Torgny Nyblom escreveu:
> On Thursday 11 March 2010 16.16.53 Thiago Macieira wrote:
> > Em Quinta-feira 11. Março 2010, às 06.59.24, Torgny Nyblom escreveu:
> > > On Thursday 11 March 2010 14.32.32 Andreas Pakulat wrote:
> >
cha/submodule you'll have to create a separate
> > repository.
>
> Ok, thanks for that.
>
> But how do I get the files (in the separate branch) into the correct
> subdir?
You can't create history that doesn't exist in SVN, that includes getting files
in
Em Quarta-feira 10. Março 2010, às 12.41.31, Niko Sams escreveu:
> 2010/3/10 Thiago Macieira :
> > Em Quarta-feira 10. Março 2010, às 12.18.50, Niko Sams escreveu:
> >> Hi,
> >>
> >> I'm trying to write conversion rules for quanta.
> >
catch this case?
That's the branching Quanta 3.1, done the cvs2svn way.
That branch was originally a CVS branch. When it got converted to SVN, cvs2svn
branches /trunk, then deletes what didn't get branched. That's what the commit
says.
--
Thiago Macieira - thiago (AT) macieira.info -
Is this possible?
Hi Torgny
It doesn't reprocess history. The recurse action lists the directory you're in
and applies the actions to the files and directories contained within. It's a
one-level deep action and only for the current revision.
--
Thiago Macieira - thiago (
You can rewrite the tool :-)
As it stands, you need to carefully write the rule so that only that file
matches.
Directories are ignored in import rules. They are useful only in recurse
rules.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager -
Em Domingo 28. Fevereiro 2010, às 01.16.40, Oswald Buddenhagen escreveu:
> On Sun, Feb 28, 2010 at 12:31:17AM +0100, Thiago Macieira wrote:
> > Em Domingo 28. Fevereiro 2010, às 00.08.15, Andreas Pakulat escreveu:
> > > Well, its ugly compared to those tags that don't have a
out where the tag
and/or branch was copied from, so it couldn't generate the proper Git history.
I can think of three possibilities:
1) bug in the tool
2) the SVN commits were not copy-with-history
3) the SVN paths they were copied from did not match any rules
Note that "match any ru
oesn't work for partial checkouts.
> Last but not least it seems there are some tags that don't relate to any
> history at all. I'm guessing they've been created from branches that
> don't exist anymore right now. Am I right that adding a rule for the
> path with a spe
e the tool not crash on it
> ;)
Handling slashes properly requires a major refactoring of the SVN code inside
the tool.
But it should be done, because it avoids those mistakes. And especially
because some rules must not end in a slash (the ignore ones).
--
Thiago Macieira - thiago (A
I can't imagine cloning 2 million commits when I want to clone
> kdelibs.git by 2015 :-O
Cloning everything now with Git isn't much larger than cloning everything now
with SVN.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Develop
g repositories is the way to go, I didn't pay
attention to the list of requirements from Oswald.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
PGP/GPG: 0x6EF45358; fingerprint:
E067 918B B660 DBD1
ens of git repositories:
>
> http://source.android.com/download/using-repo
Repo helps with downloading and maintaining the source tree. It doesn't help
with the dependency tree -- that's actually normal Makefiles.
And it doesn't do atomic commits across repositories nor mov
s on
server2. Done, you no longer have atomicity because of a technical decision.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
PGP/GPG: 0x6EF45358; fingerprint:
E067 918B B660 DBD1 105C 966C
-balancing servers. Imagine
the case that someone is using a repository somewhere else while working in a
branch.
What's the point in a distributed VCS that needs to talk to one specific
server?
Anyway, the point being: unless it's distributed, the Git community will not
accept the
ne server.
But with a client-driven locking, the servers might want to time out the
locking, which means you'll be limited in time (and thus amount of data) that
you can "atomically" push.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manag
Em Sexta-feira 12. Fevereiro 2010, às 23.59.43, Oswald Buddenhagen escreveu:
> On Fri, Feb 12, 2010 at 11:19:29PM +0100, Thiago Macieira wrote:
> > Em Sexta-feira 12. Fevereiro 2010, às 19.05.51, Oswald Buddenhagen
escreveu:
> > > one thing is truly transparent submodule h
you what's missing".
