You are viewing a plain text version of this content. The canonical link for it is here.
Posted to adffaces-user@incubator.apache.org by Renzo Tomaselli <re...@tecnotp.it> on 2006/12/30 11:12:56 UTC

[Trinidad] css generation and rule ordering

Hi, I noticed that Trinidad CSS generation occurs through some kind of
optimization, which groups together CSS rules having exactly the same
contents. This mixes up component rules and custom rules, thus scrambling
their original postions.  
Since CSS rule matching is highly position-dependent, the above mixup yields
unpleasant side effects. E.i. html elements with multiple classes get rendered
by the last class attribute found in the CSS file. If position changes,
rendering can change.  
I worked around such issue by adding dummy rule elements, so that comparison
for merging fails and this works fine, since position is preserved.  
Is there any cleaner way to prevent this scrambling or can we even disable it
at all ?  
Thanks -- Renzo  
  


Re: [Trinidad] css generation and rule ordering

Posted by Adam Winer <aw...@gmail.com>.
No, there isn't a way to prevent the scrambling.  I generally
make a practice of avoiding order-dependent rules.  IIRC,
ordering isn't important if you have varying specificities;
and if you're relying on ordering, your CSS may be getting
just a touch overwhelming...

-- Adam



On 12/30/06, Renzo Tomaselli <re...@tecnotp.it> wrote:
>
>  Hi, I noticed that Trinidad CSS generation occurs through some kind of
> optimization, which groups together CSS rules having exactly the same
> contents. This mixes up component rules and custom rules, thus scrambling
> their original postions.
>  Since CSS rule matching is highly position-dependent, the above mixup
> yields unpleasant side effects. E.i. html elements with multiple classes get
> rendered by the last class attribute found in the CSS file. If position
> changes, rendering can change.
>  I worked around such issue by adding dummy rule elements, so that
> comparison for merging fails and this works fine, since position is
> preserved.
>  Is there any cleaner way to prevent this scrambling or can we even disable
> it at all ?
>  Thanks -- Renzo
>
>