wescott@LNIC1.HPRC.UH.EDU (Andrew M. Wescott) (04/24/89)
Hey does anyone wonder what OSF/Motif coupled with the
DM would look like? Well take a look at pages 230-231
of the May 1989 BYTE MAGAZINE.
I'm ready!
Andrew Wescott
University of Houston
Department of Chemical Engineering
weiner@novavax.UUCP (Bob Weiner) (04/28/89)
In article <8904241457.AA00540@lnic1.hprc.uh.edu> wescott@LNIC1.HPRC.UH.EDU (Andrew M. Wescott) writes: > Hey does anyone wonder what OSF/Motif coupled with the > DM would look like? Well take a look at pages 230-231 > of the May 1989 BYTE MAGAZINE. > > I'm ready! I have to put a couple cents in here. I have not seen Byte but I have heard Motif 'looks' really nice. Fine, pretty 3-D windows and buttons. What we need is not only easy to use programs but powerful ones. Examples: One of the easiest ways for a program to make your keyboard (remember, the thing that you can interface with on every computer and terminal) very useful is to support key SEQUENCES that can be bound to arbitrary functions just as most Emacs versions do. Menus and mice movements are just too cumbersome in many respects and no one yet has a good model for building up powerful abstractions with ONLY a mouse based graphical interface. Who wants to bet Motif's style guide does not suggest a powerful keyboard model? Just look at all the silly macro makers on Macintoshes and you'll understand how widely ignored an issue this is. What about the complexity of writing applications for Motif? X Windows already extremely complicates the development process (though it does aid portability). Unless the OSF is ready to release a complete object-oriented development environment that supports Motif application development, this represents a major developer headache. That translates directly into more expensive software. (The price you pay for technology tradeoffs that make a complete interface environment portable between PC-type and UNIX machines.) How come no one is talking about the hypertextual capabilities of HP's New Wave which is part of Motif. There some real progress could be made. Is it because few people understand this technology? I recently attended a presentation by some significant OSF representatives. The development guy knew what he was talking about but acknowledged that there were many BASIC problems with their ideas that they had not figured out yet. The spokesman and the federal systems representative were very close to Zippy the Pinhead, in bloodlines at least. I want the OSF to succeed but lets not act like lemmings and follow them to the death simply because of their well known backers. SUN reps claim that IBM (the guys OSF paid about $80 million for AIX) has publicly stated that they do not need to develop OSF-compatible environments since they already have AIX and can develop it themselves! -- Bob Weiner, Motorola, Inc., USENET: ...!gatech!uflorida!novavax!weiner (407) 738-2087
nazgul@apollo.COM (Kee Hinckley) (04/29/89)
In article <1235@novavax.UUCP> weiner@novavax.UUCP (Bob Weiner) writes: >In article <8904241457.AA00540@lnic1.hprc.uh.edu> wescott@LNIC1.HPRC.UH.EDU (Andrew M. Wescott) writes: > >> Hey does anyone wonder what OSF/Motif coupled with the >> DM would look like? Well take a look at pages 230-231 >> of the May 1989 BYTE MAGAZINE. >> >> I'm ready! > >I have to put a couple cents in here. I have not seen Byte but I have >heard Motif 'looks' really nice. Fine, pretty 3-D windows and buttons. >What we need is not only easy to use programs but powerful ones. > >Examples: > >interface. Who wants to bet Motif's style guide does not suggest a >powerful keyboard model? Just look at all the silly macro makers on >Macintoshes and you'll understand how widely ignored an issue this is. Motif supports full keyboard access to all of the functionality. In addition to the standard "accelerators" there are mnemonics for pulling down all menu items, selecting from lists, moving from one button to the next, and so on. Believe me, after six years using an Apollo I am not about to move to an interface that doesn't let me have a next-window, pop and shell key, or one that forces me to grab my mouse everytime I want to do something useful. That was one of the things I stressed when I worked on the UEC selection team at OSF. What Motif does *not* do is have the kind of macro programming functionality that the Display Manager gets you. There are ways to add this functionality to X and Motif in particular, but it is non-trivial. (See "An Object Oriented Extension Language for Integrating Disparate Applications", Interact '87). >aid portability). Unless the OSF is ready to release a complete >object-oriented development environment that supports Motif application >development, this represents a major developer headache. That The X Intrinsics, on which Motif is based, support an object oriented model. We are in the process of integrating Motif with Open Dialogue to provide even more object oriented and dialogue management features. In either case programming an interface with Motif should *substantially* reduce development time over trying to do one from scratch in *any* graphics toolkit. >How come no one is talking about the hypertextual capabilities of HP's >New Wave which is part of Motif. There some real progress could be >made. Is it because few people understand this technology? Motif has nothing to do with New Wave. New Wave was not submitted and does not currently (so far as I know) run on Unix. Motif has the same *feel* as the Presentation Manager, but that's about the only thing that relates them. >SUN reps claim that IBM (the guys OSF paid about $80 million for AIX) >has publicly stated that they do not need to develop OSF-compatible >environments since they already have AIX and can develop it themselves! So? This whole question of whose OS your are going to run is really nonsense. So long as they are Posix/XOpen compatible do you really care? Yes, certain things would be nice. Shared libraries, typed file systems, etc.. When those things get added we need to standardize on them too. But as to whether you bought it from AT&T or Sun or OSF or IBM? You're going to spend a lot more time trying to port your User Interface from Open Look to Motif than you ever would porting the OS parts. All opinions are mine, of course. Kee Hinckley User Environment Apollo Computer -- ### User Environment, Apollo Computer Inc. ### Public Access ProLine BBS ### ### {mit-eddie,yale}!apollo!nazgul ### nazgul@pro-angmar.cts.com ### ### nazgul@apollo.com ### (617) 641-3722 300/1200/2400 ### I'm not sure which upsets me more; that people are so unwilling to accept responsibility for their own actions, or that they are so eager to regulate everyone else's.
