On 3/4/06, Anselm R. Garbe <garbeam_AT_wmii.de> wrote:
> And as we wrote in dynamic WM, I really doubt, that static
> layouts ever will properly work and can be easily adapted to new
> requirements or frequently changing tasks.
I believe you missed the entire point of my email. Dynamic window
managers are optimized for not thinking about how to layout windows,
which provides benefits when you're rearranging windows constantly. I
don't care about that. The number of actions I have to take in order
to arrange windows is miniscule compared to the number of times I need
to switch between applications. Moreover, I arrange windows before I
start a task, so taking 5, 30 or even 60 seconds less to arrange my
windows doesn't matter because it's not interrupting my workflow.
By explicitly arranging my windows into an extremely static setup, I
know, without thinking, how to get to any window on my computer
without having to search for the window. This optimizes window
switching, which occurs during normal workflow and so optimizing it is
far more important.
Think about how easy it is to work with two windows using the MS
Windows, Mac OS X, or Metacity window managers. You alt+tab between
the two without thinking about it, it's a O(1) operation. When you add
a third window, the process breaks down and you must search through a
window listing to get to the third window. This introduces a new task
and interrupts workflow, it's an O(n) operation even not considering
the workflow interruption.
Using my wmi setup I can switch between at least 8 windows as an O(1)
operation and no workflow interruption, which is the only reason I use
wmi rather than a more mainstream window manager. This is not possible
using wmii, as the window manager rearranges windows all over the
place so jumping to a given window involves finding the window,
providing no benefit over more mainstream window managers. Switching
between windows is an O(n) operation in wmii and not tacit at all.
Received on Mon Mar 06 2006 - 22:24:44 UTC
This archive was generated by hypermail 2.2.0 : Sun Jul 13 2008 - 16:00:45 UTC