W3C home > Mailing lists > Public > www-style@w3.org > June 2012

Re: [css4-ui] Scrollbar tracking control

From: Sebastian Zartner <sebastianzartner@gmx.de>
Date: Sat, 30 Jun 2012 00:41:07 +0200
Cc: derhoermi@gmx.net, jackalmage@gmail.com, www-style@w3.org
Message-ID: <20120629224107.247480@gmx.net>
To: Brad Kemper <brad.kemper@gmail.com>
> You mean it just doesn't feel right?
Well, yes.

> > and having a longer name isn't always bad as long as it is more
> descriptive as a short one.
> 
> Sure. But you have to admit, 'overflow-attachment-y' is pretty darn 
> long.
Sure, it's not the shortest name. And I am open for other suggestions.

> I don't think it is really describing it that well. 'overflow-
> anchor[-x | -y]' would be better
Sounds ok to me.

> > My proposal would also have the benefit that you could combine the
> > value in the 'overflow' property and use it as shorthand property. 
> > E.g. like this:
> > 
> > overflow: scroll normal bottom;
> 
> Sure. You could still do that with my naming variation, since I am not
> touching the values. I think your better argument is to have a 
> 'overflow-attachment' shorthand property that combined x and y, like 
> this:
> 
> overflow-attachment: <overflow-attachment-x> <overflow-attachment-y>
> 
> Which could then be shorthanded into 'overflow' as you suggested:
> 
> overflow: <overflow-attachment>

Yes, that's one benefit of my proposal.

> So, my version would be like this:
> 
> scroll-anchor: <scroll-left> <scroll-right>
> overflow: <scroll-anchor>

Normally longhand property names are having the shorthand property name as prefix. Having 'scroll-anchor' people probably wouldn't know that they can put it into 'overflow'.

> > > And it could be writing mode dependent, if we also had those 
> > > keywords (er, begin/stop and head/foot?
> > Yes, I didn't consider this before. What about begin/end for both
> properties? What are other properties using here?
> 
> I think flex-box is doing that (start/end for both vertical and
> horizontal), but only because, uh, well, I a, not the best one to 
> explain that. I think because it needs to have a vertical and 
> horizontal that are more divorced from writing mode. Or something. 
> Sorry I haven't been following that quite as closely.
I don't believe people would mix up the writing mode with flexbox. But that's not the topic of this issue.

> > > Or whatever it is this week?).
> > Of course it should be something consistent with other properties'
> values.
> 
> Agreed. I'm being a bit snarky because start/end are apparently well
> established for bidirectional writing, but still hard for me to 
> remember, because the opposite of "start" should be "stop" or "finish", 
> and the opposite of "end" should be "begin" (you even paired begin/end):
> 
> ~ One can start smoking and stop smoking, but it is quite a different
> thing to end smoking. 
> 
> ~ One can start running in a race when the starting gun is fired, and 
> stop running when they reach the finishing line, but the race doesn't 
> end until everyone is done or some time limit is reached. 
> 
> ~ books, articles, movies, and songs all have beginnings and endings. 
> You can start or stop reading, watching, or listening at any point, but 
> the piece generally begins at the beginning and ends at the ending (you 
> can personally start reading, watching, or listening at the beginning, 
> and stop reading, watching, or listening at the end, or you could just 
> start and stop reading, watching, or listening to a middle part).
Ok, thanks for the good examples. They really clarify this. So it sounds I wasn't that wrong with begin/end even when I'm not a native English speaker. :-)

> Anyway, if you only wanted two directions, it would be start/end, 
> because those keyword are established for bidi. But I think you'd not 
> get away with that. You need all 4 logical directions to be consistent 
> with most other things that (will) have four logical directions. The 
> other two are under discussion to change from before/after to 
> head/foot.
As I said as long as it complies with other directional property values I think it's fine.

> > The alternative (when staying with left/right and top/bottom) would be
> to use the :dir() pseudo-class [1]. So somebody would have to write
> this to cover different writing modes:
> 
> I don't like that. It seems clumsy and awkward to me.
I agree. I just mentioned it to point out an alternative.

Sebastian
-- 
Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir
belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de
Received on Friday, 29 June 2012 22:41:36 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:55 GMT