- From: Amos Jeffries <squid3@treenet.co.nz>
- Date: Sat, 15 Nov 2014 22:46:50 +1300
- To: ietf-http-wg@w3.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 15/11/2014 9:17 p.m., Martin Nilsson wrote: > On Fri, 14 Nov 2014 23:54:28 +0100, Daniel Stenberg wrote: > >> >> While an interesting exercise, do you have any HTTP >> client/browser authors backing up this idea and syntax? >> >> This seems like a futile effort to me with little hope of >> adaption. I'm the author of a HTTP client using the simplest >> possible User-agent but I'm also fully aware of how often users >> have to mimic popular browsers' user-agent strings to make >> servers respond appropriately. >> >> There's a lot of pain and many side-effects in doing even minor >> modifications of the user-agent string. >> > > On the apps-discuss list it was stated that all major browsers > already conformed to this specification, which makes it easy to > conform to, but also a bit pointless. > > If we could get agreement on a single user-agent string and have > all browsers move to that one at the same time, that would be a > standard I would be enthusiastic to endorse. An agreement in a > secret back room would do just as fine as a standard I suppose... It would be even better if browsers just started to conform to RFC 7231 section 5.5.3. There is no conspired agreement or timing required on their part to do so. * "MUST NOT generate advertising or other nonessential information within the product identifier" no dumping comments about OS version, unrelated software installed, other browsers installed!, patch levels, etc. "Mozilla/" is conformant but realistically unnecessary these days as it can readily be inferred from the browser Product/x.y identifier instead. * "SHOULD limit generated product identifiers to what is necessary to identify the product" no dumping "like Gecko", "like Konquerer", "like Firefox" , "like Chrome" into the UA string. The draft-karcz-uuas as written does nothing to advance improvement of U-A usage over RFC 7231, and in fact the browser-string section goes so far as to recommend violations of the RFC 7231 requirements. PS. Apparently the browser "like" is winning the browser war for market share. No matter how much advertizing or downloads the other so-called mainstream browsers put out this product just keeps coming out on top. ;-P Amos -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (MingW32) iQEcBAEBAgAGBQJUZyEJAAoJELJo5wb/XPRjp0gH+wa6VbvoTVwrXf5uusbjzxLy w1JmHqzFS1Y2mCNuPy8skHxv5H/piqRHA1nQ1jM6p/naq2NZ2rx3vmeLQYn5Ynrc OpQm/WgbvUb7drSM4Of77X/Nf+rTtw+xDIjC67isCzpm4pr7ZEB82pxrXqmYq6KC P7msLkErGjp/EMNTLqyC5E2OYL6ZgA2NCYeOQGWNrfAqMhDSyCYjW87jxfTb9x+I cb6hulPleGa39xxbjDSlBaimE6Qng1oOsCY8ieInXA5bKcphV+RuHazAxOV39VT9 zg3A/FS8bXDJAsBd2vewJp1W0G1xrso0KQJHti9izfEZw3HvMb+9afkBMeZSVoM= =RKyg -----END PGP SIGNATURE-----
Received on Saturday, 15 November 2014 09:47:32 UTC