hvr@eng.sun.COM (Heather Rose) (09/18/90)
First, I'll list all the pieces (much of this information can be found in
"The NeWS Book" by Gosling, Rosenthal, and Arden which lists additional
window systems and in greater technical detail):
o SunView is a user interface toolkit built on top of SunWindows.
SunWindows is a kernel-based window system which was one of the first
widely used window systems on UNIX. It was one of the first window
systems to deal with the problems of multiple processes and address
spaces. SunWindows was first released in 1983. The SunView user interface
toolkit was first released in 1984 (the same year Apple introduced the
Mac--remember the Orwellian ads on TV?). SunView had no style guide
or functional specification.
o X began life as a distant relation to the W (1982) window system from
Stanford. In 1983, Project Athena (research project at MIT funded by
DEC and IBM) first began using X on Vax computers.
In 1985, two students from CMU on winter break ported X10 from the Vax
to other popular workstations which made it much more widely used
(some of you may remember the hardcoded key layouts from the Vax Stations.)
In 1987, X10 was reworked into X11 which we are using today. X made
window management into a separate process, and did not specify a look
or feel. The X Toolkit (now referred to as the Intrinsics) first
started life on X10, then provided much feedback for the development
of X11 and Xt on X11 today.
o NeWS (Networked, extensible Window System) was first released in 1986
by Sun Microsystems. NeWS is based on the PostScript imaging model to
leverage a standard graphics model introduced by Adobe and widely used
for printing. NeWS is like the X server or SunWindows in that it also
just manages windows and does not provide a toolkit or user interface.
The first toolkit for NeWS, Lite, was an experimental toolkit written in
object-oriented PostScript. The most widely used toolkit for NeWS
today is the publically available GoodNeWS/HyperNeWS from the Turing
Institute.
o XView is the next generation of SunView on X11. Many of the problems
with the SunView toolkit were fixed in the XView toolkit. XView was
first released with OpenWindows 1.0 in 1989. The source to XView was
made publically available later that year. XView does not use the
Xt Intrinsics, but it's own object-oriented foundation layer. XView
implements the OPEN LOOK GUI with a SunView API.
o X11/NeWS is a combined X11 and NeWS window server. X and NeWS share
the same event stream, the same graphics substructures, the same
window hierarchy. X11/NeWS was first released by Sun Microsystems
with OpenWindows 1.0 in 1989. For example, one can create an X window,
then open a connection to the NeWS side of the server, hand over the
X window id, and then begin rendering PostScript in the X window.
(This is how the XVPS package added to XView works.)
o Xt+ (now called OPEN LOOK Intrinsics Toolkit, OLIT) developed by USL
(UNIX Software Labs was spun off by AT&T as a separate company so other
vendors besides AT&T can have more influence over the direction of UNIX.)
OLIT is a distant relation to the HP widgets donated to the X Consortium.
USL calls OLIT "OPEN LOOK." Sun uses the name OLIT to avoid confusing the
the L&F with an implementation. OLIT implements an OPEN LOOK GUI with
an Xt API.
o The NeWS Toolkit (now called TNT, formally called NDE) was developed
by Sun Microsystems for NeWS. It completely replaces Lite which was
the predecessor to TNT built on NeWS before it was merged with X11.
What distinguishes TNT from most other toolkits is the ability to
dynamically change the L&F via dynamic subclassing. TNT implements
OPEN LOOK with a TNT API.
o OPEN LOOK is a graphical user interface specification and style guide.
Sun, AT&T, and Xerox worked together (with industry review) to define
the look and feel of OPEN LOOK over a 2.5 year period. The goals of
OPEN LOOK were to create a user interface for a multi-tasking,
networked display which would be familiar to users coming from the
PC space, to provide a look and feel which is legally safe (no chance
of a lawsuit), and to provide the right to license the look and feel
at no charge (the GUI L&F is free). With many companies suing over
L&F copyright and others charging outrageous rates just to use their
L&F (not the implementation, just the specification), these were basic
requirements for the GUI.
OPEN LOOK was prototyped with several toolkits and window systems:
SunView on SunWindows,
XView on X11R2, X11R3 and X11/NeWS,
OLIT on Xt on X11R2 and X11R3, and
The NeWS Toolkit on NeWS.
The functional specification and style guide are toolkit and window
system independent.
The OPEN LOOK user interface can be licensed at no charge for any
type of computer.
About one month after OPEN LOOK was announced to be the user interface
for System 5 Release 4, the OSF began an RFT (Request for Technology)
process for what is now called OSF/Motif. OPEN LOOK was submitted by
AT&T to the OSF as a candidate for the graphical user interface
specification and style guide. About one to two months later, OSF
announced the Motif decision. Motif is a combination of technologies
from DEC, HP, and Microsoft.
o OLWM (OPEN LOOK Window Manager) is one name for two implementations.
USL (formally a division within AT&T) releases a version of olwm
with a session manager, swm, with the OLIT toolkit. Olwm from
Sun Microsystems is a window manager for X clients which has some
rudimentary session management built in. Olwm from Sun Microsystems
was first donated to the X Consortium with the XView toolkit in 1989.
The latest donation of XView and olwm was August 1990 based on the
source from OpenWindows Version 2.
Olwm currently does not manage NeWS windows since the NeWS Toolkit
has a built-in window management library much like previous window
systems to X. Since the NeWS Toolkit can be changed at runtime,
customization can happen at any time; thus, there is no need for a
separate window management process for user interface customization.
o OpenWindows is the name of Sun's networked window products. OpenWindows
includes many pieces: X11/NeWS server, Open Fonts, Xt/OLIT, XView,
TNT, olwm, DeskSet, and various libraries and utilities. Open Fonts
are scalable fonts in a format called F3 and includes all the fonts
on a LaserWriter II plus more. DeskSet is a suit of personal
productivity tools including a file manager, mail tool, and calendar.
The rest of the pieces are mentioned above. OpenWindows also will
run SunView programs in binary compatibility mode, and will allow
direct access to the frame buffer (punch through the window system).
OpenWindows has been an unbundled product (have to pay extra); however,
a user's set of OpenWindows is bundled with every diskfull SunIPC.
OpenWindows will eventually become the default window system for all
Sun systems.
Now, having the pieces defined, onto the specific questions...
In article <1990Sep13.182045.16787@elroy.jpl.nasa.gov>,
pjs@aristotle.JPL.NASA.gov (Peter Scott) writes:
|>
|> Hi. I'm trying to get some things straight for a seminar I'm
|> presenting
|> shortly... typically, I'm sure about the more detailed and technical
|> stuff
|> but not some of the more general things.
|>
|> o Open Look is a look-and-feel put out by Sun and AT&T. It
|> comprises a
|> style guide, widget set, and window manager, just like Motif. The
|> source
|> code for all of these items is freely available via FTP.
No, OPEN LOOK is just a user interface specification and application
style guide. i.e. it is an idea not an implementation of an idea.
|> o SunView is the proprietary windowing system developed by Sun that
|> they
|> originally shipped with their machines. SunTools is the API for
|> SunView.
|> Source code to SunView is not available. Does Sun still support
|> or develop SunView?
No, Suntools is really just a program that one runs to start up the SunView
environment. The source to suntools is on Sun systems in
/usr/share/src/sun/suntool/suntools.c. SunView used to be called SunTools.
