576936 - skydrive.live.com Office Powerpoint html viewer uses mobile? layout for non-Firefox UAs

Status: RESOLVED INVALID

(Web Compatibility :: Site Reports, defect)

Description

[Carl Fletcher]

User-Agent:       Mozilla/5.0 (X11; U; Linux i686; en-US; rv:2.0b2pre) Gecko/20100630 SUSE/4.0b-1.1 Minefield/4.0b2pre
Build Identifier: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:2.0b2pre) Gecko/20100630 SUSE/4.0b-1.1 Minefield/4.0b2pre

Click an image
Layout is lost

Reproducible: Always

Comment 1

[Kevin Brosnan]

Spoofing Firefox 3.6 causes the site to render as expected. This is a live.com issue, please contact the live.com support and ask them to check for Gecko instead of Firefox. http://geckoisgecko.org/

Comment 2

[Kevin Brosnan]

According to bug 571755 live.com picks up the pre in the user agent string and confuses it with the Palm Pre. Thus giving the page the mobile css.

Comment 3

[Kevin Brosnan]

Fixed by placing Firefox in the trunk user agent, bug 588874.

Comment 4

[Dão Gottwald [:dao]]

(In reply to comment #3)
> Fixed by placing Firefox in the trunk user agent, bug 588874.

Firefox is not the only Gecko browser.

Comment 5

[Dão Gottwald [:dao]]

(If you think this is a duplicate of bug 571755, you should mark it accordingly.)

Comment 6

[Kevin Brosnan]

This was reported because the user agent was Minefield. If there are issues with release versions SeaMonkey, Camino...I would consider that to be a separate bug.

Comment 7

[Chris Lawson (gone)]

(In reply to comment #6)
> This was reported because the user agent was Minefield.

Well, it's not entirely clear what, exactly, the problem was. It may have been that the original UA string contained "pre", or it may have been that the original UA string did not include "Firefox", or both. Either way, it's certainly not something that is unique to Firefox nightly builds that ID as "Minefield".

> If there are issues
> with release versions SeaMonkey, Camino...I would consider that to be a
> separate bug.

That's not how we've generally handled TE issues in the past, since the whole point of getting authors to fix their sniffing is that they a) probably shouldn't be sniffing in the first place, but if they must, they should b) be sniffing by rendering engine, not user-agent name. One bug per RE (or, put another way, one bug for all Gecko browsers) seems logical. It doesn't make much sense to tell people they should ignore UA names and look for RE versions, and then have bugs that differ only in the UA name they're intended to evangelise.

Comment 8

[Karl Dubost💡 :karlcow]

Moving to Desktop and closing as invalid.
We try to contact sites for getting things right for most people, not necessary for all kind of weird UAs such as the Minefield one, which is likely to increase the UA footprint instead of reducing it. :)