Well, as far as mod_webkit is concerned, if apache has access to the header,
we can get it.  Just gotts figure out how.

What does the HTTP 1.1 spec say about this?

Jay

> -----Original Message-----
> From: Chuck Esterbrook [mailto:[EMAIL PROTECTED]]
> Sent: Thursday, March 14, 2002 10:18 PM
> To: [EMAIL PROTECTED]
> Subject: Re: [Webware-discuss] HEAD?
> 
> 
> On Thursday 14 March 2002 04:16 pm, Chuck Esterbrook wrote:
> > On Wednesday 13 March 2002 10:43 am, Chuck Esterbrook wrote:
> > > I just looked at my apache access_log and all the requests for the
> > > style sheet are GET and POST. No HEAD:
> > >
> > > 68.5.28.172 - - [13/Mar/2002:13:23:12 -0500] "GET /StyleSheet.css
> > > HTTP/1.1" 200 10928
> > >
> > >
> > > I was expecting HEAD for StyleSheet.css since it is just a static
> > > file being served by UnknownFileTypeServlet.
> > >
> > > A grep for HEAD reveals no more HEAD requests after March 5th. And
> > > the few HEADs that occurred were for anything but the StyleSheet.
> > > This adds 10K per request and will slow down my app server.
> >
> > Okay, the deal is that the GET from modern browsers is 
> accompanied by
> > an If-Modified-Since header (so they don't bother with HEAD). I
> > checked this out using Mozilla, Apache and a static .html 
> file served
> > directly. The requests are always GET, but the subsequent responses
> > return 0 bytes of document.
> >
> > Having dumped the attrs of HTTPRequest, I don't see that it has
> > access to that information, or even the request headers in 
> general. I
> > also don't find any source code in webware referencing 
> "modified" and
> > "since" in the same line.
> >
> > I presume that we can somehow get this out of mod_webkit, 
> but I don't
> > know about CGI.
> >
> > I'll investigate further tonight.
> 
> Bottom line: Most browsers use a "conditional GET" such as a GET with 
> an If-Modified-Since: header as opposed to the old "HEAD and GET" 
> technique. However on Apache, the If-Modified-Since HTTP header does 
> not get passed to WebKit (via mod_webkit or CGI), so for static files 
> such as a style sheet, WebKit sends the full content every time.
> 
> One workaround would be to put your static files outside your webkit 
> location. For my special case of a style sheet, I may just may use a 
> symbolic link that points to its current location.
> 
> 
> Here is some of the research I did (mostly via Google):
> 
> - I found that some HTTP servers will offer an HTTP_IF_MODIFIED_SINCE 
> to their CGI scripts, but Apache does not.
> 
> - Searching for HTTP_IF_MODIFIED_SINCE hit a lot of PHP sites 
> that were 
> dumping globals, but when I visited the pages (and reloaded) 
> HTTP_IF_MODIFIED_SINCE was missing. Perhaps it was available in an 
> earlier version of either Apache or PHP.
> 
> - The only examples of being able to hook into If-Modified-Since were 
> things (such as mod_perl) that lived in Apache. But the hook I saw 
> wasn't to get the value, but call a function with higher level 
> semantics.
> 
> - It might be possible to tweak Apache to include 
> HTTP_IF_MODIFIED_SINCE, but I'm suspicious that it's not already done 
> or at least available as a patch. There might be some special 
> treatment 
> of this header by Apache that make it infeasible.
> 
> - Apache 2.0?
> 
> 
> 
> -Chuck
> 
> _______________________________________________
> Webware-discuss mailing list
> [EMAIL PROTECTED]
> https://lists.sourceforge.net/lists/listinfo/webware-discuss
> 


----------------------------------------------------------------------------

This e-mail and any attachments may be confidential or legally privileged.
If you received this message in error or are not the intended recipient, you
should destroy the e-mail message and any attachments or copies, and you are
prohibited from retaining, distributing, disclosing or using any information
contained herein.  Please inform us of the erroneous delivery by return
e-mail. 

Thank you for your cooperation.

----------------------------------------------------------------------------



_______________________________________________
Webware-discuss mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/webware-discuss

Reply via email to