[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Re[2]: [Q] GSWeb Cache Again
From: |
Sungjin Chun |
Subject: |
Re: Re[2]: [Q] GSWeb Cache Again |
Date: |
Mon, 27 Nov 2000 22:30:53 +0900 |
So, current state of GSWeb performs caching but cached data
is not used for efficient, faster response. And this could be
"corrected". Do I understand rightly?
Architecture of GSWeb seems very good for me and it
can solve my problem of mixed view and model, controller
problem of script language such as PHP, JSP and etc.
Only problem is performance and I hope that I can improve
the performance of GSWeb and be happy. :-)
----- Original Message -----
From: "Manuel Guesdon" <ml@sbuilders.com>
To: "Sungjin Chun" <ninja@linuxone.co.kr>
Cc: <discuss@gnustep.org>; <discuss@gnustep.org>
Sent: Monday, November 27, 2000 10:25 PM
Subject: Re[2]: [Q] GSWeb Cache Again
> On Mon, 27 Nov 2000 21:26:50 +0900 Sungjin Chun <Sungjin Chun
<ninja@linuxone.co.kr>> wrote:
>
> >| GSWeb makes html from template and as far as I know this
> >| conversion occurs for every request even the results are
> >| the same as old ones. And I think caching these results and
> >| return without conversion makes faster response.
> >| Am I wrong?
>
> You're right.
> GSWeb cache paring result (you can see a debug indication of
> this:
> Component THENAME language XXX (Cached) or Component THENAME
> language XXX (Not Cached)
>
> Each page result is stocked in a GSWResponse object. I think
> these GSWResponse object are stored somewhere, probably in
> GSWSession. As far as I remember, contexts are stocked with
> their response. Response signature may be used for this.
> There's obscure parts in GSWSession: I think contextRecords and
> contextArrayStack are not very well implemented.
> To re-send a cached response, you have to be sure that
> the browser want it with same parameters (for exemple the visitor
> didn't changed any field value in a form before re-submiting
> it).
>
> Hope this helps.
>
> Manuel
>
>
>
>
>
>