Re: [wmii] Column Layouts

From: Jonas Domeij <jonas.domeij_AT_gmail.com>
Date: Sat, 29 Apr 2006 04:05:12 +0900

Hi. I've been trying to catch up with the development of wmii after
some time of absence, trying the last bunch of snaps for a while.

To me it sounds like David Arroyo is proposing a more acme-esque
approach. Instead of making columns behave differently with modes,
just provide ways of manipulating clients within a column.

On 28/04/06, Anselm R. Garbe <garbeam_AT_wmii.de> wrote:
> Anyway. Your approach has the big disadvantage, that it sounds
> to need much more interactions than pressing a simple key or
> clicking a tag and arranging the clients of a column
> differently. You always need to invoke a resize action, in the
> end this feels less dynamic than the current approach.

It does require more interaction, yes. However, the current way has
the disadvantages that the default mode quickly grows useless for more
than identifying clients (as they're too small to work with), and
stack/max only lets you view a sole client. (As a fun sidenote,
Anselm: default mode will gets more and more useful while stack/max
gets less the bigger your screen is)

Something I've noticed reading through all the old mail is the
dogmatic tendency of referring to a solution as dynamic, while there's
confusion about exactly what dynamic means. This is understandable if
you look it up in a dictionary - especially since many of us are
europeans with a native tongue similar to english, but with subtle
differences.

As a swede, I label something "dynamisk" as kind of a mix of these two
definitions:

2. Characterized by continuous change, activity, or progress: a dynamic market.
3. Marked by intensity and vigor; forceful. See synonyms at active.
(answers.com)

...meaning 'adapting easily, and therefore effective'.

To get to the point, I'd say that Arroyo's approach is in fact /more/
dynamic, Anselm's is only more semi-automatic. This "stack file"
allows the use of more than one client in a column conveniently, but
still with the great limitation that they would have to be
neighbouring clients.

All that being said, in the experimental spirit of wmii, it's probably
good to try both ways (and the some). My only conclusion is that
wmii-3 still has a way to go to perfection.

> > To reiterate, the current implementation of column "modes" is
> > overengineered.
>
> Agreed, and if you think about my proposal, you will notice that
> there is a single algorithm which keeps track of the above
> arrangements based on the stack value.

Isn't it possible to call that single algorithm overengineered as
well? Every time Anselm is optimistic about a general solution with a
consistent concept I get a bit nervous, with things like nested
layouts in memory.

To make a long thread even longer, ironically I'd like to bring up the
discussion about an abandoned mode - the exclusive flag. It seemed
forgotten in the intense debate over empty columns. Many users
switching from wmii-2.5 to a recent snap seem to miss larswm's way of
doing things, after which they're told to get used to columns. Well, I
still miss larswm's simplicity and automacy, which is unsurpassed in
my opinion. I believe the exclusive flag can merge the two in a nice
way.

The problems discussed can be solved by restrictions:
* only allow one exclusive column per view (why would we need more?)
* push all other clients to the right (if a client moves to the
exclusive column, push the other client to the right)
* fetch a client from the right if the sole client in the exclusive
column is destroyed

Unfortunetaly there are probably new problems because of tagging, but
I'll let someone smarter than me ponder about them :)

Regards, zahod.
Received on Fri Apr 28 2006 - 21:05:14 UTC

This archive was generated by hypermail 2.2.0 : Sun Jul 13 2008 - 16:03:59 UTC