SunView and SunWindows can be licensed from Sun Microsystems as part of
SunOS. This is necessary since SunWindows is partially implemented in
the SunOS kernel.
The latest release of SunView/SunWindows was with SunOS 4.1. SunView
programs will also run with the OpenWindows environment in binary
compatibility mode. However, most customers prefer native OPEN LOOK to
SunView compatibility.
|> o NeWS (Network-extensible Windowing System) is based on a
|> PostScript
|> interpreter and is a client-server model like X. It corresponds
|> to the X Protocol/Xlib levels of X. Is the source code available
|> via FTP? Does anyone supply (or run) a pure NeWS server or just
|> the combined X/NeWS Open Windows server?
NeWS source is not ftp'able. It can found on the graphics update tape
of System 5 Release 4 or licensed from Sun. NeWS was available from
SGI, Architec, Parallax, the Grasshopper Group, TAG Inc, and more.
For more information, contact:
Scott Manville, at Open Vistas Association,
80 East 11th St., Suite 222,
New York, New York 10003.
(212) 979 5337
|> o The Open Look window manager (olwm) can run applications written
|> for
|> NeWS or X. It does this by running a server from Sun called
|> X/NeWS
|> which serves both windowing systems. Is the source code for this
|> server available via FTP?
Olwm is ftp'able with the XView source. I sent out an announcement on this
a while back. If you need to see it again, send the message:
send intro xview2-announce
send intro xview2-more
send intro xview2-ftp
to the email address: xvstuff@norge.eng.sun.com.
|> o XView is an X toolkit, from Sun, based on the Open Look GUI. I
|> assume that it therefore contains a Sun version of the Intrinsics
|> and a widget library. How different are calls to XView from
|> calls
|> to Xt? Is the source code available via FTP?
XView's foundation is similar to Xt in that they both provide a way to
do subclassing in C. However, the implementation is significantly
different from Xt.
Yes, source is ftp'able. See above.
|> o I'm told that Open Look runs applications written for SunView.
|> Do
|> they need to be modified in any way? Does this mean that I could
|> have on one screen a window containing Xrn and a window
|> containing
|> mailtool? What configuration would I need to make that happen?
OpenWindows will run unmodified SunView binaries in compatibility mode.
|> o Xt+ is a toolkit from AT&T based on the X Intrinsics. Is it then
|> a
|> widget library? A widget library plus some additional Xt-like
|> routines? Is it related to OLIT?
XT+ was renamed to OLIT. OLIT is an OPEN LOOK widget set. It uses
the generic Xt routines with one convenience routine, OlInitialize.
|> o The ballyhoo about GUIs on Unix boxes appears to pit Open Look
|> against Motif. Are there any other contenders at this level
|> that run on many architectures? NeXTStep runs on NeXTs and
|> has been licensed by IBM; is IBM shipping iron with NeXTStep
|> running yet?
Just OPEN LOOK and OSF/Motif are being pushed as standard, licensable
GUI's. As far as I know, NeXTStep is not readily licensable, and it's
not tax free. (i.e. you pay taxes, royalties, for the privilege of
using the L&F which is separate from licensing an implementation of a L&F.)
|> o A tally of applications shipping for different GUIs (_Personal
|> Workstation_, 9/90) shows Open Look eclipsing the competition.
|> Is this lead due to their counting in applications that run under
|> SunView and therefore automatically run under Open Look? What
|> would
|> the tally look like without them? (Tally shows 55 for Open Look
|> vs.
|> 23 for NeXTStep, 22 for OS/2 PM, and 17 for Motif.)
Don't know. Also, all the SunView OPEN LOOK applications (I can only think
of four) are being ported to X11/OPEN LOOK, so it's a moot point. However,
the real reason for the large number of applications for OPEN LOOK is the
volume of Sun and Sun-compatible machines already on the market and the
number shipping today and in the next year. According to Goldman & Sachs,
HP + DEC + IBM combined are scheduled to ship a total of 105K workstations
in the next year, and Sun alone is scheduled to ship 160K workstations
in the next year of which the majority (> 95%) will be Sparc running
OpenWindows/OPEN LOOK.
One can be portable between OPEN LOOK and Motif with either careful
coding structure (perhaps there are GUI-builder tools to address this
point, or special toolkits) or by using Xt-based toolkits which provide
a very similar API.
Regards,
Heather Rose hvr@eng.sun.com
Sun Microsystems, Inc.cflatter@ZIA.AOC.NRAO.EDU (Chris Flatters) (09/18/90)
> About one month after OPEN LOOK was announced to be the user interface > for System 5 Release 4, the OSF began an RFT (Request for Technology) > process for what is now called OSF/Motif. OPEN LOOK was submitted by > AT&T to the OSF as a candidate for the graphical user interface > specification and style guide. About one to two months later, OSF > announced the Motif decision. Motif is a combination of technologies > from DEC, HP, and Microsoft. As a matter of morbid curiosity, has OSF ever made the reasoning that lead to their choice of a combination of elements from DEC, HP and Microsoft public? NB: I'm not trying to start a religious debate here, I just don't understand why OSF did what they did. Chris Flatters
dbrooks@osf.org (David Brooks) (09/19/90)
In article <9009181549.AA29380@zia.aoc.nrao.edu>, cflatter@ZIA.AOC.NRAO.EDU (Chris Flatters) writes: |> |> As a matter of morbid curiosity, has OSF ever made the reasoning that lead |> to their choice of a combination of elements from DEC, HP and Microsoft |> public? Sure; a Rationale is an essential part of our process. The User Environment Component rationale is quite a thick document, and certainly went to all Members. I can't find my copy right now. Kee? -- David Brooks dbrooks@osf.org Systems Engineering, OSF uunet!osf.org!dbrooks Experience Hackvergnuegen!
