- From: Dean Jackson <dino@apple.com>
- Date: Thu, 13 Dec 2012 11:05:15 +1100
- To: Sylvain Galineau <sylvaing@microsoft.com>
- Cc: fantasai <fantasai.lists@inkedblade.net>, "www-style@w3.org" <www-style@w3.org>
On 13/12/2012, at 8:03 AM, Sylvain Galineau <sylvaing@microsoft.com> wrote: > [fantasai:] >> >> I don't see that changing the spec here is particularly useful or >> necessary. >> Just put >> >> html { >> widows: 1; >> orphans: 1; >> } >> >> in the UA style sheet. >> > > Er, not breaking existing content is quite useful and necessary :) > > We could say 'fix the UA stylesheet' every time the initial value of > a new property breaks existing content. Generally, we don't. I do not > think the purpose of UA stylesheets is, or should be, to work around > initial values that conflict with existing content, be it on screen > or paper. The default behavior should be compatible when possible > and reasonable. Any specific reasons why it's too onerous here? Some more thoughts on using UA stylesheet: - I'm not sure why having the UA stylesheet force a value that disagrees with the specification is any better than having the implementation do it. As far as the developers are concerned, both are non-standard. - UA stylesheets are limited to things that can be selected by CSS. This is purely theoretical, but imagine I need to implement a different "auto" behaviour on something that I can't write a selector for. - There is the potential to conflict with author styles, because my UA stylesheet might be more specific (see above about wanting different behaviours). I think I may have said this on Twitter (the greatest place for technical arguments), when we change existing content we get complaints from people. It doesn't matter if we think we've improved their content, as should be the case for these properties. The good news is that with 'auto', nothing gets worse. And anyone who has specified 'widows' or 'orphans' gets exactly what they asked for. If this discussion was actually on Twitter, I would have been so limited by space that this email would have just been "I agree with Sylvain". Dean
Received on Thursday, 13 December 2012 00:05:57 UTC