[Plone-UI] Re: Plone, restructured text, STX and CSS

Dave Turvene dturvene at comcast.net
Thu Mar 22 16:38:47 UTC 2007

Alexander Limi <limi at ...> writes:

> Yeah, I would be much more likely to use ReST if they cleaned up its  
> output somewhat, maybe you can help champion that?
> See http://article.gmane.org/gmane.comp.web.zope.plone.documentation/1498  
> for an example of how bad it gets.

Ouch.  Yes, it seems to tag tokens at an unnecessarily fine granularity.

> Other pet peeves are that they are supposed to live inside other systems  
> (like wikis, CMSes, etc), and thus should not use super-generic class  
> names like "document" — it would be better if they used a prefix on all of  
> them, like "rest-document".

Again, I see what you mean.  I don't suppose anyone else in the world uses CSS
classes named "field" or "reference".

> Firebug is the best CSS tool ever:

Nice, much easier than WevDeveloper.  Thanks!

> If ReST outputs a "docutils" class, you should probably make use of that,  
> but I don't really know the specifics here. :)

Here's where the really weird things begin.  authoring.css defines individual
border-top, border-bottom, etc. declarations for .stx table th, .stx table td
EXCEPT for the two missing table borders.  When I add the two missing border
sides to my skin, or simply use the "border:" declaration, everything shows. 
Why go to the trouble of writing six declaration statements for six of eight
sides, when two four-sided "border:" declarations do the trick?  

The second weird thing is these two sides can also be rendered using the ReST
.docutils class but none of the six originally displayed sides do this.  So I
have several ways to patch this (for my application), but it seems very much a
trial-and-error process.  It also seems like there a lot of ways to do the same
thing in CSS.  I think Larry Wall had a hand in it. 


More information about the UI mailing list