ben@hpcvlx.cv.hp.com (Benjamin Ellsworth) (09/19/90)
> As a matter of morbid curiosity, has OSF ever made the reasoning that > lead to their choice of a combination of elements from DEC, HP and > Microsoft public? I'll be interested to read an official OSF statement myself. In the meantime, here's my personal opinion of why OL didn't make it. Given: - Multi-tasking Commercial/Business workstations are a hot market; perhaps *the* hot workstation market for the next decade. - These workstations should be running un*x. - The current single tasking workstations are either MSDOS or Mac's. - Microsoft appears to be much more open (and less anxious to litigate) than Apple. It seems quite reasonable to conclude: - Microsoft compatible GUI's are going to be much more acceptable to the Commercial/Business folks that some new unknown interface. Once we have come to this conclusion, we notice: - OL is not Microsoft compatible, and to make it so destroys its open-lookness. - HP's CXI was nearly compatible, and DEC's could be force fitted to the HP CXI model. The decision to go with a combined HP/DEC offering rather than an OL offering becomes obvious. ----------------------------------------------------------------------- Benjamin Ellsworth | ben@cv.hp.com | INTERNET Hewlett-Packard Company | {backbone}!hplabs!hp-pcd!ben | UUCP 1000 N.E. Circle | (USA) (503) 750-4980 | FAX Corvallis, OR 97330 | (USA) (503) 757-2000 | VOICE ----------------------------------------------------------------------- All relevant disclaimers apply. -----------------------------------------------------------------------
brett@SPIFF.DEN.MMC.COM (Brett Ballantyne) (09/19/90)
In article <9009181549.AA29380@zia.aoc.nrao.edu>, cflatter@ZIA.AOC.NRAO.EDU (Chris Flatters) writes: |> |> As a matter of morbid curiosity, has OSF ever made the reasoning that lead |> to their choice of a combination of elements from DEC, HP andMicrosoft |> public? David Brooks of OSF then replied: Sure; a Rationale is an essential part of our process. The User Environment Component rationale is quite a thick document, and certainly went to all Members. I can't find my copy right now. Kee? Not to start another Motif vs. OPEN LOOK holy war, but how thick does a document have to be to state they adopted a GUI L&F submitted by those companies who give them money? I've read several places that OPEN LOOK is technically superior to Motif (a debatable issue to be sure, but that seems to be the general consensus -- at least in the trade rags I read), so why would they choose Motif? Simple. IBM, DEC, HP, etc., give OSF money and Sun and AT&T don't. I'm not saying that's wrong, but to claim to have industry's best interest at heart and not just OSF member's interest at heart is misrepresentation and *that* is wrong. If I'm mistaken, please set me straight, but that's the way the world appears from the window I look out of. A disclaimer: I'm not religiously tied to Motif or OPEN LOOK -- as long as the UI is easy to use and is coherent, most of my users don't care what the widgets look like. Brett Ballantyne brett@spiff.den.mmc.com
nazgul@alphalpha.com (Kee Hinckley) (09/20/90)
Some nits and questions. In article <9009171845.AA02659@kimba.Eng.Sun.COM> hvr@eng.sun.COM (Heather Rose) writes a lot: > widely used window systems on UNIX. It was one of the first window > systems to deal with the problems of multiple processes and address > spaces. SunWindows was first released in 1983. The SunView user interface It lagged Apollo's Display Manager by several years in dealing with that "problem". I guess if there was nothing inbetween that counts as "one of the first". > at no charge (the GUI L&F is free). With many companies suing over > L&F copyright and others charging outrageous rates just to use their > L&F (not the implementation, just the specification), these were basic What company is charging for using the L&F? > About one month after OPEN LOOK was announced to be the user interface > for System 5 Release 4, the OSF began an RFT (Request for Technology) Is it a required part of 5.4? I thought they backed off of that. > process for what is now called OSF/Motif. OPEN LOOK was submitted by > AT&T to the OSF as a candidate for the graphical user interface > specification and style guide. About one to two months later, OSF > announced the Motif decision. Motif is a combination of technologies Five months. >XView's foundation is similar to Xt in that they both provide a way to >do subclassing in C. However, the implementation is significantly Is that foundation availble to non-toolkit writers who wish to add a "widget"? >Don't know. Also, all the SunView OPEN LOOK applications (I can only think >of four) are being ported to X11/OPEN LOOK, so it's a moot point. However, >the real reason for the large number of applications for OPEN LOOK is the >volume of Sun and Sun-compatible machines already on the market and the This doesn't necessarily follow. As others have pointed out, there are a number of applications shipping (even as their first machine!) on Sun machines that use Motif rather than Open Look. >One can be portable between OPEN LOOK and Motif with either careful >coding structure (perhaps there are GUI-builder tools to address this >point, or special toolkits) or by using Xt-based toolkits which provide >a very similar API. I haven't seen any of those yet :-). -kee -- Alphalpha Software, Inc. | motif-request@alphalpha.com nazgul@alphalpha.com |----------------------------------- 617/646-7703 (voice/fax) | Proline BBS: 617/641-3722 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.
nazgul@alphalpha.com (Kee Hinckley) (09/20/90)
In article <1990Sep18.172142@osf.org> dbrooks@osf.org (David Brooks) writes: >In article <9009181549.AA29380@zia.aoc.nrao.edu>, >cflatter@ZIA.AOC.NRAO.EDU (Chris Flatters) writes: >|> >|> As a matter of morbid curiosity, has OSF ever made the reasoning that >lead >|> to their choice of a combination of elements from DEC, HP and >Microsoft >|> public? > >Sure; a Rationale is an essential part of our process. The User >Environment Component rationale is quite a thick document, and >certainly went to all Members. I can't find my copy right now. Kee? What? You want me to scan in it and post it? :-) Seriously though (to cflatter), what question did you have about the rationale? About the technical reasoning or the political poppycock? -kee -- Alphalpha Software, Inc. | motif-request@alphalpha.com nazgul@alphalpha.com |----------------------------------- 617/646-7703 (voice/fax) | Proline BBS: 617/641-3722 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.
marbru@auto-trol.UUCP (Martin Brunecky) (09/20/90)
In article <9009191654.AA05730@spiff> brett@SPIFF.DEN.MMC.COM (Brett Ballantyne) writes: >I've read several places that >OPEN LOOK is technically superior to Motif (a debatable issue to be sure, >but that seems to be the general consensus -- at least in the trade >rags I read) .... It may sound like I want to start yet another war. But I'v seen statements like the one above before - but NEVER without any justification. I can think of many reasons which could be used to justify such statements: - toolkit features: "pushpins" versus "keyboard traversal", "drag and drop" versus "UIL" (comparing incompatible features is purely intentional-). - style guide consistency and completness: how well it covers needs of wide range of applications, how it deals with new/unknown/undefined issues - toolkit efficiency: memory usage, number of server requests to bring up a UI component, popup appearance speed (and any other areas where toolkits can "cheat" to get better than "the other one"-) - ease of programming: availability of pre-configured objects, UI definition tools such as (but better not-) UIL - upwards compatibility between multiple releases - toolkit "openness", i.e. how much is the toolkit suitable for add-ons, enhancements and customization - toolkit code quality/maturity: how many bugs, how serious - toolkit support quality: how long does it take to fix a problem or add a desirable feature I am sure that my list is not complete. But just what I have liste above makes me feel that anyone who claims ANY toolkit SUPERIOR to any other one, should either choke and die, or better have an aggregated experience of a large software house behind such a statement. Note, I won't buy a judgement of an "independent" press even if supported by "laboratory evaluation" - toolkit is not a disk drive that you can take, shake and measure. You'v got to live with it, at least for a while. So. Can anyone come up with really justified statement of SUPERIORITY, or should we just burry such statements and get going on real issues ? -- =*= Opinions presented here are solely of my own and not those of Auto-trol =*= Martin Brunecky marbru@auto-trol.COM (303) 252-2499 {...}ncar!ico!auto-trol!marbru Auto-trol Technology Corp. 12500 North Washington St., Denver, CO 80241-2404
chan@hpfcmgw.HP.COM (Chan Benson) (09/20/90)
>> About one month after OPEN LOOK was announced to be the user interface >> for System 5 Release 4, the OSF began an RFT (Request for Technology) >> process for what is now called OSF/Motif. OPEN LOOK was submitted by >> AT&T to the OSF as a candidate for the graphical user interface >> specification and style guide. About one to two months later, OSF >> announced the Motif decision. Motif is a combination of technologies >> from DEC, HP, and Microsoft. > >As a matter of morbid curiosity, has OSF ever made the reasoning that lead >to their choice of a combination of elements from DEC, HP and Microsoft >public? Well I don't have the rationale handy (so hit me with a stick if I'm wrong), but I think one of the reasons Open Look was not chosen was because AT&T only offered the spec and a style guide, not an sample implementation. -- Chan
