?

Log in

No account? Create an account
Lord Yupa

February 2010

S M T W T F S
 123456
78910111213
14151617181920
21222324252627
28      
Powered by LiveJournal.com
Danger Mouse

Mozilla oddity? [Updated]

I finally got around to setting up mod_gzip again (I broke it quite a while back, and just haven't gotten around to fixing it).

One thing I noticed, is that Mozilla doesn't always "work" correctly with it.

For example, when I try to get http://topher.zyp.org (a test page I use for playing with HTML, XHTML, CSS, etc), Mozilla somehow manages to incorrectly request the page. Here's the apache log:

tconl171104.tconl.com - - [14/Nov/2001:02:17:29 -0600] "GET / HTTP/1.1" 304 - "-" "Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:0.9.5) Gecko/20011011" mod_gzip: SEND_AS_IS:NO_200 In:0 Out:0 Ratio:0pct.

Here's what it should look like (taken from lynx and w3m):

topher.zyp.org - - [14/Nov/2001:02:16:03 -0600] "GET / HTTP/1.0" 200 852 "-" "Lynx/2.8.4rel.1 libwww-FM/2.14 SSL-MM/1.4.1 OpenSSL/0.9.6b" mod_gzip: OK In:1794 Out:852 Ratio:53pct.

topher.zyp.org - - [14/Nov/2001:02:20:21 -0600] "GET / HTTP/1.0" 200 852 "-" "w3m/0.2.1-inu-1.5" mod_gzip: OK In:1794 Out:852 Ratio:53pct.

So, why isn't Mozilla getting it? What's really odd, too, is if I manually gzip index.html (to index.html.gz) and request index.html (as opposed to "/"), then mozilla will properly get the gzipped file just fine.

And, I know mozilla can handle gzipped files (I've played with manually compressed files before, with it). Then, I remembered that in the Preferences, under Debug, there's a list of acceptable Encoding types. So, I glanced at that:

gzip, deflate, compress;q=0.9

Well, I'm confoozled. I can't think of any reason why mozilla wouldn't be getting/requesting the file correctly. Oh, and what's really strange, is that every once in a while Mozilla will get the file properly, compressed. And it works fine when I request non-static files with Mozilla:

tconl171104.tconl.com - - [14/Nov/2001:02:25:04 -0600] "GET /test/check.php HTTP/1.1" 200 396 "-" "Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:0.9.5) Gecko/20011011" mod_gzip: DECHUNK:OK In:607 Out:396 Ratio:35pct.

I love mod_gzip, I just wish things were a little less finicky with regards to it.

[Update]: I'm not so sure this is Mozilla, after all. In fact, I'm not pretty sure it isn't. I just did a little playing with IE, and one thing I noticed is that it seems to correctly get the page, compressed, the first time it gets it. After that, it doesn't. I'm wondering if this is due to cache reasons, and Mozilla/IE are checking to see whether the page has been updated since it was last viewed, and if not, then they're simply redisplaying the same page.

[Update II]: Okay, I'm almost positive my update theory is right. After touching the index.html file and then doing another page refresh, Mozilla/IE get the page correctly.

Chalk this one up to me not thinking things through entirely before posting about it. (Although, I did finally get my thoughts where they needed to be, and only a few minutes after my original post. ;-)

Comments

Yeah. . .

I figured it was the browser just checking with the webserver to ensure that the copy it had cached was still the most recent copy.

Thanks for the details, though. It's always good to have a theory verified, and learn the low level details that explain your high level theory. ;-)