Re: Conceptual models (was: Re: Detachable menubar idea)

[raster redhat com]

On  3 May, Jurgen Botz shouted:
->  Not only doesn't the detachable menubar seem terribly useful, but the
->  situation is much worse... it is symptomatic of a major problem with
->  Gnome; the lack of coherent conceptual models.
->  
->  Ok, so we have a detachable menu bar... not a bad idea in principle; by
->  detaching it I can prevent it from taking up space in my window, which
->  can be a good thing.  But now that I've detached it how can I tell it
->  apart from other detached menu bars?  Also, how do I re-attach it (at
->  the moment the only way to do so is to close it via an ICCCM message)?

1. drag it back where you detached it from and it will plonk back into
place
2. have your window manager put decorations on it (fvwm has options, so
 do all its derivatives (afterstep, fvwm95 etc.) - windowmaker will
 deocrate the window reagrless with a title, so will enlightenment
 0.13.3 and earlier (0.14 will always not put decorations on these as a
 mater of honoring the hints set and aesthetics). 

->  If there's a single application that can have multiple windows (like,
->  say, the gnome-terminal) will there be a single detached menu for all
->  windows (not currently) or do new windows still have their own menu
->  bars (currently so)?  If we had a concept of application layers (a'la
->  Macintosh or NeXT) then detached menubars would have a more
->  identifyable identity and a single detached menubar for all an
->  application's windows would make sense... but not as it is.

Hmmmmm. Call me a biggot and flame away but I'm goignt to poitn a big
finder at the user who detaches 10 menus and forgets whihc ones are
which and say "thats you're problem - you deteached them" It's like
forgetting what "the button with the X does" on the titlebar. Sometimes
we have to assume some competancy on the user becuase bending over
backwards for even the most newbie user will more often than not make
the iunterface a kludge, unisable and horrible for the power user or
those who have learnt about it.. personally i'd liek the option of
"hidng" the menu and turnign it into a "floating" menu liek gimps, or
ee's (right mouse button anywhere in the app bring sup the main menus
unless context dictates another menu for that area).

->  Currently there just aren't any easy answers to these questions, and 
->  frankly, even if we come up with sepecific solutions to some of the
->  most glaring problems (i.e. detached menubar identity), I'm afraid the
->  solution will be just another special case that doesn't fit into a 
->  coherent whole.

yup.. and fix it for person A - you make it hoorid for person B.. thats
wat I'm trying to say...

->  A GUI framework like Gnome should consist of a small set of conceptual
->  models from which all behaviors can be somewhat intuitively inferred...
->  the Macintosh and NeXT GUIs adhere very well to this design principle and

macos's menubar at the top is BAD for any reasonably competeant user..
why? to access ANY meny uption you have to take a jorney of discovery
ALL the way to the top of the screen to get to the menu with the pointer
from where it was in the app.. now the amiga's magicmenu whihc did what
i suggested wiht floating menus (turned the apps menu into a right-mouse
anywhere in the apps window brigns up the aps menu) is already much
better, and faster to use...

->  even Windows is slowly by surely getting better in this regard (until the
->  '95 GUI Windows was an absolute disaster from a conceptual model
->  perspective, and even in the modern versions there are a lot of
->  inconsistencies.)
->  
->  I am afraid that Gnome is falling into the trap of just kind of "growing
->  together"... we may make good UI design decisions on particular details,
->  solving particularly hairy inconsistencies one at a time, but will end up
->  with something that simply doesn't have any unifying overall models, thus
->  constantly surprising the user with unexpected behaviors.
->  
->  I am therefor calling for everybody who is interested in the larger GUI
->  issue to take a step back and think about conceptual models for user
->  interaction, maybe go play with a Mac or a NeXTstep box, and start some
->  discussion here on the list.  We need to fix this now, while everything is
->  still in flux, so that we can have a GUI that's truely better than Windows
->  from the start.

that won't happen. GNOMe needsa to reach critial mass - it needs lots
of development, apps and a usable environment.. we simply can't afford
to suspend development for months in a flame war of whose idea of a gui
is best... in my view next, macos and windows all are a kludge, ugly
and are a pain to use.. amigaos (3.0+) was better (especially with MUI)
but it also needed improvement.. the key is user configurability.. dont
enforce a policy on everyone - give the user the choice what policy
they want from a large list. MUI did this with its gui conig tool and
it was brilliant - it let the power user re-design his look ANd feel
ANd function of his desktop. The default settings that gnome gets "out
of the box" can eb argued about once such a system is in place.. but
creating such a system will take time.. lots of it.. and right now we
don't have that luxury...

->  --
->     ~~/            /~)
->      /.. 	 /-< 
->   \_/ u r g e n  /_ _) o t z      "Unix?  What's that?  Is that like Linux?"
->  
->  
->  

-- 
--------------- Codito, ergo sum - "I code, therefore I am" --------------------
raster@rasterman.com       /\___ /\ ___/||\___ ____/|/\___  raster@redhat.com
Carsten Haitzler           | _ //__\\ __||_ __\\ ___|| _ /  Red Hat Advanced
218/21 Conner Drive        || // __ \\_ \ | |   \ _/_|| /   Development Labs
Chapel Hill NC 27514 USA   ||\\\/  \//__/ |_|   /___/||\\   919 547 0012 ext 282
+1 (919) 929 9443, 801 4392   For pure Enlightenmenthttp://www.rasterman.com/