nazgul@alphalpha.com (Kee Hinckley) (09/20/90)
In article <9009191654.AA05730@spiff> brett@SPIFF.DEN.MMC.COM (Brett Ballantyne) writes: >Not to start another Motif vs. OPEN LOOK holy war, but how thick does >a document have to be to state they adopted a GUI L&F submitted by >those companies who give them money? I've read several places that *FLAME ON* No holy war here, this is personal. I get really pissed when people accuse me of dishonesty. You didn't mean it personally I'm sure, but guess what, there were real people on the Motif selection team. Perhaps you should ask Bob Scheifler, he was there as an observer throughout the process. He even refused to be paid for his consulting (unless you count the chocolate cake :-). >OPEN LOOK is technically superior to Motif (a debatable issue to be sure, *Which* Open Look? I've seen nothing detailing which L&F is technically superior, so you must be talking implementation (I haven't seen any papers on that either, but I'd like to, do you have references?). The only implementation of XView submitted was Xt+. We went over the design and implementation - including looking at the code, of every submission that made the first cut (yes, Xt+ did). Of the toolkits submitted, I believe we chose the best implementation. There are things I might change in how the code was merged, there are things that I think went wrong (how did UIL become part of the AES!), but we took the submissions we had and we made a *technical* decision. Sure, it jibed with the political ones, and if it hadn't I have no idea who would have won, maybe the politics would have won out - but it didn't come to that. The closest thing we made to a "political" decision was to go with C and the Intrinsics. I'm not 100% comfortable with either of those decisions, but using C++ would have limited market acceptance, and the non-Intrinsics options weren't capable enough so I think it was probably the right way to go. So, maybe Xt+ is better now, I don't know. Maybe XView is better too, but Sun declined to submit it, so it wasn't an option. That's hardly OSF's fault. >but that seems to be the general consensus -- at least in the trade >rags I read), so why would they choose Motif? Simple. IBM, DEC, HP, etc., >give OSF money and Sun and AT&T don't. I'm not saying that's wrong, >but to claim to have industry's best interest at heart and not just >OSF member's interest at heart is misrepresentation and *that* is wrong. >If I'm mistaken, please set me straight, but that's the way the world >appears from the window I look out of. What's your address? I'll send you a bottle of Windex. Here. I'll grant you something. It's no accident that OSF chose submissions from its members. You're absolutely right. Consider the odds. There are two of Sun and AT&T, on the other side you have submissions from Apollo, HP, DEC, IBM and dozens of other companies, many OSF members. All things being equal, of course the OSF members stand a better chance - pure statistics. I'll tell you something else you didn't even notice, and it's much more real then your political bullshit and much more of a problem. All of the selections were from large companies. Why? Because small companies can't afford to give up their software to OSF at a loss. All that development effort, all the potential market advantages - all gone for some minimal royalty. That's something people forget when they criticize OSF for charging for software. (Personally I think OSF can get funding for the stuff from the member companies, but it still wouldn't be enough to make it viable for a small company.) But this "us" vs. "them" crap is nonsense. If you don't like what OSF is doing, then tell them - God knows I do. *FLAME OFF* -kee I speak for myself; always have, always will. -- Alphalpha Software, Inc. | motif-request@alphalpha.com nazgul@alphalpha.com |----------------------------------- 617/646-7703 (voice/fax) | Proline BBS: 617/641-3722 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.
rhoward@msd.gatech.edu (Robert L. Howard) (09/21/90)
In <1990Sep20.165120.14156@alphalpha.com> nazgul@alphalpha.com (Kee Hinckley) writes: >In article <9009191654.AA05730@spiff> brett@SPIFF.DEN.MMC.COM (Brett Ballantyne) writes: >>Not to start another Motif vs. OPEN LOOK holy war... >*FLAME ON* >No holy war here, this is personal... Perhaps this is why you are a little less than rational... [[ BTW, pick a window size that fits on most (80 col) screens ]] >>OPEN LOOK is technically superior to Motif (a debatable issue to be sure, >*Which* Open Look? There is only one Open Look spec. It involves a functional spec and a style guide. > I've seen nothing detailing which L&F is technically superior, ^^^ >so you must be talking implementation(I haven't seen any papers on that either, ^^^^^^^^^^^^^^ ^^^^^^ >but I'd like to, do you have references?). From a *user* standpoint it's the L&F that matter (along with consistency of the applications that 'comply' to that L&F). I understand developers wanting a good toolkit, but for the users the L&F is what counts. I don't think any developer is going to be able to say to his customers "Hey, we chose the best toolkit to enhance our productivity. We got the app to market faster that way." Well, this is a nice argument considering that if the app is long getting to market nobody benefits but that won't win you any long term customers. They ultimately want the 'best' product. As for papers, if you count the trade rags, I've seen lots of opinions flying. Lots of them do refer to the L&F. For example, pro-OL people argue that the OL spec is a lot 'tighter' thus ensuring that 'compliant' applications will be 'more consistent'. > The only implementation of XView submitted ^^^^^ >was Xt+. I am sure you meant OpenLook (as Xview is also a toolkit). > We went over the design and implementation - including looking at the code, >of every submission that made the first cut (yes, Xt+ did). Of the toolkits submitted, >I believe we chose the best implementation. This is what I feel is the flaw in the OSF reasoning. Sure, I can understand the need to take care of their members needs, but I thought OSF was formed to created standards that would help us users. The real end users (assuming UNIX/workstation vendors want to penetrate the business markets) aren't going to care about programming issues (toolkits, implementations, etc.) but rather about how the resulting apps are going to make their live better. This is not to say that OSF made a bad choice in Motif (I can think of a lot of good end user reasons to pick it) but I think justifying it with the implementation arguments is questionable at best. > The closest >thing we made to a "political" decision was to go with C and the Intrinsics. I'm not >100% comfortable with either of those decisions, but using C++ would have limited >market acceptance, and the non-Intrinsics options weren't capable enough so I think >it was probably the right way to go. Again more programmer issues rather than end user issues. Now, I know someone is going to say, "Hey! If we ignored the developer we wouldn't have had a chance at the end user because he would have never gotten an application on his desk." That's a good point but I'm not sure I can exactly buy it. After all there have been MS Windows applications for along time know (because the market demanded them) yet from what I here it was one of the most hoorid things to have to write code for. >*FLAME OFF* Thank goodness... Robert -- | Robert L. Howard | Georgia Tech Research Institute | | rhoward@msd.gatech.edu | MATD Laboratory | | (404) 528-7165 | Atlanta, Georgia 30332 | | UUCP: ...!{allegra,amd,hplabs,ut-ngp}!gatech!msd!rhoward |
sarima@tdatirv.UUCP (Stanley Friesen) (09/22/90)
In article <833@auto-trol.UUCP> marbru@auto-trol.UUCP (Martin Brunecky) writes: >In article <9009191654.AA05730@spiff> brett@SPIFF.DEN.MMC.COM (Brett Ballantyne) writes: >>I've read several places that >>OPEN LOOK is technically superior to Motif (a debatable issue to be sure, >>but that seems to be the general consensus -- at least in the trade >>rags I read) .... > I can think of many reasons which could be used to justify such > statements: > - toolkit features: "pushpins" versus "keyboard traversal", "drag and drop" > versus "UIL" (comparing incompatible features is purely intentional-). > - style guide consistency and completness: how well it covers needs of > wide range of applications, how it deals with new/unknown/undefined > issues [long list of toolkit level features deleted] How about adding: - GUI internal consistancy and ease of learning (e.g. intuitiveness, simplicity of operation, ease of description). Really, since we are (supposedly) comparing User Interfaces, then technical superiority should be judged on the basis of *user* interaction, not programming ease, implementation quality, or documentaion quality. (For one thing multiple implementations are possible of the same GUI - for instance there are *four* Open Look toolkits, OLIT, XView, NeWS, and SunView Open Look) In my experience the Open Look interface is much more intuitive, and much more consistant than the Motif interface. For instance, the designated MENU button on the mouse *always* brings up a menu under Open Look. And it does *nothing* else. (This is usually the right button). It is this simplicity and consistancy that makes Open Look superior.
