>-----Original Message----- >From: Robert Menschel [mailto:[EMAIL PROTECTED] >Sent: Wednesday, August 11, 2004 8:46 PM >To: Chris Santerre >Cc: [EMAIL PROTECTED] >Subject: Re[2]: Header rule performance > > >Hello Chris, > >Wednesday, August 11, 2004, 7:11:42 AM, you wrote: > >>>I suspect that the "exists" version would be more efficient, use less >>>resources. ... > >CS> I believe this is true. exists: is much faster. But isn't that just >CS> for the header name, and no the contents in that header? I'm trying >CS> to remember this rule. But it looks like it also looks for the >CS> contenets within that header. Which I don't think exists: >checks for. > >Yes, exists: doesn't check for contents. However, our rule >doesn't check >for contents either -- it just checks to see whether the header's name >exists as "name:" anywhere in the headers (and would match even if that >name: were found in a Received header). > >Looks like I'll be converting to the exists: format. > >CS> But I could be wrong. > >No, you? Never!
I used to think that, then I got married. Now it seems I'm always wrong. ;) --Chris
