On Nov 28, 2007 4:00 AM, Alexander Knöller
<alexander.knoeller@neuland-bfi.de> wrote:
> We using Subclipse 1.2.4 with SVN-Kit with Eclipse 3.2 on macOSX 10.4
> (MacBook) for a JAVA project with a lot of small feature projects (week to
> week deployments).
> Everyting is developed in branches.
> So we often merge.
> We all experience the merging being very slow accompanied with heavy kernel
> activity of the CPU.
> Last time a colleague included the spring2-source-files in his branch (lots
> of small JAVA files).
> Yesterday he startet to merge this branch to a new branch. The merge is
> still running: Each little JAVA file which exists in the branch which is
> merged to the working copy takes about 1 minute to be merged, although those
> files do not exist in the working copy, yet.
> We wonder why it takes that long to detect, that the file does not exist in
> the working copy an ist to be added there? And why do the two CPU-cores
> suffer such a heavy kernel-usage?
> Is there a way to improve this setting?
Have you tried doing the same merge from the command line? Subclipse
does not do the work of the merge, we just run an API that is
equivalent to the command line. If you install this OSX binary, you
will get the command line client AND also be able to use JavaHL in
Subclipse:
http://downloads.open.collab.net/binaries.html
I would suggesting doing that so that we can rule out whether this is
an issue with SVNKit. The next thing would be to review the merge
options you are specifying.
--
Thanks
Mark Phippard
http://markphip.blogspot.com/
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subclipse.tigris.org
For additional commands, e-mail: users-help@subclipse.tigris.org
Received on Wed Nov 28 15:39:29 2007