Re: [naviserver-devel] gzip compression fixes

[Daniel S. <moo...@av...>]

On Fri, May 23, 2008 at 1:54 PM, Stephen Deasey <sd...@gm...> wrote:
> What are the bugs that are fixed here? The message doesn't say.

When gzip is enabled for adp's, Internet Explorer 6 quietly rejects it
and renders nothing.  Firefox is forgiving of malformed gzip pages, IE
is not.   Curl also grumbles:  curl --compressed --trace - -i
mydomain.com

I had fixed this in AOLserver 4.0.10 cvs over 3 years ago and was
surprised it had crept back in.

When compress2() is done, it pads the destination buffer with 4 extra
bytes.  To be honest, I don't remember why it does this.  I did long
ago, but I don't now.  I do recall verifying it by comparing the
output of ns_gzip to the output of unix command line gzip.

The changes to the code rewind those 4 bytes and then appends the
footer.   Network byte ordering is not used.  The net result is an RFC
1952 complaint gzip file, which is what Internet Explorer 6 and curl
seem to demand.

Also, manually calculating the size destination buffer is deprecated.
compressBound() is advised.

Daniel
-- 
| ---------------------------------------------------------------
| Daniel P. Stasinski | http://www.saidsimple.com
| moo...@av... | http://www.disabilities-r-us.com
| XMMP: moo...@av... | http://www.avenues.org
| Google Talk: mooooooo | http://www.scriptkitties.com