cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Andy Lewis <>
Subject RE: xBug -Reply
Date Mon, 24 Jan 2000 18:44:19 GMT
I agree here . The cache should be able to handle more than just an HTML

The first point I guess though is philosophical. I personally adhere to the
thought that a cached object will almost always be faster than a generated
one. Having seen Cocoon 1.5 under load, I am convinced that scalability
depends on caching. The objective should be to never process the item twice.

Andy Lewis
eBusiness Development Manager
VERITAS Software, Heathrow, Florida
Voice:  407-531-7584  -  Fax:  407-531-7686  -  Cell:  407-718-4718
Pager:  -  EMail:

" Some days, it is best to keep reality at arms length..."

		-----Original Message-----
		From:	Michael French []
		Sent:	Monday, January 24, 2000 10:03 AM
		To:	''
		Subject:	RE: xBug -Reply

		> -----Original Message-----
		> From: Andy Lewis []
		> Sent: Friday, January 21, 2000 3:44 PM
		> To: ''
		> Subject: RE: xBug -Reply
		> one-time querystring-based URLs). I used to pester Stefano
		> about the cache system from time to time but haven't had
		> time lately. Sounds like a good problem for y'all to
		> some resources towards fixing ;).
		> Is there any general interest in making the cache system
		> more robust?

		If the cache system is going to be reexamined, then I
		we allow binary output pages (e.g. jpg). One way to do this
		be to have two subclasses of Pages (character and binary),
		similarly a binary formatter interface that has handle()
		which takes an output stream, rather than a character
		An example application would be a server-side SVG renderer.

		I could help with this (when I finally get around to moving
to 1.6!).


View raw message