Atomicity and multiple repositories is impossible. Forget it.
Choose one or the other.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
PGP/GPG: 0x6EF45358; fingerprint:
's up to the distro to decide.
No, we can't do that.
The moment that we start splitting the repositories, the CMakeLists.txt files
will start to drift. It will be impossible to do in-module and out-of-module
builds with the same source code.
--
Thiago Macieira - thiago (AT) maciei
Em Sexta-feira 5. Fevereiro 2010, às 22.38.58, Ingo Klöcker escreveu:
> On Sunday 31 January 2010, Thiago Macieira wrote:
> > Em Domingo 31. Janeiro 2010, às 14.39.31, Maciej Mrozowski escreveu:
> > > Source distro packagers (like here in Gentoo) surely prefer
> > > spl
Em Domingo 31. Janeiro 2010, às 14.39.31, Maciej Mrozowski escreveu:
> On Sunday 31 of January 2010 08:59:17 Thiago Macieira wrote:
> > Em Domingo 31. Janeiro 2010, às 02.06.58, Oswald Buddenhagen escreveu:
> > > afaict, it has worked so far (except for the mess in kdebase).
&g
what. We
want you to use this script to build KDE in your distribution (yeah, forget
your current dependency mechanisms too)."
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
PGP/GPG: 0x6EF45358; fingerp
Em Sábado 30. Janeiro 2010, às 23.22.56, Albert Astals Cid escreveu:
> A Dissabte, 30 de gener de 2010, Thiago Macieira va escriure:
> > Em Sábado 30. Janeiro 2010, às 16.34.27, Oswald Buddenhagen escreveu:
> > > > You don't link to applications. You link to librar
Em Sábado 30. Janeiro 2010, às 16.56.17, Oswald Buddenhagen escreveu:
> On Sat, Jan 30, 2010 at 04:49:40PM +0100, Thiago Macieira wrote:
> > Em Sábado 30. Janeiro 2010, às 16.34.27, Oswald Buddenhagen escreveu:
> > > > You don't link to applications. You link to librar
he interface
> - the application may provide an internal library to link against
It's not internal. It's a public library, with installed headers.
It is so because the hypothetical plugin we have in mind here comes from a
separate module.
--
Thiago Macieira - thiago (AT) macieir
Em Sábado 30. Janeiro 2010, às 12.19.20, Oswald Buddenhagen escreveu:
> On Sat, Jan 30, 2010 at 09:22:58AM +0100, Thiago Macieira wrote:
> > Em Sábado 30. Janeiro 2010, às 02.35.44, Michael Jansen escreveu:
> > > We would loose all that flexibility. if we consider moving with
&
all that flexibility. if we consider moving with history as
> an essential point.
According to the plan I posted, that's not allowed.
You can only depend on Libraries. You cannot depend on applications installed.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Se
Em Sexta-feira 29. Janeiro 2010, às 18.19.11, Oswald Buddenhagen escreveu:
> On Fri, Jan 29, 2010 at 05:52:49PM +0100, Thiago Macieira wrote:
> > Em Sexta-feira 29. Janeiro 2010, às 14.36.32, Matt Williams escreveu:
> > > The only problem I see here is how to move a game
need common history to perform a merge.
> This is also a problem when we want to move libs or apps
> out of a module (into playground/unmaintained or another main
> module).
Now that's difficult to keep history. Not impossible, but difficult.
--
Thiago Macieira - thiago (AT)
e dependency file could be used by
packagers to update their packages' build and install-time dependencies.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
PGP/GPG: 0x6EF45358; fingerprint:
E06
Em Quinta-feira 28. Janeiro 2010, às 00.58.59, Oswald Buddenhagen escreveu:
> On Wed, Jan 27, 2010 at 10:46:52PM +0100, Thiago Macieira wrote:
> > [...] The build rules are: [...]
>
> why not (mostly) keep our current layout in form of a meta structure (be
> it git submodules or
's orthogonal to any discussion on Git or how you make your patch land in
both branches.
The fact is you have to test your fix.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
PGP/GPG: 0x6EF45358; fingerpr
Em Quarta-feira 27. Janeiro 2010, às 22.46.52, Thiago Macieira escreveu:
> Q: What are the flaws in this design?