cflatter@ZIA.AOC.NRAO.EDU (Chris Flatters) (09/22/90)
>>XView's foundation is similar to Xt in that they both provide a way to >>do subclassing in C. However, the implementation is significantly >Is that foundation availble to non-toolkit writers who wish to add a >"widget"? The requirements for developing an XView package (which corresponds to an object class) is documented --- see chapter 22 of Dan Heller's "XView Programming Manual". However you would write packages as a means of extending the toolkit and not as a matter of routine application development. XView programming is very different to programming using an Xt based widget set. This is probably not a bad thing. I think that you can start producing useful XView applications without the same level of investment in learning the toolkit than is required for Xt plus widgets -- the API is far less complex (this is, of course, a subjective judgement and should be treated as such). Chris Flatters
marbru@auto-trol.UUCP (Martin Brunecky) (09/22/90)
In article <rhoward.653919896@romeo> rhoward@msd.gatech.edu (Robert L. Howard) writes: > >From a *user* standpoint it's the L&F that matter (along with consistency >of the applications that 'comply' to that L&F). I understand developers >wanting a good toolkit, but for the users the L&F is what counts. I don't >think any developer is going to be able to say to his customers "Hey, we >chose the best toolkit to enhance our productivity. We got the app to >market faster that way." Well, this is a nice argument considering that >if the app is long getting to market nobody benefits but that won't win >you any long term customers. They ultimately want the 'best' product. > Thanx for finally trying to define how we measure the toolkit SUPRIORITY. From your posting (not entirely replicated here, sorry), it seems that the SPECIFICATION is what matters. Thoug right obove you say the 'best' product. Now, if you allow me to disagree (flame on?). I have the OL Specification on my bookshelf (I even went through it...). I also have OSF/Motif style guide there, along with it's AES So I have 3 documents: OL: Specification telling the TOOLKIT developer how the individual element must look, feel etc. Motif Style Guide: Specification teling me what CONCEPTS should the TOOLKIT and APPLICATION programmer follow, and what controls should be used (feel) for example component classes (not which classes should be implemented and how). Motif AES: Specification describing OSF's implementation of Motif Style Guide, i.e. programmer's reference manual. I am finding it rather difficult to judge OL versus Motif based on the documents above. Sure, It's easier to implement Open Look compliant toolkit, as everything is defined (just go and code). On the other hand, Motif leaves me more free to implement my objects, for my specific needs. Is that bad ? I don't think so. So, if I would accept your approach of judging toolkits by their specifications, I find it difficult to pick a winner. Now, when it comes to real judgement, I agree that the customer is the judge (though press would like to jump in). The customer ultimatly wants the 'best' product. And for that matter, I would let the customer decide, over the period of few years, what is the 'best' product, and quit any statements about superiority for now. Besides. It is IBM PC (and clones) with MS-DOS that the customer found the 'best' - and buys it, and buys it. Has this ever been a SUPERIOR technology ? -- =*= Opinions presented here are solely of my own and not those of Auto-trol =*= Martin Brunecky marbru@auto-trol.COM (303) 252-2499 {...}ncar!ico!auto-trol!marbru Auto-trol Technology Corp. 12500 North Washington St., Denver, CO 80241-2404
nazgul@alphalpha.com (Kee Hinckley) (09/22/90)
In article <1210046@hpfcmgw.HP.COM> chan@hpfcmgw.HP.COM (Chan Benson) writes: >Well I don't have the rationale handy (so hit me with a stick if I'm >wrong), but I think one of the reasons Open Look was not chosen was because >AT&T only offered the spec and a style guide, not an sample implementation. No. -- Alphalpha Software, Inc. | motif-request@alphalpha.com nazgul@alphalpha.com |----------------------------------- 617/646-7703 (voice/fax) | Proline BBS: 617/641-3722 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.
nazgul@alphalpha.com (Kee Hinckley) (09/22/90)
In article <rhoward.653919896@romeo> rhoward@msd.gatech.edu (Robert L. Howard) writes: >[[ BTW, pick a window size that fits on most (80 col) screens ]] Sorry, I post mail in a variable width font, and auto-wrap isn't scheduled until release two of the mail product :-). Next time. You then follow with a series of arguments as to why we should have looked at it from the user standpoint, not the toolkit. I can assure you we did both. I was answering what I thought was a toolkit discussion. The rationale document deals with both sides of the issue. I urge you to contact OSF and see if you can get a copy. -kee -- Alphalpha Software, Inc. | motif-request@alphalpha.com nazgul@alphalpha.com |----------------------------------- 617/646-7703 (voice/fax) | Proline BBS: 617/641-3722 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.
jacobi@jooby.Eng.Sun.COM (Jennifer Jacobi) (09/26/90)
In article <1990Sep20.165120.14156@alphalpha.com> nazgul@alphalpha.com (Kee Hinckley) writes: >but I'd like to, do you have references?). The only implementation of XView submitted >was Xt+. Huh??? Can you please describe what you mean here. XView and Xt+ are two completely different toolkits that implement OPEN LOOK L&F. XView is a Xlib based toolkit for porting sun view apps, and Xt+ is an X Intrinsic based toolkit. So how can an implementation of XView be Xt+? jacobi@eng.sun.com ----- my comments are mine -- only mine.
fgreco@govt.shearson.COM (Frank Greco) (09/27/90)
>I'll be interested to read an official OSF statement myself. In the >meantime, here's my personal opinion of why OL didn't make it. Huh? Please don't tell all the companies on Wall Street, cuz the vast majority of them are using Open Look (on Suns), not Motif. Sun's (actually SPARC's) large percentage of the workstation market will cause OPEN LOOK to have a higher percentage of the workstation UI "market". It doesn't matter a bit what OSF does. The market *always* decides the standard. Why don't people learn from history? Recall the Microsoft BASIC vs "real" BASIC battle just a few years ago. Despite the standardness of real BASIC, Microsoft won the marketing and numbers game. OPEN LOOK vs Motif will result in a similar decision. > >It seems quite reasonable to conclude: > - Microsoft compatible GUI's are going to be much more acceptable > to the Commercial/Business folks that some new unknown > interface. Yow! What's your rationale for this? I don't agree at all. There are more Microsoft GUI's (what's a Microsoft compatible GUI??) than X-based GUI's because there are many more PC's out there than workstations. Over time, this will change. I would surmise at around 1996 or so, workstation shipments will match PC shipments... this is only an intuitive guessimation on my part based on current shipment data. > >Once we have come to this conclusion, we notice: ^^ Whatta you mean "we" ? I believe you mean "I". > - OL is not Microsoft compatible, and to make it so destroys its > open-lookness. First of all, OPEN LOOK can be made to run on DOS. It is a user interface, not software. I believe there are two companies out there that are developing an OL interface for DOS and/or OS/2 (there was some blurb in InfoRag or PC Weak about this recently). So OL can be implemented on a Microsoft platform; hence it can be "Microsoft compatible". > > - HP's CXI was nearly compatible, and DEC's could be force fitted > to the HP CXI model. That sounds very elegant ;-) > >The decision to go with a combined HP/DEC offering rather than an OL >offering becomes obvious. Blech! You weren't on the Debating Team at your school were you? ;-) Frank D. Greco - Consultant +-----------------------------------+-----------------------------------+ |On Assignment at: |Office: | | email: fgreco@shearson.com | email: frank5@mars.njit.edu | +-----------------------------------+-----------------------------------+ My comments reflect my own opinions, not my clients.
