[Zlib-devel] zlib 1.2.3.1 released for testing

[Greg Roelofs]

>> There's no particularly good reason why Firefox couldn't use zlib  
>> 1.1.x, right?

> Only speed.  inflate is used to decompress HTTP content encodings  
> gzip and deflate, which are pretty common.  So the speed may make a  
> difference.

I think we both know that compression speed isn't the bottleneck these
days.  Home computers, even 5-year-old ones, are ridiculously overpowered,
and the stuff that gets gzip- or deflate-encoded isn't the bulk of what
gets downloaded, anyway--already-compressed binary formats are.

No, the reasons for poor web-browsing performance are primarily in the
network (bandwidth latency, DNS latency, multi-request latency) and in
the page-design/browser-architecture combo (failure to render until all
CSS finishes downloading, 2-request limit for simultaneous downloads
from the same site, unnecessary redirects, poorly compressed images,
overweight Flash, swapping due to memory leaks, you name it...).  Even
doubled performance in inflate()--and I don't believe it was anything
like that, was it?--would not make a noticeable difference in the
rendering speed of most pages.

> 1.1.x would have to be 1.1.4 to avoid a security vulnerability in  
> 1.1.3 and earlier.

Yup.  I couldn't remember if 1.1.4 had any known issues of its own,
though.

Greg