On Fri, Mar 6, 2009 at 4:18 PM, Jsbeginner <jsbegin...@monarobase.net>wrote:
> Thankyou for the help, > > I will contact the script maintainer to ask him to work on this problem for > future updates, and for the moment I've commented out the line that > specifies the length. > I suppose that it won't cause any problems with files that are sent through > this script that are not gzipped (this script allows html, js, css (all > compressed with gzip) as well as images that aren't compressed. > > From what I understand the best would be to detect if gzip is activated for > each file and only allow the header to specify the length if the file is not > planned to be compressed with gzip (images etc). As we discussed, the declaration of size of the file is COMPLETELY unnecessary in this case. Lenin www.twitter.com/nine_L > > > Thanks again :) > > Nisse Engström a écrit : > > On Thu, 5 Mar 2009 19:14:20 +0100, Nisse Engström wrote: >> >> >> >>> On Thu, 05 Mar 2009 15:45:35 +0100, Jsbeginner wrote: >>> >>> >>> >>>> <?php >>>> header('Content-Type: application/x-javascript'); >>>> header('Content-Length: '.filesize('test.js')); >>>> readfile('test.js'); >>>> ?> >>>> >>>> test.js is only a few lines long, and if I remove the header content >>>> type the file loads instantaniously do it's not a problem with readfile. >>>> I thought about zlib gzip taking maybe a long time to load but I've >>>> changed the compression level from 6 to 1 and the file still has a the same >>>> lag. >>>> >>>> >>> Content-Length MUST NOT be sent when using a >>> Transfer-Encoding (eg. gzip). >>> >>> See: <http://tools.ietf.org/html/rfc2616#section-4.4> >>> >>> >> >> And I goofed: Transfer-Encoding is the stuff that says >> "chunked". I was thinking of Content-Encoding. But the >> section I refered to also says: >> >> "If a Content-Length header field (section 14.13) is present, its >> decimal value in OCTETs represents both the entity-length and the >> transfer-length. The Content-Length header field MUST NOT be sent >> if these two lengths are different" >> >> [And a little further down, it mentions "that the >> recipient can arse it". I'm not sure quite what >> to make of that. :-)] >> >> - - - >> >> Come to think about it, I've written some output handlers >> to convert utf-8 to utf-16 or -32, which uses either >> 'Content-Length' or 'Transfer-Encoding: chunked' depending >> on the size of the output. Perhaps I should change that to >> always 'chunked'... >> >> >> /Nisse >> >> >> > >