“squid actually changed their default policy with caching dynamic URLs with their 2.7 release:
http://wiki.squid-cache.org/ConfigExamples/DynamicContent

The issue with Squid is that it is almost 1.1, but not completely. From the top of my head the biggie is chunked encoding. Hopefully it will become 1.1 one day. But the code seems to be very hard to change.

0 – http://www.aol.com/
0 – http://www.ebay.com/
13 – # http://www.facebook.com/
0 – http://www.google.com/search?q=flowers
4 – http://search.live.com/results.aspx?q=flowers
1 – http://www.msn.com/
1 – http://www.myspace.com/
12 – http://en.wikipedia.org/wiki/Flowers
0 – http://www.yahoo.com/
1 – http://www.youtube.com/

Facebook has 118 resources (!), so only having 13 of them suffer from a querystring isn’t bad, but URLs like “facebook_logo.gif?0:67387″ could be fixed. For Wikipedia, 12 out of 20 cacheable resources contain a querystring (e.g., “ajax.js?169″).

Many of these sites have clearly worked hard to avoid querystrings, with URLs like “js_3.011.js” and “/4_7_0_227490/main.js”. That’s great, but there’s still more cleanup to do. A comment from my blog post pointed to a technique for versioning filenames posted by Kevin Hale that gives the specifics about how to tackle this in Apache with mod_rewrite.

On possible way around this is to put the version into the path and use mod_rewrite or a similar technique to remove it.