fgreco@govt.shearson.COM (Frank Greco) (09/28/90)
>>XView's foundation is similar to Xt in that they both provide a way to >>do subclassing in C. However, the implementation is significantly >Is that foundation availble to non-toolkit writers who wish to add a >"widget"? Yes. There is a document floating around that clearly details the manner in which a programmer can add your own XView objects. If you wish, contact me via email and I'll send you a copy (in Postscript) that I obtained off the net sometime ago. >>Don't know. Also, all the SunView OPEN LOOK applications (I can only think >>of four) are being ported to X11/OPEN LOOK, so it's a moot point. However, >>the real reason for the large number of applications for OPEN LOOK is the >>volume of Sun and Sun-compatible machines already on the market and the >This doesn't necessarily follow. As others have pointed out, there are a number >of applications shipping (even as their first machine!) on Sun machines >that use Motif rather than Open Look. Yes, that's true, however on the Sun workstation, the OL apps far outweigh the Motif ones... And as long as Sun does not formally support Motif, this will continue to be so. When big corporations commit to a workstation vendor, they like the warm fuzzies they get from *official* Sun support. Frank Greco My opinions are mine, not my clients.
fgreco@govt.shearson.COM (Frank Greco) (09/28/90)
> > Besides. It is IBM PC (and clones) with MS-DOS that the customer found > the 'best' - and buys it, and buys it. Has this ever been a SUPERIOR > technology ? >-- Precisely! It was the vast numbers of PC's and a bundled (low-tech) MS-DOS (practically every vendor bundled MS-DOS with each PC sold) that dictated the "standard" operating system. That is exactly why OPEN LOOK will prevail, regardless of any technical superiority/non-superiority, source code availability/non-availability, Xt-based/non-Xt-based matter. The amount of Sun (actually SPARC is more appropriate) workstations in the commercial marketplace is growing incredibly fast (in the Wall Street arena, there *is* no other workstation). Guess which UI is standard on those machines? Hint: Its not Motif. Frank Greco fgreco@shearson.com Comments are mine, not my clients
rlh2@ukc.ac.uk (Richard Hesketh) (09/28/90)
In article <9009272035.AA09060@islanders.> fgreco@govt.shearson.COM (Frank Greco) writes: [more OL vs Motif] > incredibly fast (in the Wall Street arena, there *is* no other workstation). > Guess which UI is standard on those machines? Hint: Its not Motif. Strange. Then why have I just read an article posted in comp.windows.x.motif concerning a Motif UI builder sent by a guy at "Citibank, 111 Wall Street" ?
jordan@morgan.COM (Jordan Hayes) (09/29/90)
Frank D. Greco <fgreco@govt.shearson.COM> writes: Please don't tell all the companies on Wall Street, cuz the vast majority of them are using Open Look (on Suns), not Motif. The guy was talking about why OpenLook didn't get selected by OSF as the technology they would back, not why "OpenLook" didn't have market penetration (which wouldn't make sense anyway). [ note: I don't necessarily agree, but that's what he said ] What is your sample size for "vast majority [of Wall Street]" ...? I didn't think the "vast majority" of anybody is doing anything (except maybe flaming about this issue :-). Sun's (actually SPARC's) large percentage of the workstation market will cause OPEN LOOK to have a higher percentage of the workstation UI "market". I think this is naive. what's a Microsoft compatible GUI?? Uh, Motif is (supposed to be) PM +/- 7 things. GUI, of course, not API. > - OL is not Microsoft compatible, and to make it so > destroys its open-lookness. First of all, OPEN LOOK can be made to run on DOS. It is a user interface, not software. Duh? When this guy says "Microsoft Compatable" he's talking about the GUI layer, not the API layer for XView ... he means Joe Hodedo could transition from PM to Motif in a "compatable" way -- making Open Look applications that would feel the same way would make them no longer Open Look. For the purposes of this thread, "Microsoft" == "PM" or "Windows" "Microsoft" != "DOS" or "PCs" My comments reflect my own opinions, not my clients. Hope so! ;-) /jordan
prc@erbe.se (Robert Claeson) (09/29/90)
In a recent article rlh2@ukc.ac.uk (Richard Hesketh) writes: >In article <9009272035.AA09060@islanders.> fgreco@govt.shearson.COM (Frank Greco) writes: > [more OL vs Motif] >> incredibly fast (in the Wall Street arena, there *is* no other workstation). >> Guess which UI is standard on those machines? Hint: Its not Motif. >Strange. Then why have I just read an article posted in comp.windows.x.motif >concerning a Motif UI builder sent by a guy at "Citibank, 111 Wall Street" ? Also, remember that using a Sun don't have to mean using OpenLook. Many large Sun sites are using Motif on their systems. -- Robert Claeson |Reasonable mailers: rclaeson@erbe.se ERBE DATA AB | Dumb mailers: rclaeson%erbe.se@sunet.se | Perverse mailers: rclaeson%erbe.se@encore.com These opinions reflect my personal views and not those of my employer.
fgreco@donald.GOVt.shearson.COM (Frank Greco) (10/01/90)
> > incredibly fast (in the Wall Street arena, there *is* no other workstation). > > Guess which UI is standard on those machines? Hint: Its not Motif. > > Strange. Then why have I just read an article posted in comp.windows.x.motif > concerning a Motif UI builder sent by a guy at "Citibank, 111 Wall Street" ? > When I said "in the Wall Street arena, there *is* no other workstation", I meant "In the Wall Street workstation market, the overwhelming majority of workstations bought is SPARC-based (mostly Suns)". Of course there are DEC 3100's, IBM 6000's, HP whatevers..., but Sun has the largest percentage by far. As to your Citibank reference, I only have this to say: I have a PC. I have Unix running on my PC. Does that make Unix the operating system of choice for all PC's? Of course not. I also have a 200 watt amp in my Toyota. Does that mean most Toyota owners like to hear ZZ Top at 130 decibels? Of course not. As usual, the net is somewhat deficient when it comes to clearly indicating hyperbole in textual form. I should've looked up the correct "smiley face" to use in this situation. Frank G. again, my comments are mine, not my (human) clients.
