Sander van Dijk <a.h.vandijk_AT_gmail.com> once said:
> On 4/21/06, Denis Grelich <denisg_AT_ueberl33t.info> wrote:
> > Hello all,
> >
> > I find it quite cumbersome that views that don't have any clients are
> > destroyed automatically.
>
> Well, creating them automatically, but having to destroy them manually
> would be quite weird and inconsistent I think...
>
> > Firstly, it adds some unpredictability: you don't know where you end up
> > when your current view is destroyed.
>
> True.
>
> > Secondly, the current behaviour is quite annoying in that respect that it
> > forces me to re-type the tagname if I accidently destroyed a window or if I
> > want to remove a tag from a window but add another client to that view.
>
> The same applies to columns, if you want to remove a client from a
> column, but add another one to it, you just have to either do that in
> the correct order (add new client, than remove the other one), or
> recreate the column; the same does and should apply to views.
Everyone seems to be thinking about views as what would be called
workspaces in the classical sense. There is only _one_ view, the
set of clients you're currently looking at. Think of all clients
being in a big pool each of them with a certain attribute that
sets them apart in groups, the tag. When you construct your
current perspective, the view, you're calling forth all clients
with a certain attribute.
All this talk about destroying views and selecting other views
shouldn't even be a concern. When all clients bearing the specific
tag of your perspective are gone, you should be viewing nothing at
all. This doesn't mean you're in a new view .. or in an empty view,
it just means you're viewing the set of clients with the "nil" tag
which no client should have.
Correct me if I'm wrong, but that's how I interpret views.
> > Last but not least, this behaviour was introduced as a means to add some
> > consistency to column behaviour, but actually, there is about zero relation
> > between clients and views?
>
> There's definitaly a relation between frames, columns and views:
> allowing empty ones is undynamic, and on top of that allowing some of
> them to be empty but not others would be inconsistent as well; if you
> cannot select empty views, views shouldn't be allowed to remain in
> existence when they become empty either; as soon as something becomes
> empty, it becomes obsolete, I believe it's good that this applies to
> all 'objects' in wmii (frames, columns, and views).
In my opinion, columns and views are inherently different. One
deals with layout and the other deals with content.
regards,
ality
-- "The right to be left alone is indeed the beginning of all freedom."Received on Fri Apr 21 2006 - 15:10:07 UTC
This archive was generated by hypermail 2.2.0 : Sun Jul 13 2008 - 16:03:05 UTC