gt1111a@prism.gatech.EDU (Vincent Fox) (02/28/91)
I for one am a little sick of products only written for particular windowing systems. I run a small net of about 15 machines. Plenty of stuff runs under regular MIT X. But my boss has just informed me of a few new products we are buying, one of which needs Open Windows, the other requiring Motif. Now tell me why I must go through the agony of trying to make 3 windowing packages share my disk? Can't X code be written so that it takes advantage of OL or MOTIF's features if you have it, but still run under MIT? I realize this is not as simple as say dealing with the presence or absence of an FPU, but the same principle applies. At least I can reasonably keep MIT X on all our various platforms up to date and working. Servicing 3 packages though.....maybe I should find a new job. -- Vincent Fox (That's Mr. Bucko to you)|"Fleshy-headed mutant, are you friendly?" Georgia Tech, Atlanta GA |"No way, eh! Radiation has made me an SR-71: gt1111a@prism.gatech.edu | enemy of civilization." Pony Express:...!gatech!prism!gt1111a| - Bob & Doug in Strange Brew
rthomson@mesa.dsd.es.com (Rich Thomson) (02/28/91)
In article <22979@hydra.gatech.EDU> gt1111a@prism.gatech.EDU (Vincent Fox) writes: >I for one am a little sick of products only written for particular windowing >systems. I run a small net of about 15 machines. Plenty of stuff runs under >regular MIT X. But my boss has just informed me of a few new products we are >buying, one of which needs Open Windows, the other requiring Motif. Now tell >me why I must go through the agony of trying to make 3 windowing packages >share my disk? Open Look and Motif are not "windowing systems". X windows is a windowing system, on top of which Open Look and Motif are layered. They are just application libraries. If you buy a binary package that "needs" motif, the only thing it should really need to execute are the applications defaults file, etc. It won't need the libraries, include files, etc, etc. You only need those things if you are developing code that uses the UI libraries. -- Rich -- ``Read my MIPS -- no new VAXes!!'' -- George Bush after sniffing freon Disclaimer: I speak for myself, except as noted. UUCP: ...!uunet!dsd.es.com!rthomson Rich Thomson ARPA: rthomson@dsd.es.com PEXt Programmer
warren@cbnewsh.att.com (warren.a.montgomery) (02/28/91)
Gee, it sure would be more convenient if the whole world spoke the same language, or if we all ate the same kinds of food, or if we all wore the same style of clothes, BUT NONE OF THESE THINGS WILL EVER HAPPEN, BECAUSE PEOPLE DON'T REALLY WANT THAT, SO WHY IS ALL OF THIS ENERGY BEING WASTED ON GUI WARS! The OpenLook versus Motif war is like the war between emacs and vi, or ASCII and EBCIDIC, or Fortran and Cobol, all of which are still healthy. Once users learn how to do something, they want to be able to keep doing that thing the same way. Since millions of users have already learned how to do all of the things that GUI's do in several different ways, one size will not fit all. I'm sure most users out there simply want all of their tools to be usable with the interface they already know (which is likely to be niether OpenLook nor Motif). Toolkits and libraries that let the user pick the interface seem the obvious solution, so why not spend our energies on these rather than debate which one will win. They will probably all be around a lot longer than we will. -- Warren Montgomery att!ihlpf!warren
guy@auspex.auspex.com (Guy Harris) (03/01/91)
>They are just application libraries. If you buy a binary package that >"needs" motif, the only thing it should really need to execute are the >applications defaults file, etc. It won't need the libraries, include >files, etc, etc. You only need those things if you are developing >code that uses the UI libraries. Or if the binary package was linked, on systems supporting shared libraries, with *shared* versions of the OL/Motif/whatever libraries, rather than having its very own N hundred KB copy of the libraries in its executable image....
nazgul@alphalpha.com (Kee Hinckley) (03/01/91)
In article <22979@hydra.gatech.EDU> gt1111a@prism.gatech.EDU (Vincent Fox) writes: >I for one am a little sick of products only written for particular windowing >systems. I run a small net of about 15 machines. Plenty of stuff runs under >regular MIT X. But my boss has just informed me of a few new products we are >buying, one of which needs Open Windows, the other requiring Motif. Now tell >me why I must go through the agony of trying to make 3 windowing packages >share my disk? Can't X code be written so that it takes advantage of OL or >MOTIF's features if you have it, but still run under MIT? Motif applications *do* run under both OL and Mit X. Motif doesn't not require any specific type of X server. If you've got products that require Open Windows, then run that and everything else will be fine. -- Alfalfa Software, Inc. | Poste: The EMail for Unix nazgul@alfalfa.com | Send Anything... Anywhere 617/646-7703 (voice/fax) | info@alfalfa.com 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.