fgreco@donald.GOVt.shearson.COM (Frank Greco) (10/01/90)
> > What is your sample size for "vast majority [of Wall Street]" ...? I > didn't think the "vast majority" of anybody is doing anything (except > maybe flaming about this issue :-). Gees, I really don't agree Jordan. I've been involved in over a dozen high-level meetings for different clients of mine within the past 2 months in which the primary discussion has been which GUI to use. These were definitely *not* informal hacker tech-talks. --------- > > Sun's (actually SPARC's) large percentage of the workstation > market will cause OPEN LOOK to have a higher percentage of the > workstation UI "market". > > I think this is naive. I think your reply is incomplete.;->..... Well...what's your reason? cat -v got your tongue? I'm basing my "naive" opinions on the fact that if OL comes bundled with the machine, it will be a hard sell (of course, not impossible) to have corporate users utilize, and sys admins install, a non-supported (by Sun) user interface. ---------- > > what's a Microsoft compatible GUI?? > > Uh, Motif is (supposed to be) PM +/- 7 things. GUI, of course, not API. ^^^^^^^^^^^^^^^^^^^^^^^ I'm glad *you*'re stating your case explicitly. The author wasn't. ---------- > First of all, OPEN LOOK can be made to run on DOS. It is a > user interface, not software. > > Duh? When this guy says "Microsoft Compatable" he's talking about the > GUI layer, not the API layer for XView . Well perhaps that how *you* interpreted his email. I read it differently sir. And what does "Duh?" mean big guy?.......btw, its "compatible" not "Compatable" dude.... ---------- > For the purposes of this thread, > > "Microsoft" == "PM" or "Windows" > "Microsoft" != "DOS" or "PCs" Your definitions Jordan? Or the author's? Are you submitting your interpretation as a replacement for his? ---------- > > My comments reflect my own opinions, not my clients. > > Hope so! ;-) Oh, how clever....;-> Frank G. ...my comments still reflect my own opinions, not my clients...
marbru@auto-trol.UUCP (Martin Brunecky) (10/03/90)
In article <9010011558.AA13076@islanders.> fgreco@donald.GOVt.shearson.COM (Frank Greco) writes: >> > And what does "Duh?" mean big guy?.......btw, its "compatible" not > "Compatable" dude.... > Frank G. > ...my comments still reflect my own opinions, not my clients... O.K. I am now convinced. Wall Street is going to use SPARSstations and Open Look. It's better and it's blessed by Sun. But can't we (I man the rest of us not picking up cream on Wall Street) have an "OPEN" DISCUSSION (without violating Sun Microsystem trademark-) about technical issues, without throwing mud on "the other dude" ? -- =*= Opinions presented here are solely of my own and not those of Auto-trol =*= Martin Brunecky marbru@auto-trol.COM (303) 252-2499 {...}ncar!ico!auto-trol!marbru Auto-trol Technology Corp. 12500 North Washington St., Denver, CO 80241-2404
ldm@texhrc.UUCP (Lyle Meier) (10/05/90)
In article <9009272035.AA09060@islanders.>, fgreco@govt.shearson.COM (Frank Greco) writes: > The amount of Sun (actually SPARC is more > appropriate) workstations in the commercial marketplace is growing > incredibly fast (in the Wall Street arena, there *is* no other workstation). > Guess which UI is standard on those machines? Hint: Its not Motif. What this shows is that different industries are adapting different toolkits. At the last Society of Exploration Geophysicists show, MOTIF lead by at least 10 to 1 over open look in the number of displays. Since many Oil company's are heavily into PM and Windows, the path there where workstations will always occupy a niche, will be easiest with open look. What will likely happen is that different industries will adopt different GUI's depending upon what their needs and requirements are
graham@fuel.dec.com (kris graham) (10/16/90)
>> The amount of Sun (actually SPARC is more >> appropriate) workstations in the commercial marketplace is growing > > incredibly fast (in the Wall Street arena, there *is* no other workstation). > > Guess which UI is standard on those machines? Hint: Its not Motif. > What this shows is that different industries are adapting different > toolkits. At the last Society of Exploration Geophysicists show, > MOTIF lead by at least 10 to 1 over open look in the number of > displays Also, at Wall Street's most popular financial/commercial show, Motif trounced Open Look by a 10 to 1 margin! I dunno where Frank Greco gets his data from. I will look again at the October 31 UNIX EXPO here in New York City...another big show. Christopher Graham Digital Equipment Corp Ultrix Resource Center New York City Internet: graham@fuel.enet.dec.com UUCP: ...!decwrl!fuel.enet.dec.com!graham
don@zardoz.coral.COM (Don Dewar) (10/16/90)
) Return-Path: <uunet!expo.lcs.mit.edu!xpert-mailer> ) Date: 15 Oct 90 22:41:09 GMT ) From: uunet!decwrl.dec.com!bacchus.pa.dec.com!deccrl!shlump.nac.dec.com!riscy.enet.dec.com!fuel.dec.com!graham (kris graham) ) Subject: Re: What's what in OPEN LOOK/OpenWindows (long) ) References: <507@texhrc.UUCP>, <9009272035.AA09060@islanders.> ) Sender: uunet!expo.lcs.mit.edu!xpert-request ) To: xpert@expo.lcs.mit.edu ) ) >> The amount of Sun (actually SPARC is more ) >> appropriate) workstations in the commercial marketplace is growing ) > > incredibly fast (in the Wall Street arena, there *is* no other ) workstation). ) > > Guess which UI is standard on those machines? Hint: Its not Motif. ) ) > What this shows is that different industries are adapting different ) > toolkits. At the last Society of Exploration Geophysicists show, ) > MOTIF lead by at least 10 to 1 over open look in the number of ) > displays ) ) Also, at Wall Street's most popular financial/commercial show, Motif trounced ) Open Look by a 10 to 1 margin! I dunno where Frank Greco gets his data ) from. ) ) I will look again at the October 31 UNIX EXPO here in New York City...another ) big show. ) ) Christopher Graham ) Digital Equipment Corp ) Ultrix Resource Center ) New York City ) ) Internet: graham@fuel.enet.dec.com ) UUCP: ...!decwrl!fuel.enet.dec.com!graham ) ) As an addendum to Christopher's comment, the recent Interop '90 also seemed overwhelmingly to favor Motif, even on SPARCstations. Of those network management programs written on UNIX platforms, I heard of far more using Motif than OpenLook. In addition, the SPARC was also a common sight at many of those same booths. To me the muddied message is becoming clearer. The UI software battle is not going to be decided by any one hardware's predominance in the marketplace. So please stop using the lame argument that Sun's large market share is going to make OpenLook the defacto standard. Go back to the argument on technical merit -- it was far more interesting. Let me couch my last request so there are no misunderstandings. I believe it will be the market that sets the defacto standard UI in the UNIX arena. But it will be the software marketplace, not the hardware marketplace that renders the final decision. Everyone who is trying to make predictions at this point is still trying to figure out who is going to win the superbowl on the first day of the season. +---------+ | Coral | |@@@@@*@**| |@@*@@**@@| Don Dewar |*@@**@@@@| Coral Network Corporation, Marlborough, MA |@***@@@@@| Internet: don@coral.com |@@**@@@@@| Phone: (508) 460-6010 |*********| Fax: (508) 481-6258 |Networks | +---------+
fgreco@dprg-330.GOVt.shearson.COM (Frank Greco) (10/16/90)
