[comp.windows.x.motif] Mwm question

nazgul@alfalfa.com (Kee Hinckley) (05/18/91)

Several people have noted that my response to the question about
Mwm's placement of icons was not terribly substantive and rather
rude in tone.  My apologies, it certainly would have been better
sent via mail, or not at all.  I would like to clarify however that
I wasn't just being flip when I suggested that one of the reasons
mwm (and Motif) are difficult is because they offer so many options.
I consider that to be both an advantage and disadvantage.  GUIs 
like Open Look considerably limit your options, but the lack of
choices does make things easier.

Back to the question at hand though.  After playing with it I'm
not entirely sure what algorithm Mwm *does* use.  If I deiconify
a window, and then create a new window and iconify it, it fills
the space left by the last window (this is Mwm 1.1).  In my mind
that space ought to (and this is what I had thought Mwm did) be
left open so long as a window corresponding to it exists.  In 
other words each window should have an icon position that is
remembered and not disturbed unless the user explicitly moves
things around.  Thoughts?

					-kee

don@atlantis.coral.com (Don Dewar) (05/18/91)

) Return-Path: <uunet!alfalfa.com!motif-request>
) Subject: Mwm question
) To: motif@alfalfa.com (The Motif Discussion List)
) Cc: kddlab!cs.titech!wnoc-tyo-news!dclsic!sjc!leia!harkcom (Alton Harkcom)
) X-Mailer: Poste 1.0
) From: uunet!alfalfa.com!nazgul (Kee Hinckley)
) Date: Fri, 17 May 91 20:01:48 -0400
) 
) Several people have noted that my response ...
) 
) Back to the question at hand though.  After playing with it I'm
) not entirely sure what algorithm Mwm *does* use.  If I deiconify
) a window, and then create a new window and iconify it, it fills
) the space left by the last window (this is Mwm 1.1).  In my mind
) that space ought to (and this is what I had thought Mwm did) be
) left open so long as a window corresponding to it exists.  In 
) other words each window should have an icon position that is
) remembered and not disturbed unless the user explicitly moves
) things around.  Thoughts?
) 
) 					-kee
) 
) 

And here we go...  I disagree and it is a matter of personal
perference and style.  I like to conserve screen space as much as
possible and I bring up lots of clients and "iconify" and "deiconfify"
them all the time.  I think the current algorithm makes better use of
screen realestate.  I believe your method gives a better sense of
spacial reference of where the icon will be, but I don't find that
particularly useful, whereas some else might.

This brings to mind your original statement about mwm being too
flexible.  Here is a case where I would like to do it one way and you
would like to do it another.  Should this be yet another user
cutomizable feature of mwm?  I don't believe you could argue that
either algorithm is better.  They both have advantages and
disadvantages, so which do you choose.

Here is my interpretation of the alogorithm by the way.  A client will
be iconified back to the same place if that space is available.  If
you iconify a client that has never been iconified before, it will be
iconfied in the first available space.  If a new client has been
iconfied into a space used by an existing client, the existing client
will be iconified into the first available space.

  +---------+
  | Coral   |
  |@@@@@*@**|
  |@@*@@**@@|     Don Dewar
  |*@@**@@@@|     Coral Network Corporation, Marlborough, MA
  |@***@@@@@|     Internet: don@coral.com
  |@@**@@@@@|     Phone:    (508) 460-6010
  |*********|     Fax:      (508) 481-6258
  |Networks |
  +---------+