RE: [css-writing-modes][CSS21] propagation of 'direction' from <body>

 On 03/12/2015 10:33 AM, Simon Pieters wrote:
> >
> > direction:rtl anywhere, 83,121 pages (~63.94%). This is *way* higher than I
> expected.
> 
> Yes, that's clearly absurd. RTL pages or even mixed-use ones aren't 64% of
> the Web. Maybe it's set in some library?
> 
> > I'm not sure what to make of this. I suppose the httparchive dataset
> > has a different bias than the data Greg uses. I can see a few options for
> next steps if we want to move this to HTML:

My query is over a randomized set of sites that are a max of 10 days old (the sample size varies but is normally a unique set of 5 - 10 million unique sites and all linked files [css, js, html on the same domain]); not sure how the web archive's data is searched.

> > * Someone implements new use-counters/telemetry/etc to gauge the
> impact.
> > * Someone tries to implement and ship dir special-casing in HTML and see
> what breaks.

Option two seems like the best option as it would answer our question faster than going through all the found sites and doing reductions to see if the change of where it propagates from would break the site. We could possibly have it turned on to default in the certain region we're expecting the issues for ~10% of users initially (I believe Blink has done this with other changes recently).

IE is not in a current state that we will be able to take on this change and flighting it, so it would be best if a Gecko/Blink based vendor could carry the test out. That said, it's understandable if this isn't possible for your teams either, just wanted to get it on record that we won't be able to do this for the foreseeable future.

Greg

Received on Friday, 13 March 2015 18:26:18 UTC