> > >> The amount of Sun (actually SPARC is more > >> appropriate) workstations in the commercial marketplace is growing > >> incredibly fast (in the Wall Street arena, there *is* no other workstation). > >> Guess which UI is standard on those machines? Hint: Its not Motif. > > > What this shows is that different industries are adapting different > > toolkits. At the last Society of Exploration Geophysicists show, > > Also, at Wall Street's most popular financial/commercial show, Motif trounced > Open Look by a 10 to 1 margin! Yeah, I agree. There were many more people saying Motif is inferior to OPEN LOOK...;-) Seriously, I get my data by working in the Wall Street Financial sector for the past 5 years. I've only seen *2* DEC Unix workstations in that time; plenty of VMS machines but only *2* DEC workstations running Unix (no one runs VMS on a real workstation right Chris?). And those DEC workstations were evaluation machines. Heck, I've seen more IBM RIOS machines than DEC boxes. Of course, *these* machines were running Motif. I've even seen a few Sun's and Solbournes running Motif. But overwhelmingly, the standard GUI is OPEN LOOK. Its not just opinion, but observation. Geesus, what financial/commercial show were you at Chris? I have several press releases (ie, not personal opinion from yours truly but facts from the software vendors) from "Wall Street's most popular financial/commercial show" (ugh...gimme a break) that indicate that Open Look, not Motif, is the GUI of choice for the financial sector. It might not be the choice for extremely PC-minded institutions (I don't mean that pejoratively), but OPEN LOOK seems to be growing faster than its competitor in the financial arena. Being in DEC marketing support, you wouldn't be biased would you ? ;->> > I will look again at the October 31 UNIX EXPO here in New York City...another > big show. Be There and Be Square...;-) BTW, did you see the article in today's *Digital Review* regarding multimedia in the 90's? The photo was that of X/NeWS and OPEN LOOK. There also was a picture of an OPEN LOOK screen on the inside cover too. > > Christopher Graham > Digital Equipment Corp > New York City Frank G. [Put your favorite standard opinion disclaimer here]
cook@sgi.com (Doug Cook) (10/17/90)
In article <9010161109.AA02249@zardoz.noname> don@zardoz.coral.COM (Don Dewar) writes: >Let me couch my last request so there are no misunderstandings. I >believe it will be the market that sets the defacto standard UI >in the UNIX arena. But it will be the software marketplace, not the >hardware marketplace that renders the final decision. Everyone who is >trying to make predictions at this point is still trying to figure >out who is going to win the superbowl on the first day of the season. I disagree. Motif was being pushed as "the standard" well before it was even released; there were people committed to use it practically before they saw it. There's far more politics than market decision going on here. If it is a Super Bowl, it's a Super Bowl of lemmings; you *can* predict what will happen to all of them from observing the first one jump off the cliff. Make no mistake; I'm not saying anything at all about which toolkit I think people should be using. I just don't think that the "market" is what is deciding things, at least to a large degree. -Doug Doug Cook | My opinions do not necessarily Video Group, Advanced Systems Division | reflect those of my employer. Silicon Graphics, Inc. | Mountain View, CA |
marbru@auto-trol.UUCP (Martin Brunecky) (10/18/90)
In article <1990Oct16.185848.2617@odin.corp.sgi.com> cook@sgi.com (Doug Cook) writes: >In article <9010161109.AA02249@zardoz.noname> don@zardoz.coral.COM (Don Dewar) writes: >>Let me couch my last request so there are no misunderstandings. I >>believe it will be the market that sets the defacto standard UI >>in the UNIX arena. But it will be the software marketplace, not the ..... (rest deleted). I used to think that this newsgroup is about X, but it seems that we are engaged in predicting the future. I did some meteorology in my times ... predicting is difficult... especially the future. What about getting back to (technical) grounds, and really look what's what in Open Look and what's what in Motif. To me, the fact that one or the other is pushed by a company I personally dislike should not matter. What matters are real features, differences, advantages of one or the other. If I am allowed to express my opinion (without being stoned here), the evolution should bring both UIs close to each other, not further apart. Taking stands like "over my dead body" and "we have the winner" may be makes some business impact, but gives very little help to the user. And certainly does not stimulate evolution and inovation. So why not discuss what are the MAJOR incompatibilities of both GUI, and IF thay can be overcome (or WHY not) - if there is a merging path. -- =*= Opinions presented here are solely of my own and not those of Auto-trol =*= Martin Brunecky [BORN TO BASH UIL] marbru@auto-trol.COM (303) 252-2499 {...}ncar!ico!auto-trol!marbru Auto-trol Technology Corp. 12500 North Washington St., Denver, CO 80241-2404
dgh@Unify.Com (David Harrington) (10/19/90)
In article <1990Oct16.185848.2617@odin.corp.sgi.com> cook@sgi.com (Doug Cook) writes: >In article <9010161109.AA02249@zardoz.noname> don@zardoz.coral.COM (Don Dewar) writes: >>Let me couch my last request so there are no misunderstandings. I >>believe it will be the market that sets the defacto standard UI >>in the UNIX arena. But it will be the software marketplace, not the >>hardware marketplace that renders the final decision. Everyone who is >>trying to make predictions at this point is still trying to figure >>out who is going to win the superbowl on the first day of the season. > >I disagree. Motif was being pushed as "the standard" well before it >was even released; there were people committed to use it practically before >they saw it. There's far more politics than market decision going on here. >If it is a Super Bowl, it's a Super Bowl of lemmings; you *can* predict what >will happen to all of them from observing the first one jump off the cliff. I agree. I have always thought that the main two reasons you see workstation vendors like DG, Concurrent, Motorola, etc (*not* your biggies!) choose Motif are: 1. Motif is easily licensed from OSF, *and* it's cheap. (Isn't that what OSF stands for? Our Software's Free? Aren't they in business to provide cheap software for the iron pushers? Have I started another (flame) thread here?) 2. Motif "style" is more flexible than Open Look, thus allowing the hardware vendors to offer, to a certain extent, their own unique look and feel using a "standard" toolkit. > >Make no mistake; I'm not saying anything at all about which toolkit I >think people should be using. I just don't think that the "market" is what >is deciding things, at least to a large degree. > > -Doug > >Doug Cook | My opinions do not necessarily >Video Group, Advanced Systems Division | reflect those of my employer. >Silicon Graphics, Inc. | >Mountain View, CA | -- David Harrington internet: dgh@eire.unify.COM Unify Corporation ...!{csusac,pyramid}!unify!eire!dgh 3870 Rosin Court voice: (916) 920-9092 Sacramento, CA 95834 fax: (916) 921-5340
cflatter@ZIA.AOC.NRAO.EDU (Chris Flatters) (10/22/90)
Dave Harrington writes: > 2. Motif "style" is more flexible than Open Look, thus allowing the hardware > vendors to offer, to a certain extent, their own unique look and feel using a > "standard" toolkit. This seems contrary to the aims of OSF/Motif: "Applications must maintain a consistent look and feel on all platforms to enable the easy transfer of skills learned on one system to any other in the network."[%] That said the Motif style guide is relatively flexible (read loose) and software vendors do seem to like pushing at the edges of that envelope of flexibility. Chris Flatters
moraes@cs.toronto.edu (Mark Moraes) (10/22/90)
dgh@Unify.Com (David Harrington) writes: >2. Motif "style" is more flexible than Open Look, thus allowing the hardware >vendors to offer, to a certain extent, their own unique look and feel using a >"standard" toolkit. Sigh -- has the wheel turned full circle? As I remember, the great toolkit and GUI hoo-hah started when people wanted a more uniform look and feel across different applications and different platforms from different vendors. (and different vendors wanted their own unique look and feel to be the "de facto" standard) What I've found amusing lately is the subtle implication that Motif and Open Look applications will not co-exist on the same screen, and that to program X one must either use Motif or Open Look... Mark.