mercurial/crew at 7727: 18 outgoing changesets

[Mads Kiilerich]

Martin Geisler wrote:
> Mercurial Commits <hg at intevation.org> writes:
>   
>> http://hg.intevation.org/mercurial/crew/rev/6a888d491eaf
>> changeset:   7725:6a888d491eaf
>> user:        Benoit Allard <benoit at aeteurope.nl>
>> date:        Thu Feb 05 18:15:33 2009 +0100
>> summary:     coal/paper: switch to XHTML 1.1
>>     
>
> Are we sure we want this?
>
> I'm not a big webdesigner, but I have tried to follow along from the
> sideline, and it is my understanding that many sites chose to generate
> strict HTML 4 since this is the best understood variant.
>
> One problem with XHTML is what content-type to use, the web tells me
> that the "correct" type breaks IE.
>
> With the current template system it should be just as easy to generate
> correct HTML 4 as correct XHTML -- the individual files in the
> template wont validate anyway.
>   

That change "almost" fixes 
http://www.selenic.com/mercurial/bts/issue1464 (IE fails to render 
white-space: pre in quirks mode). Testing shows that it works better 
than before on most pages. But IE 7 rendering still fails on pages with 
these crazy characters - see for example /rev/fff6e253e1f6 .

So you might be right that strict HTML 4 would be better. But perhaps a 
filter for HTML &-encoding of all unicode characters is needed in any 
case. But that requires knowledge of which encoding is used...

/Mads