Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-29 Thread David Cassidy
ie6 has many "features" maybe this is just another one On Fri, 2008-02-29 at 21:03 +0800, J. Peng wrote: > Thanks for all replies. > I thought then the primary problem is IE6 with http/1.0 can't accept > the Vary: header. > I don't know why, but this really happened on all our ie6 browsers.

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-29 Thread J. Peng
Thanks for all replies. I thought then the primary problem is IE6 with http/1.0 can't accept the Vary: header. I don't know why, but this really happened on all our ie6 browsers. Our users have also reported this case. Maybe it's an IE's bug? ie7/firefox don't have this problem. Thanks. //joy O

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-29 Thread Eric Covener
On Fri, Feb 29, 2008 at 4:24 AM, J. Peng <[EMAIL PROTECTED]> wrote: > On Fri, Feb 29, 2008 at 5:14 PM, Matus UHLAR - fantomas > <[EMAIL PROTECTED]> wrote: > > On 29.02.08 11:44, J. Peng wrote: > > > I talk nothing about cache. > > > if mod_deflate didn't compress a file, why it still send a V

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-29 Thread David Cassidy
mod deflate from my understanding will always send the Vary header. it does it because it *might* do something to the content on subsequent requests. Whilst for you proxies are not important it doesn't mean it isn't important for everyone else - ditto using the standards :) I guess your applicati

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-29 Thread J. Peng
On Fri, Feb 29, 2008 at 5:14 PM, Matus UHLAR - fantomas <[EMAIL PROTECTED]> wrote: > On 29.02.08 11:44, J. Peng wrote: > > I talk nothing about cache. > > if mod_deflate didn't compress a file, why it still send a Vary header? > > This is what I'm always talking about. > > you were advised to r

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-29 Thread Matus UHLAR - fantomas
On 29.02.08 11:44, J. Peng wrote: > I talk nothing about cache. > if mod_deflate didn't compress a file, why it still send a Vary header? > This is what I'm always talking about. you were advised to read HTTP specificationm (rfc 2616), did you? -- Matus UHLAR - fantomas, [EMAIL PROTECTED] ; http

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-28 Thread J. Peng
I talk nothing about cache. if mod_deflate didn't compress a file, why it still send a Vary header? This is what I'm always talking about. On Fri, Feb 29, 2008 at 11:25 AM, Joshua Slive <[EMAIL PROTECTED]> wrote: > On Thu, Feb 28, 2008 at 8:46 PM, J. Peng <[EMAIL PROTECTED]> wrote: > > I mean I a

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-28 Thread Joshua Slive
On Thu, Feb 28, 2008 at 8:46 PM, J. Peng <[EMAIL PROTECTED]> wrote: > I mean I access to apache directly, didn't go through any cache/proxy. > ie6 with http/1.0 has problems with a Vary: header, if the response > was not gziped. > What I think is if mod_deflate DIDN'T compress a file, it SHOULD

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-28 Thread J. Peng
I mean I access to apache directly, didn't go through any cache/proxy. ie6 with http/1.0 has problems with a Vary: header, if the response was not gziped. What I think is if mod_deflate DIDN'T compress a file, it SHOULD NOT send a Vary header. Is it really hard for the author? Not at all. I wrote a

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-28 Thread Joshua Slive
On Thu, Feb 28, 2008 at 9:57 AM, Nils Jeppe <[EMAIL PROTECTED]> wrote: > > > On Thu, 28 Feb 2008, Joshua Slive wrote: > > > For example, why are you forcing MSIE to use HTTP/1.0 requests in the > > first place? > > MSIE defaults to http 1.0 when using a proxy. This can be changed in its > pref

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-28 Thread Nils Jeppe
On Thu, 28 Feb 2008, Joshua Slive wrote: For example, why are you forcing MSIE to use HTTP/1.0 requests in the first place? MSIE defaults to http 1.0 when using a proxy. This can be changed in its preferences, but of course this requires manual action on the user's part. Best wishes, -