raney@yertle.Colorado.EDU (Scott Raney) (03/02/91)
In response to the comment that Motif and Open Look are not windowing systems and should be able to run on vanilla X, I must point out that it may not be as simple as that. One thing that has bitten me a number of times is that both Open Look and Motif require parts that are not included in the base X. For example, Open Look requires a special glyph font to draw its rounded-rectangle buttons. If the font isn't loaded in the server, the app won't run (I've run into this several times, and usually the machines are controlled by different departments so nothing can be done about it.). Motif requires a special file to remap keysyms. Again, if you don't have it you'll have problems. With 20-20 hindsight, it seems to me the problem is due to MIT's "no policy" philosophy. In my opinion these problems, which are getting worse every day, could have been avoided had MIT not been so reluctant to take a stand. I don't really think it matters which look and feel is used, what is important that it be the same everywhere. But it's too late now. The best examples of standards are those that are de facto, and not the result of long commercial battles, or long reviewing periods by standards organizations. The take home lesson: when producing a new product, especially one that requires a new technology, don't be timid about specifying the interfaces and annointing them as the standard. Even if you make mistakes, and parts must be redone, the community of users will be better off in the long run. -- ====================================================================== Scott Raney No other person or organization raney@gabor.colorado.edu can be held responsible for my (303)499-9855 opinions or actions
grp@Unify.com (Greg Pasquariello) (03/03/91)
In article <1991Mar1.005945.13918@alphalpha.com>, nazgul@alphalpha.com (Kee Hinckley) writes: > In article <22979@hydra.gatech.EDU> gt1111a@prism.gatech.EDU (Vincent Fox) writes: > >I for one am a little sick of products only written for particular windowing > >systems. > > Motif applications *do* run under both OL and Mit X. More correctly, Motif applications run under the OpenWindows X11/NeWS server, which is not the same as OL. > -- > Alfalfa Software, Inc. | Poste: The EMail for Unix > nazgul@alfalfa.com | Send Anything... Anywhere > 617/646-7703 (voice/fax) | info@alfalfa.com --- Greg Pasquariello grp@unify.com Unify Corporation Be good and never poison people
jim@ncd.COM (Jim Fulton) (03/03/91)
With 20-20 hindsight, it seems to me the problem is due to MIT's "no
policy" philosophy. In my opinion these problems, which are getting
worse every day, could have been avoided had MIT not been so reluctant
to take a stand
... which would very likely have prevented a number of vendors from
supporting X in the first place. Clearly there is a middle ground that might
have worked, but everybody had their hands too full then to risk it.
I don't really think it matters which look and feel
is used, what is important that it be the same everywhere.
I tend to agree with you, but at the time there were a significant number of
people who didn't.
But it's too late now.
I'm not so sure of this. As other people have pointed out, the two are
beginning to converge in many ways. Also, the market is still getting
its teeth (i.e. large customers requiring one or the other).
The best examples of standards are those that are de facto
Which is the reason why MIT didn't provide a look and feel. Such de facto
standards typically have a lot of time and effort already invested in them by
someone, and there were no large toolkits available then.
The take home lesson: when producing a new product, especially one
that requires a new technology, don't be timid about specifying the
interfaces and annointing them as the standard.
Gee, some people have complained that X has done too much of this! :-)
Even if you make
mistakes, and parts must be redone, the community of users will be
better off in the long run.
Perhaps. But, if you try to standardize on something that is clearly
unsuitable, people will give up on you. I'm not arguing that it couldn't
have been done, just that it's not a simple matter.
Jimimp@Solbourne.COM (Warner Losh) (03/03/91)
In article <1991Mar1.210030.26900@csn.org> raney@yertle.Colorado.EDU (Scott Raney) writes: >For example, Open Look requires a >special glyph font to draw its rounded-rectangle buttons. More accurately, there exists at least one Open Look toolkit that requires this. However, there are toolkits that implments Open Look that doesn't require this special font.... Warner -- Warner Losh imp@Solbourne.COM We sing about Beauty and we sing about Truth at $10,000 a show.