On Fri, Mar 25, 2011 at 1:03 PM, Mark Rowe <mr...@apple.com> wrote: > On 2011-03-25, at 12:56, Peter Kasting wrote: > > On Fri, Mar 25, 2011 at 12:54 PM, Mark Rowe <mr...@apple.com> wrote: > >> Is there some reason why these examples use manufactured Safari build >> numbers? It's implausible that a version of Safari with a build number of >> 534.24 would ever claim to be version 5.0.3. >> > > Sorry, I wasn't sure what the right numbers were. What would be a more > accurate number? I'd be happy to change it. > > > I'm not sure what you're asking here. Safari 5.0.3 will always report a > build version of 533.19.4. Safari 5.0.4 will always report a build version > of 533.20.27. You already used the former in your "before examples". You > have also been inconsistent about the WebKit version number. In the > "before" version for the Mac user agent string you've included the "+" > suffix that indicates an engineering build (from a local build or nightly). > This isn't present in the Windows "before" version or either of the "after" > versions. Since that component isn't relevant to the point you're trying to > make it seems like it would be preferable for it to be consistent between > examples. >
Ideally, I would like the Safari-on-Windows and Safari-on-Mac strings from a trunk build as of around the time my change landed, which I'd use as the identical before and after strings, and only change the portions actually affected by the UA changes that landed. I don't have the ability to build either of those myself right now, so what you're seeing is a combination of internet research, inference from Chromium builds, and mistakes in copy and pasting. PK
_______________________________________________ webkit-dev mailing list webkit-dev@lists.webkit.org http://lists.webkit.org/mailman/listinfo.cgi/webkit-dev
