[squid-users] clientProcessHit

[joe]

>Ah, you said you understood Vary. But this mail indicates that you do
>not quite understand it well enough.
i do understand may be the way sometime i try to explain it goes buggy lol
like those bug any way

>I prune away the extra headers to clarify:

>>Firefox:

>> GET /en_US/embeds.js HTTP/1.1
>> Accept-Encoding: gzip, deflate
 
>> HTTP/1.1 200 OK
>> Vary: Accept-Encoding
>> ----------------------------------------------------------

>Chrome:

>> GET /en_US/embeds.js HTTP/1.1
>> Accept-Encoding: gzip, deflate, sdch

>... notice how Accept-Encoding changed.

>> HTTP/1.1 200 OK
>> Vary: Accept-Encoding


>The Vary header tells Squid that these objects are to be stored with the
>*content* of the requests Accept-Encoding header added to their URL on
>the cache index.

>The Accept-Encoding details are effectively being added to the URL
>query-string without actually being added to the public URL users see.
yup
>Since the Accept-Encoding headers are different on the Chrome requests
>these objects are as different in cache.
yup
>Firefox shares Accept-Encoding value with IE, and with some command-line
>tools. So they have no problems caching for one and re-using the cached
>object for another.
yup
>But Chome do the Google sdch thing, so it cant use those objects and has
>to have a separate object cached just for Google products.

yup but event if i force gzip only as
# Normalize Accept-Encoding to support compression for better caching 
request_header_access Accept-Encoding deny all 
request_header_replace Accept-Encoding gzip

chrome still get miss so its forcing the use of Google sdch or what im
missing 

so what will happen. dose chrome use this instead of  gzip
since Vary: Accept-Encoding   ar same reply 
and squid request only  gzip
and all reply Vary: Accept-Encoding is it bad ?? to do so ?? for other app
in mobile




--
View this message in context: http://squid-web-proxy-cache.1019090.n4.nabble.com/clientProcessHit-tp4676484p4676587.html
Sent from the Squid - Users mailing list archive at Nabble.com.