> A: Points not addressed:
> - most libraries depend on Qt, but if Qt is in KDE Support, then there's
> an order problem
> - it's impossible to have addons/
essed:
- most libraries depend on Qt, but if Qt is in KDE Support, then there's an
order problem
- it's impossible to have addons/plugins to apps
- it's impossible to have libraries depend on other libraries
And other stuff I've missed.
--
Thiago Macieira - thiago (AT) macieira
Em Quarta-feira 27 Janeiro 2010, às 14:46:21, Oswald Buddenhagen escreveu:
> On Wed, Jan 27, 2010 at 02:01:28PM +0100, Thiago Macieira wrote:
> > Git submodule isn't atomic at all. It cannot be used to indicate
> > atomicity because it requires at *least* two operations (two
kdegit-build script to clone, check out and update the
code, to track dependencies, etc.
So to bootstrap your KDE build, you'd clone the repository that contains the
manifest and the build tool. When you run the tool, it clones the remaining
repositories. Next time, it'll simply update
u change the stable branch, you should be really careful and test it. I
mean, who backports stuff and doesn't test that the fix worked?
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
PGP/GPG: 0x6EF45358;
Em Quarta-feira 27 Janeiro 2010, às 12:38:21, Thomas Zander escreveu:
> On Wednesday 27. January 2010 10.45.09 Thiago Macieira wrote:
> > You don't update the supermodule unless you want to say something.
> >
> > For example, when we're about to tag, tag all reposit
po$ touch INSTALL; git add INSTALL; git commit -m "Added INSTALL"
[master 83b2950] Added INSTALL
0 files changed, 0 insertions(+), 0 deletions(-)
create mode 100644 INSTALL
/tmp/superrepo$ git pull --rebase . b2
From .
* branchb2 -> FETCH_HEAD
First, rewinding head to r
gt;
> ...unless it's possible to write a set of hooks that commit to the super-
> module every time any of the submodules' repos gets a commit?
> it sounds like it's got other problems, though, and makes it easy to make
> mistakes. and I feel like I'm forgetting som
bugfixes. Just apply all bugfixes
to the lowest/earliest branch where the fix makes sense.
> Also, that probably just moves the problem around: If I created a branch
> for each bugfix, I would then need a list of _branches_ that are not
> merged yet, together with a way to block those.
g
;t accept your push."
exit 1
fi
fi
git cat-file will return with an error if the commit doesn't exist. If it
doesn't exist, it can't be part of any branch.
git log A..B lists the commits between A and B. The -n1 option makes
d gaps filled in
>
> in theory cvs2svn should have already mapped that. obviously some
> entries were missing or an entire case class was not considered at all.
Just merge accounts with disabled-accounts before generating the map file.
--
Thiago Macieira - thiago (AT) macieira.info - thi
ng commits, yet there are still a lot of
unreviewed commits going in. More importantly: there is no hook to try and
enforce the policy.
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
PGP/GPG: 0x6EF45358; fin
Em Sexta-feira 11. Dezembro 2009, às 00.29.59, Chani escreveu:
> On December 10, 2009 15:13:44 Thiago Macieira wrote:
> > Em Quinta-feira 10. Dezembro 2009, às 18.09.14, Jeff Mitchell escreveu:
> > > > We aren't going to switch to Gitorious with the current merge reques
ture we didn't have in SVN?
--
Thiago Macieira - thiago (AT) macieira.info - thiago (AT) kde.org
Senior Product Manager - Nokia, Qt Development Frameworks
PGP/GPG: 0x6EF45358; fingerprint:
E067 918B B660 DBD1 105C 966C 33F5 F005 6EF4 5358
signature.asc
Description: This is
go back to problem #1 above, which I think requires doing two sweeps
of the SVN history: the first one backwards, starting from the current
location, finding out where it came from, then the second one forwards, doing
the import.
Maybe what we can learn from this is we can have a &qu
Em Quinta-feira 03 Dezembro 2009, às 11:03:27, Thiago Macieira escreveu:
> Em Quinta-feira 03 Dezembro 2009, às 10:20:11, Thomas Zander escreveu:
> > On Thursday 12. November 2009 19.41.21 Thomas Zander wrote:
> > > As some movement has been detected I want to point out that the
1 - 100 of 218 matches
Mail list logo