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

Re: [TSVN] Technical questions on status caching

From: Will Dean <svn_at_indcomp.co.uk>
Date: 2004-12-22 00:37:39 CET

At 23:38 21/12/2004 +0100, you wrote:

> Using svn_client_status, the entries fiel is only accessed one, the
> status of all files in the folder is retrieved, which is not efficient in
> my case because I have 3000 files in my folder.

Obviously there will be a cross-over at some point, where a small number of
very inefficient accesses is quicker than a large number of efficient
ones. However, I am surprised that you see it with only 3000 files.

I have just modified a part of TSVN to get the status of selected files
within a folder by getting the status of the entire folder, then discarding
the ones I don't want in the get_status callback. (Much as you were
advised to do on the SVN mailing list a few days ago, actually.)

This approach seems very fast indeed - I don't find that fetching the
status of the 3500 files is slow. Admittedly I'm on a very vast machine,
with a very fast disk-system, so my figures might not be what everyone
sees. I have made sure that my filtering algorithm doesn't introduce some
O(n^2) nonsense into the callback.

>It's right that svn_wc_status is not efficent if you have few files in the
>directory.

I'll take your word for it. It would be interesting to see some numbers,
to get an idea of the number of files at which the two techniques
cross-over in benefit.

Will

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tortoisesvn.tigris.org
For additional commands, e-mail: dev-help@tortoisesvn.tigris.org
Received on Wed Dec 22 00:38:40 2004

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

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