Login | Register For Free | Help
Search for: (Advanced)

Mailing List Archive: Apache: Dev

UXBITHACK/GXBITHACK Patches

 

 

Apache dev RSS feed   Index | Next | Previous | View Threaded


andrew at www

Mar 13, 1995, 11:16 AM

Post #1 of 4 (117 views)
Permalink
UXBITHACK/GXBITHACK Patches

Hi,

I've upped the patches for the following to /incoming:

B5) XBITHACK not honored on <!--#include-->ed files [Andrew Wilson]

o patch.UXBITHACK

modifies http_include.c

E7) Send Last-modified header for server-side-included docs if
group XBIT is set. [Rob Hartill, by way of Andrew Wilson]

o patch.GXBITHACK

modifies http_get.c

Each patch has it's own compile-time directive, so you can switch it off
for testing purposes.

Apologies for the delay, I'm trying to break the habit of a lifetime and
actually do non-internet things on the weekends for a change ;)

Ay.


hartill at ooo

Mar 13, 1995, 11:37 AM

Post #2 of 4 (112 views)
Permalink
Re: UXBITHACK/GXBITHACK Patches [In reply to]

>
> >E7) Send Last-modified header for server-side-included docs if
> > group XBIT is set. [Rob Hartill, by way of Andrew Wilson]
>
> Would this not be far more neatly implemented by a new shtml directive?

It's a 2 line hack to solve a problem with uncachable html using
server-side includes. As a 2 line hack it works well.

Andy Wilson was looking into the problems of solving this issue
properly - to do so, would require a lot of coding - we'd need
to look into each type of include, and decide how it should affect
the http headers, e.g. the Last-modified date should be the date
of the newest include. So, to respond to a "get-if-modified-since"
the server would have to create the document, check the date, then
decide whether to send the document or a 304 response instead. This
is something we need to look into later - a lot later.


rob h


rst at ai

Mar 13, 1995, 1:49 PM

Post #3 of 4 (112 views)
Permalink
Re: UXBITHACK/GXBITHACK Patches [In reply to]

Date: Mon, 13 Mar 95 18:29 GMT
From: drtr [at] ast (David Robinson)

>E7) Send Last-modified header for server-side-included docs if
> group XBIT is set. [Rob Hartill, by way of Andrew Wilson]

Would this not be far more neatly implemented by a new shtml directive?

David.

Actually, there already is something like this in the spec ---
<META http-equiv="last-modified">, although for maximum convenience
the server would have to be hacked to plug in the actual last-modified
date if no other data were supplied.

The real problem with implementing this is that it requires nontrivial
surgery on the http_includes code --- the server needs to do a prepass
over the document which covers at least the <HEAD> (and, for the sake
of efficiency, to be smart enough to figure out where that ends
without explicit </HEAD> tags) before printing the MIME headers.

After that, assuming that the client isn't just being told to
USE_LOCAL_COPY, the server has to go back and do a second pass in
which it actually scans the document (shipping most of it back to the
client, and interpreting the remaining directives).

This is obviously the right thing, but it's also a nontrivial amount
of work, so it'll have to be deferred until someone is motiviated
enough to roll up their sleeves and code it.

rst


drtr at ast

Mar 13, 1995, 6:29 PM

Post #4 of 4 (112 views)
Permalink
Re: UXBITHACK/GXBITHACK Patches [In reply to]

>E7) Send Last-modified header for server-side-included docs if
> group XBIT is set. [Rob Hartill, by way of Andrew Wilson]

Would this not be far more neatly implemented by a new shtml directive?

David.

Apache dev RSS feed   Index | Next | Previous | View Threaded
 
 


Interested in having your list archived? Contact Gossamer Threads
 
  Web Applications & Managed Hosting Powered by Gossamer Threads Inc.