Geoff Hutchison (Geoffrey.R.Hutchison@williams.edu)
Wed, 08 Jul 1998 12:49:21 -0400 (EDT)
On Wed, 8 Jul 1998, Andrew Scherpbier wrote:
> If you know how to get apache to do that, let me know. We've tried to
> followed the directions in the docs, but we never got it to use a compressed
Ah!I found out what the "problem" is... Apache doesn't do on-the-fly
compression. If I compress an HTML doc, lynx decodes it perfectly (showing
the content-negotiation worked). But perhaps I can rig up a small action
CGI to compress HTML on-the-fly if you request a .html.gz file...
> I guess it depends on what type of duplicates are more common... If it has to
> do with the server having CNAMEs, then checking the mod time is fine, but if
> there are actual copies that wouldn't solve anything. It should probably be a
> configurable item.
That sounds good. The checksum duplicate-detection itself should probably
be a configuration option too. However, I bet the performance hit for
those of us with few duplicates will be much--something for testing.
This archive was generated by hypermail 2.0b3 on Sat Jan 02 1999 - 16:26:51 PST