weiner@novavax.UUCP (Bob Weiner) (04/29/89)
>>Who wants to bet Motif's style guide does not suggest a >>powerful keyboard model? Just look at all the silly macro makers on >>Macintoshes and you'll understand how widely ignored an issue this is. >Motif supports full keyboard access to all of the functionality. In >addition to the standard "accelerators" there are mnemonics for pulling >down all menu items, selecting from lists, moving from one button to >the next, and so on. Believe me, after six years using an Apollo I >am not about to move to an interface that doesn't let me have a next-window, >pop and shell key, or one that forces me to grab my mouse everytime I >want to do something useful. That was one of the things I stressed when >I worked on the UEC selection team at OSF. I think you missed much of my point here. I said, powerful keyboard models. The standard DM has a very weak one as does the Macintosh because the standard key binding facilities do not support key sequences. That is, using arbitrarily bound keys that act as prefixes to other keys, e.g. Control-x Control-s executes one well defined function. The Macintosh gives typists contortions by making them press multiple modifier keys AT ONCE, e.g. COMMAND-OPTION-SHIFT-T; hands are not constructed to do this well. Your description makes it sound as if the OSF wants to take something akin to the dedicated editor keys approach seen in Apollo's DM and DEC's EDT keypad. Now remember the home keys that every typist learns to center on? Any extreme movement from these can slow down input/editing/command productivity incredibly. Instead the OSF should at least use one of the well known, well refined through years of use, home key-based keyboard models that supports fully user defined key sequence bindings. This should be a system-level feature so that application developers do not have to waste their time reinventing it. OSF has a chance. Encourage them not to reproduce the user interface atrocities of the the past in a new generation of computers. As a side note, you might be interested to know that SUN took its very reasonable keyboard layout and shrunk the key spacing and key size and stuffed in extra keys in strange places so that it is virtually impossible for a normal sized adult hand to rest comfortably on and to use the keyboard. This is for their new SparcStations but it is their new keyboard 'standard' for future models. Someday, probably in a distant galaxy or Italy, people will recognize the value of human interface engineering for computer systems. >What Motif does *not* do is have the kind of macro programming functionality >that the Display Manager gets you. There are ways to add this functionality >to X and Motif in particular, but it is non-trivial. (See "An Object Oriented >Extension Language for Integrating Disparate Applications", Interact '87). Macros are not enough anyway. Again and again people have incrementally expanded macro languages to fully structured programming languages in order to meet design needs. Of course, starting with a well designed programming language makes the end result much better. Again, this should be a standard tool available from the system level. I'm sure hardware technology will keep up with any necessary increase in system overhead. >>Unless the OSF is ready to release a complete >>object-oriented development environment that supports Motif application >>development, this represents a major developer headache. > >The X Intrinsics, on which Motif is based, support an object oriented >model. We are in the process of integrating Motif with Open Dialogue >to provide even more object oriented and dialogue management features. >In either case programming an interface with Motif should *substantially* >reduce development time over trying to do one from scratch in *any* >graphics toolkit. This is good. Now if you could find someone who could design useful user interfaces, it would be even better. I have seen over ten applications based on different flavors of Dialogue and they all of been more primitive then some of the interfaces I have seen children design with Logo. >>How come no one is talking about the hypertextual capabilities of HP's >>New Wave which is part of Motif. There some real progress could be >>made. Is it because few people understand this technology? > >Motif has nothing to do with New Wave. New Wave was not submitted and >does not currently (so far as I know) run on Unix. Motif has the same >*feel* as the Presentation Manager, but that's about the only thing >that relates them. This is bad. Data and information overload on large workstation networks is becoming a major problem. Vendors should realize the potential benefits of all of the hypertext/user environment/control methods research work that has gone on in the last twenty years. Compare Apple's Hypercard to Brown's InterMedia project and you'll see the difference between what the masses are fed and what bright minds can produce if given the chance. (If you think that Apple's Macintosh took all of the great ideas from Xerox PARC, you are very wrong. They took only the shell. Now the OSF should work to give everyone some of the meat. >You're going to spend a lot more time trying to port your >User Interface from Open Look to Motif than you ever would porting the >OS parts. This is true, so how is the OSF helping? I'm sorry this sounds so religous, but if the researchers and engineers don't stand up for solid designs in all facets of production, the business people certainly are not going to say, "Go for it." -- Bob Weiner, Motorola, Inc., USENET: ...!gatech!uflorida!novavax!weiner (407) 738-2087