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-28 Thread Joshua Slive
On Thu, Feb 28, 2008 at 1:28 AM, J. Peng <[EMAIL PROTECTED]> wrote: > I just think Apache's mod_deflate has a weak point. > When client doesn't declare it accept encoding, apache returns an > uncompressed object, but for this object mod_deflate also returns a > "Vary: Accept-Encoding" header. >

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-28 Thread J. Peng
On Thu, Feb 28, 2008 at 4:26 PM, Nils Jeppe <[EMAIL PROTECTED]> wrote: > > Try adding: > > Header append Vary User-Agent > no use. at first time I added this statement, but not usefull. mod_deflate handler always add a Vary: accept-encoding in its outgoing response. btw, I have wrote a filter o

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-28 Thread Nils Jeppe
On Thu, 28 Feb 2008, J. Peng wrote: No.I didn't set up any proxy before Apache. I access to apache web service directly. Mh? You cc'ed the squid list, that's why I figured it might be a proxy issue. please, it's ie6sp2 with http/1.0 protocal. by default it's http/1.1 protocal. Yes and

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-28 Thread Nils Jeppe
Try adding: Header append Vary User-Agent Best wishes, - Nils On Thu, 28 Feb 2008, J. Peng wrote: ok this is the config in my httpd.conf: AddOutputFilterByType DEFLATE text/html text/plain text/xml text/css application/x-javascript SetOutputFilter DEFLATE BrowserMatch ^Mozilla/

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-27 Thread J. Peng
On Thu, Feb 28, 2008 at 3:22 PM, Nils Jeppe <[EMAIL PROTECTED]> wrote: > > > On Thu, 28 Feb 2008, Nils Jeppe wrote: > > > You do realize that, at least according to the documentation, you can > control > > that behaviour? I somehow doubt that this is a bug, as it would affect a > > whole lot o

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-27 Thread J. Peng
ok this is the config in my httpd.conf: AddOutputFilterByType DEFLATE text/html text/plain text/xml text/css application/x-javascript SetOutputFilter DEFLATE BrowserMatch ^Mozilla/4 gzip-only-text/html BrowserMatch ^Mozilla/4\.0[678] no-gzip BrowserMatch \bMSIE !no-gzip !gzi

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-27 Thread Nils Jeppe
On Thu, 28 Feb 2008, Nils Jeppe wrote: You do realize that, at least according to the documentation, you can control that behaviour? I somehow doubt that this is a bug, as it would affect a whole lot of people, but of course everything is possible. Please post the relevant part of your Apach

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-27 Thread Nils Jeppe
On Thu, 28 Feb 2008, J. Peng wrote: what I said is mod_deflate always send a "Vary: Accept-Encoding" header to clients though it didn't compress that object. this make some browsers have problems. You do realize that, at least according to the documentation, you can control that behaviour?

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-27 Thread J. Peng
what I said is mod_deflate always send a "Vary: Accept-Encoding" header to clients though it didn't compress that object. this make some browsers have problems. On Thu, Feb 28, 2008 at 2:49 PM, Nils Jeppe <[EMAIL PROTECTED]> wrote: > > > On Thu, 28 Feb 2008, J. Peng wrote: > > > I just think Apa

Re: [EMAIL PROTECTED] mod_deflate's problems

2008-02-27 Thread Nils Jeppe
On Thu, 28 Feb 2008, J. Peng wrote: I just think Apache's mod_deflate has a weak point. When client doesn't declare it accept encoding, apache returns an uncompressed object, but for this object mod_deflate also returns a "Vary: Accept-Encoding" header. This make some browsers be confused. In

[EMAIL PROTECTED] mod_deflate's problems

2008-02-27 Thread J. Peng
I just think Apache's mod_deflate has a weak point. When client doesn't declare it accept encoding, apache returns an uncompressed object, but for this object mod_deflate also returns a "Vary: Accept-Encoding" header. This make some browsers be confused. In fact my current IE6 sp2 can't download th