W3C home > Mailing lists > Public > public-i18n-bidi@w3.org > July to September 2012

RE: [cssom-view] value of scrollLeft in RTL situations is completely busted across browsers

From: Sylvain Galineau <sylvaing@microsoft.com>
Date: Mon, 20 Aug 2012 17:47:57 +0000
To: "robert@ocallahan.org" <robert@ocallahan.org>, Ojan Vafai <ojan@chromium.org>
CC: "www-style@gtalbot.org" <www-style@gtalbot.org>, www-style <www-style@w3.org>, Shane Stephens <shans@google.com>, ML publc-i18n-bidi <public-i18n-bidi@w3.org>
Message-ID: <3C4041FF83E1E04A986B6DC50F0178291BE2C336@TK5EX14MBXC227.redmond.corp.microsoft.com>

>For scrollLeft, Sylvain wants IE's behavior, 
>Glenn and Shane want the Firefox/Opera/Webkit(viewport) behavior (matching pageXOffset). 

Shane thought IE9 had it right but was OK with your bogus-data-based proposal so I don't
think it's so clear cut :)

Given the divergence in implementations it seems we can change this one way or the other.
However, I'd rather not make this kind of change in a piecemeal fashion. I would rather
see a proposal describing what RTL does to CSSOM-View properties overall. It may be that 
these properties are all physical and are unaffected, or they're logical, or we'll add an 
extra set of logical properties and make the current ones physical...Whatever the solution
it just seems an area where we should make one set of carefully considered changes across
all our implementations.

I don't mean to slow down progress on resolving this particular issue; but I've had some
spectacularly bad times messing with this Jenga tower of properties in the not so distant

Received on Monday, 20 August 2012 17:48:33 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:24:40 UTC