<div>Good!</div><div><br></div><div>Just be sure your old repos didn't actually use nested subtrees, otherwise you could experience problems later.</div><div><br></div><div>Marco</div><div><br></div>On Thu, Sep 30, 2010 at 9:58 AM, Harry Flink <span dir="ltr"><<a href="mailto:to-harry-from-bzrmailinglist@steo.fi">to-harry-from-bzrmailinglist@steo.fi</a>></span> wrote:<br>
<div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">Thank you Marco!<br>
<br>
So simple solution and it works \o/.<br>
<br>
I havent tested it thoroughly yet but it<br>
seems I have all the revision history<br>
as should be and format is now "2a".<br>
<br>
I have only one repository... or I mean<br>
branch, I havent used shared repositories<br>
with the project but tested it long time<br>
ago. Thus I dont need script, thanks.<br>
<br>
Also "bzr log" and "bzr log -v" seems<br>
to be ALOT faster now than with old format.<br>
With old format "bzr log <subdirectory>"<br>
made bzr to crash but with new one I can<br>
finally get list of history related to<br>
a path.<br>
<br>
Thank you Marco again!<br>
<br>
Regards,<br>
Harry Flink<br>
<div><div></div><div class="h5"><br>
<br>
On Thu, September 30, 2010 10:21, Marco Pantaleoni wrote:<br>
> Hi Harry,<br>
> if you are sure you are not actually using subtrees, you could be able to<br>
> do:<br>
><br>
> mv repo old-repo<br>
> bzr init-repository --default-rich-root --no-trees repo<br>
> (for each BRANCH in old-repo)<br>
> bzr init repo/BRANCH<br>
> bzr pull -d repo/BRANCH old-repo/BRANCH<br>
><br>
> I also have a bash script that automates this for a large collection of<br>
> repositories and branches, if you are interested.<br>
><br>
> Marco<br>
><br>
> P.S. I agree with you that users should be less involved in repository<br>
> format juggling (in an ideal world, at least).<br>
><br>
> On Wed, Sep 29, 2010 at 10:51 PM, Harry Flink <<br>
> <a href="mailto:to-harry-from-bzrmailinglist@steo.fi">to-harry-from-bzrmailinglist@steo.fi</a>> wrote:<br>
><br>
>> Hello!<br>
>><br>
>><br>
>> Me and now many my co-workers are using<br>
>> bazaar as main version control system and<br>
>> really like it.<br>
>><br>
>> I'm managing quite a big project with it and<br>
>> over 2 years ago during jumping from SVN to BZR<br>
>> tried the nested tree support which wasnt so<br>
>> good idea. I got rid of the subtrees by just<br>
>> "bzr rm" them and just keep all needed files<br>
>> in the same big project. It seems to work OK.<br>
>><br>
>> During the 2 years I have upgraded the BZR<br>
>> directory format to newer formats but they<br>
>> always have to be with subtree support.<br>
>> (I dont get it why BZR users even have to<br>
>> care about internal repository file formats?)<br>
>><br>
>><br>
>> Now when branching my repo bzr (v2.1.1) gives<br>
>> me the following warning:<br>
>> "Fetching into experimental format RepositoryFormatKnitPack3().<br>
>> This format may be unreliable or change in the future without an upgrade<br>
>> path."<br>
>><br>
>><br>
>> Command "bzr info --verbose" shows the<br>
>> following info about the repository format:<br>
>> control: Meta directory format 1<br>
>> working tree: Working tree format 6<br>
>> branch: Branch format 6<br>
>> repository: Packs containing knits with subtree support<br>
>><br>
>> I'm a bit worried for the support of this<br>
>> respository format because I dont want to loose<br>
>> my history of all 4000 revisions and still want<br>
>> to keep up with latest versions of bzr.<br>
>><br>
>><br>
>> Is there any kind of method to drop history<br>
>> of subtrees so that I could go back using the<br>
>> "normal" repository formats instead?<br>
>><br>
>> I was hoping for something like SVNs<br>
>> "svnadmin dump" text format. I dont know if it<br>
>> could be possible to use some rebase/merge stuff<br>
>> to do this step by step with automated script<br>
>> by recreating it on another repository but skip<br>
>> the very old revisions that relate to subtrees.<br>
>><br>
>> I know this is my own fault being so stupid<br>
>> to test experimental features long time ago<br>
>> and not understand that these tests might<br>
>> affect long time in the future too.<br>
>><br>
>><br>
>> --<br>
>> Best Regards,<br>
>> Harry Flink<br>
>><br>
>><br>
>><br>
>><br>
><br>
><br>
> --<br>
> Marco Pantaleoni<br>
><br>
<br>
<br>
</div></div>--<br>
Best Regards,<br>
<font color="#888888">Harry Flink<br>
<br>
<br>
</font></blockquote></div><br><br clear="all"><br>-- <br>Marco Pantaleoni<br><br>