Just a bit of musing

[coder at ibm.net]

On 28/02/97 at 04:04 PM, S001GMU at nova.wright.edu said:

>what's wrong with making them consciously ask for more detail?  even in a
>visual situation, you have to consciously focus on specific areas in
>order to see anything beyond a general shape.  Granted, a graphical
>client does allow for large amounts of information to be displayed with
>less interpretation time for the user, but again there is still only so
>much info you can put on a screen.  I think it's not at all unreasonable
>for designers to require some level of interaction with the player to
>determine what specifically their character is looking at.

This really comes down to interface design, which for a graphical system
currently seems to devolve into religious points.  Dr Cat on r.g.m.misc(?)
argued the comparitive benefits of each fairly thoroughly about 7 - 8
months ago.  Unfortunately I don't have his posts handy (see dejanews),
but from recollection:

  -- over head map/tile view (difficult to represent vertical placement 
     or vertical occlusion, in large scale relevancy problems can become 
     nasty, usually works best for very localised processing (local room 
     only), argued to discourage identification with character)

  -- 1st person (panning disorients many, problems with peripheral vision 
     and actions/events behind, problems with manipulation and vituality 
     of self, orientation sense acn be a BIG problem, good for 
     identification with character)

  -- Behind the head (see 1st person, but easier manipulation/awareness 
     of self, not as good at identification, but orientation much improved).

  -- 3/4 isometric view (real problems with occlusion (eg the painting on 
     the far side of the wall which is between your viewpoint and your 
     character), verticality, etc, not as good for identification, 
     manipulation often suffers)

<<Note the latest Mario Bros game takes a very high rolling (and arguably
well done) variation on #3 which can be _extremely_ disorienting,
sufficiently so to make it difficult to gut predict what effect a motion
controll will really have on Mario, or even high the viewpoint will move
in relation to Mario (side on, behind, how distant, etc)>>

Frankly I think its all a representation problem, not an interface design
problem, and actually doesn't matter a hoot.  I don't see that it should
change how the system internally reprexsents or manipulates the data set
either way.  Adroitly avoiding the questions of detail level and
verisimilitude, I see little reason why a decent mapping and spatial
system couldn't easily be represented in any of them.  (cf Descent and the
1st person view as versus the external 3D rotating map view)  Rolling this
into the game model (learnable character skills?) is a possible out which
could be a real value adder.

--
J C Lawrence                               Internet: claw at null.net
----------(*)                              Internet: coder at ibm.net
...Honourary Member of Clan McFud -- Teamer's Avenging Monolith...