Re: cvs2git profile data

On 12/14/2010 06:01 PM, Alex Bennee wrote:
> On 14 December 2010 05:59, Michael Haggerty <[email protected]> wrote:
>> The expensive thing about tags is tabulating the possible parents for
>> each tag.  This is expensive for many-tag repositories because the
>> number of possible parent combinations scales something like
>>
>>    number_of_files * number_of_revisions_per_file *
>> average(number_of_tags_per_revision^2)
>>
>> It is the average of the square that is fatal.
> 
> So if I aggressively exclude some branches and tags will that have a
> positive impact?

Excluding branches or tags (in the sense of --exclude or
ExcludeRegexpStrategyRule) won't help here, because the exclusion occurs
after the possible parent statistics have been collected.  (Of course it
will save some time in the later processing.)  But ignoring *tags* (in
the sense of IgnoreSymbolTransform, which is not accessible via the
command line) will get rid of the specified tags right at the beginning
of the conversion and decrease the work of collecting possible parent
statistics.  OTOH ignoring *branches* with IgnoreSymbolTransform would
be a disaster, because it only turns the branches into unlabeled branches.

>> The fork calls come from
>> abstract_rcs_revision_manager.py:52(get_content), which suggests that
>> you are not using the internal code for extracting content from the CVS
>> repository.  The internal code should be much faster; I suggest you give
>> it a try.
> 
> I used the default options and didn't specify --use-cvs or --use-rcs
> so I'm not sure why it did that.

For cvs2git, the default is --use-cvs.  The relevant option for using
the internal code is called --use-external-blob-generator (or
ExternalBlobGenerator in the config file).

(Perhaps the default should be changed.)

Michael

-- 
Michael Haggerty
[email protected]
http://softwareswirl.blogspot.com/

------------------------------------------------------
http://cvs2svn.tigris.org/ds/viewMessage.do?dsForumId=1667&dsMessageId=2690735

To unsubscribe from this discussion, e-mail: [[email protected]].