Re: Antwort: Re: Antwort: [Mod_gzip] Vary: header and mod_gzip

On Mon, 26 Aug 2002, Henrik Nordstrom wrote:

> > An interesting question I have to both of you is Who should care about
> > known bugs in popular browsers? Should Squid take a part in this work?
>
> Depends on what the bug affects. If the bug affects hop-by-hop features of
> HTTP then Squid will need to take part. If the bug only affects end-to-end
> features such as content-encoding (i.e. mod_gzip) then it is technically none
> of Squid's business. HTTP is rather clear on the distinction between caching
> proxies, user-agents and origin servers when it comes to reply entities,
> their criterias and responsibilities, but it takes a while to crossread all
> the sections to get the picture... (you first need to get the end-to-end vs
> hop-by-hop picture, everything else related to caching/proxying depends on
> this)
>
> If mod_gzip starts to play with transfer-encoding then it will become Squids
> business once Squid acheives HTTP/1.1, but then the playfield is set very
> differently from the question on content-encoding. The two are fully separate
> from each other. Any transfer-encoding bugs in browsers will then be Squids
> problem, and mod_gzip would need to deal with transfer_encoding bugs in Squid
> (if any).

I understand your position Henrik. Just in case...
I'm just wondering how the chain of browser->Squid->httpd->Squid->browser
could interfere with some browser's bugs, finally affecting the delivered
content.

For example, let's say that MSIE is requesting some.js file using
HTTP/1.1 via Squid. It sends usual Accept-Encoding: gzip, and it really
can uncompress gzip over HTTP/1.1. But Squid downgrades HTTP/1.1 to
HTTP/1.0, and httpd sees HTTP/1.0 from MSIE accomplished with
Accept-Encoding: gzip. We know exactly that MSIE does NOT speak gzip over
HTTP/1.0. Will it be correct to respond gzip in this case? I have no clue.
I did never have a chance to test this particular situation, but it looks
confusing for implementation of content compression.
It's not clear, but just in case, let's imagine that I send back gzipped
response over HTTP/1.0, and MSIE is lucky to uncompress it correctly.
I'm accomplishing the response with Vary: Content-Encoding, and Squid caches
it successfully.
The next NN-4.X is coming to Squid to order the same file over HTTP/1.0
with it's buggy Accept-Encoding: gzip. But Squid does not care about that.
What the content will be delivered to that poor fellow on NN-4.X?

Thanks,
Slava
Received on Mon Aug 26 2002 - 17:09:58 MDT