Integration of Mercurial into LXR

[andre-littoz]

I incur the same slow performance. I also replaced Apache by lighttpd and the timings remain the same.

Mercurial, Python and web servers are not the problem. But what?

Retrieving a fully annotated file through piping to Mercurial involves the web-server, Python and Mercurial (with a similar loop as the troublesome one) and it works like a charm: near instantaneous display despite the graphic HTML enhancement made in my code.

What I don't understand is: why is this so simple loop causing problem while a sophisticated one has lightning speed?

André



________________________________
 De : Matt Mackall <mpm at selenic.com>
À : andre-littoz <page74010-sf at yahoo.fr> 
Cc : Simon King <simon at simonking.org.uk>; Mercurial mailing list <mercurial at selenic.com> 
Envoyé le : Mercredi 19 décembre 2012 16h51
Objet : Re: Integration of Mercurial into LXR
 
On Wed, 2012-12-19 at 14:39 +0000, andre-littoz wrote:
> What happens if you replace the call to hg with a call to a simple python script that just prints some constant strings? (ie. try to eliminate hg as the source of the problem). Make sure it uses the same interpreter that the hg command is using. 
> 
> Simon
> 
> ======================================================================
> >>>>> Bad! It takes also seconds to retrieve data with the same symptoms: standard time under the shell (instantaneous) and seconds under the Web server.

Now you should try replacing Python with 'echo' or 'cat'. Python is
almost certainly not the problem either.

-- 
Mathematics is the supreme nostalgia of our time.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mercurial-scm.org/pipermail/mercurial/attachments/20121219/b257ae01/attachment-0002.html>