[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: Anyone want a server load generator called svn-annotate.py?

From: Brandon Ehle <azverkan_at_yahoo.com>
Date: 2003-07-29 10:01:12 CEST

>If a script depends on the bindings, that's one more requirement users
>have to meet before they can use it. When the bindings bring a
>significant benefit to the script, it's worth it; when they don't,
>imho it's better to not use them then.
>
>
I implemented a second mapping for the svn-annotate script that runs the
command line client instead of the SWIG bindings and I may have found a
reason why not to shell out to the command line.

This version uses cat and Python's internal diff library instead of
svn's diff command. Source for the new version is here
http://subversion.kicks-ass.org/

Running over ra_local I get these times. As you can see, the command
line client is much slower than the swig bindings. The overhead for
running the command line client seems to be pretty significant in this test.

SWIG Bindings
real 0m0.333s
real 0m0.317s
real 0m0.317s

Command Line Client
real 0m1.183s
real 0m1.169s
real 0m1.189s

When annotating http://svn.collab.net/repos/svn/trunk/README over
ra_dav, the cost of the command line isn't nearly as bad although it
still takes several seconds longer.

SWIG Bindings
real 0m53.558s

Command Line Client
real 0m57.320s

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Tue Jul 29 10:01:14 2003

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.