The author of the framework, John Slegers, recently posted on GitHub
that he is actively working on a new SASS version of the framework,
so if that would change your view of it then you might want to take
a look at it again in a couple months (I'm not sure of his time
frame).
My understanding is that the essence of OOCSS is reusable components
(objects) that can be subclassed as needed, favoring multiple
classes in the HTML over long selectors in the CSS. All of the
components on the
Components
page of the documentation - panels, tables, etc. - have basic
classes as well as several "subclasses" for variations.
I was drawn to it because its presentational grids implementation is
very similar to the OOCSS framework and it borrows other similar
concepts, e.g. the media object.
So that's my take, but I actually found out about it because John
announced it in this forum; if you search the archives for "John
Slegers" you'll see several posts about it.
As to why I didn't just use the OOCSS framework created by Nicole
Sullivan and her team, it mainly came down to the fact that Cascade
Framework is a little more user-friendly in case I want coworkers to
learn it as well; OOCSS doesn't have as many ready-to-go styles and
variations and you have to download it and build it to even see the
latest docs.
I do think that Cascade Framework could be improved, beyond just
moving it to SASS as John is already working on. For one thing, it
uses a global clearfix on all block-level elements, which isn't good
for performance (although for non JS-heavy sites it probably doesn't
matter) and also causes you to have to scroll past it every time you
use Firebug. So I created a new class called .row that has the
clearfix and just put my .col's inside of that.
The other thing I don't like about the official Cascade Framework
examples is all the .cell divs, which are only there to support IE7
and below (otherwise you could just use border-box and padding).
First of all IE7 has less than 2% market share these days but of
course it would be nice not to have the layout be
completely
broken on IE7. The new version of the OOCSS framework has a clever
solution to this which is to use a negative margin on the .line
class (which is analogous to my .row class).
On the downside for the OOCSS framework is the fact that it still
requires the use of .lastCol for the last column in every row, in
order to work in IE8 and below, which I've always found annoying.
I'm actually not clear why it's still needed though; in the previous
version of the OOCSS framework I thought it was needed for IE8 and
below because the .unit class was not self-clearing, but in the new
version the columns are self-clearing so I'd have to play with it in
IE8 to see why it's still there; in any case, it's certainly
possible to do without the .lastCol class as Cascade Framework and
many others have done.
The docs on there (at least the first few I clicked on) are using
the old version of the OOCSS framework. I'd recommend looking at the
new SASS version instead (or in addition), which is now the official
version on github:
https://github.com/stubbornella/oocss
You have to download and build it to see the docs; I've already
built them so if you want me to upload them so you can check it out
before installing let me know.
I think this might be the best option for you given what you've said
in your posts...for one thing, it was created by Nicole Sullivan and
her team so you can be sure it follows OOCSS principles. Zurb
Foundation is good too and you could still use certain components
from Zurb together with OOCSS. As others have often said here in the
past, OOCSS is more of a general philosophy/approach and you can
apply it to whatever frameworks and libraries you happen to be using
(although some make it easier than others of course). (Side note - I
would say the same about actual object-oriented programming - it's a
paradigm, not just a collection of techniques. In fact, Alan Kay
once said "I invented the term object-oriented programming and this
[C++] is not what I had in mind" because the paradigm got somewhat
lost through history...but that's another story that I can tell you
about another time if you're interested :)
Cheers,
Matt