[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[STUMP] My patches for saving window placement data in window properties

From: Michael Raskin
Subject: [STUMP] My patches for saving window placement data in window properties
Date: Wed, 29 Oct 2008 11:45:20 +0300
User-agent: Thunderbird (X11/20080904)

        Before fixing or abandoning (and making a forever private branch) my
patches for saving window placement data in one more semi-persistent
place, I would like to understand my options..

        What I want to say about intents/limitations of existing/proposed

        1. I acknowledge that in the simplest case (user creates some groups
and frames via .stumpwmrc and uses only them; for some reason StumpWM is
restarted) current solution works better than anything else could offer
(fully automatic and correct).
        2. Current solution will never be extensible enough to keep 2 layouts
to switch between. Mine requires just optional property name prefix.
        3. File dump/restore works, of course, but requires user to think about
file lifespan. Natural life cycle of such information is "as long as
original windows (and original X server) are alive", which can be longer
than StumpWM process lifespan, and hard to determine - so it is left to
manual work..
        4. Not-completely-unlikely case "curse, launch IceWM for half an hour,
return to StumpWM" is well-covered by my solution, but cannot be covered
by existing one. Even with floating groups there may be a window that is
more convenient to keep tiled 90% of time, and its home frame should not
be lost if it is moved to floating frame.
        5. The same holds for having functions defining different frame
layouts. I want to call that function and only then restore positions;
with existing solution the data is lost before I get any control.
        What I want to ask:
        1. Whether any additional functionality provided by my solution is
considered generally useful?
        2. What should I change for patch to be considered useful enough to be
accepted (and not messy enough to be rejected)?

reply via email to

[Prev in Thread] Current Thread [Next in Thread]