Hello Robert,
I would like to add that direct performance comparision of CVS vs Subversion
doesn't consider certain aspects of Subversion that could affect real
performance. For example, during update operation only file deltas are sent
over network, not full file contents, also apache HTTP server may compress
the data it sends. Comparision against "BASE" version of the file (to review
local, not yet committed modifications) is performed without establishing
network connection at all.
So, you should consider that overall Subversion client performance may be
better then CVS one in general (due to the less amount of data sent over
network).
Alexander Kitaev,
TMate Software,
http://tmate.org/
http://jetbrains.com/tmate/
> -----Original Message-----
> From: Robert Einsle [mailto:robert@einsle.de]
> Sent: Tuesday, January 31, 2006 2:27 PM
> To: users@subclipse.tigris.org
> Subject: [Subclipse-users] Performance ofConnections
>
> Hy List,
>
> we testet the Performance of the Connections to Repository
> (CVS, HTTP, SVN), and the interesting fact is:
>
> The Timereference is the CVS-Connetion to an Repository.
> Connecting to the SVN-Repo via HTTP and JAVA-HL takes about 4
> Times then CVS Connecting to the SVN-Repo via HTTP and
> JAVA-SVN takes about 2 Times Connecting to the SVN-Repo via
> SVN takes nearly the same time like CVS.
>
> We tested it on Local Network (100 MBit switched) and on
> VPN-Connections.
>
> Is it posible to speed up the Connections via HTTP, or ist it
> posible to let lvnserve authenticate via pam (to an LDAP-Server)?
>
> Thanks for Help
>
> \Robert
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe@subclipse.tigris.org
> For additional commands, e-mail: users-help@subclipse.tigris.org
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subclipse.tigris.org
For additional commands, e-mail: users-help@subclipse.tigris.org
Received on Tue Jan 31 12:38:02 2006