- From: Martin J. Dürst <duerst@it.aoyama.ac.jp>
- Date: Wed, 18 Jul 2012 10:41:27 +0900
- To: James M Snell <jasnell@gmail.com>
- CC: Karl Dubost <karld@opera.com>, Julian Reschke <julian.reschke@gmx.de>, HTTP Working Group <ietf-http-wg@w3.org>
On 2012/07/18 6:08, James M Snell wrote: > +1... I was getting ready to respond to Julian's post with a "+1 YEAH!" but > then stopped and thought about it... using a URI, while potentially sound > in theory, would likely just end up with a different form of the same mess > we're in now. Reducing the User-Agent field to nothing more than a single > token with a version identifier seems to me to be the Least Bad Option. So why did that not happen up to now, and why is the User-Agent string growing as it does? Because servers/proxies check on substrings in that string, so UAs are using the strategy to copy the whole header of some other browser (usually the most popular one around at that time) and then tack some additional stuff on at the back to make sure they can be uniquely identified. This is a cat-and-mouse game where the servers are as much (or as little) to blame as the browsers. As long as each side does what's most convenient for them, changing the name of the header, or offloading the data to a separate URI, or any such thing, is only a cosmetic change. Regards, Martin.
Received on Wednesday, 18 July 2012 01:42:04 UTC