home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!europa.asd.contel.com!emory!wupost!gumby!yale!yale.edu!qt.cs.utexas.edu!cs.utexas.edu!sun-barr!olivea!charnel!csusac!unify!openlook!openlook-request
- From: fgreco@cfdev1026.shearson.com (Frank Greco)
- Newsgroups: comp.windows.open-look
- Subject: re: XView
- Message-ID: <epwcil5@openlook.Unify.Com>
- Date: 6 Nov 92 17:49:14 GMT
- Sender: news@Unify.Com
- Lines: 101
-
- > > From various industry-wide discussions, it would seem that OLIT is a
- > > more attractive library for new development work.
-
- try it first... If you keep on putting pins in your hand,
- you'll eventually put up with the pain too. Don't let the
- industry-wide discussions *dictate* to you how to solve your
- problems (said with a Perot-twang).
-
- > If Sun dumps XView, it leaves a large number of XView programmers
- > out in the cold. I use XView and GUIDE for all my window development,
- > and Harris has production software based on XView. If Sun drops XView,
- > we take a serious maintenance and development hit.
-
- I don't think Keith B. (as well as many other programmers within
- Sun) understands how the average MIS (there... I said it) shop
- thinks and works. Dropping support for XView is a stupid move
- on their part. Perhaps the same thing will happen when Sun
- announced that Sunview was going away (I recall they originally
- said by 1988, Sunview would no longer be available), that is
- there will be enuf complaints that XView will be kept afloat
- until folks have migrated to something ... better (OLIT is
- NOT better).
-
- > I like the XView toolkit, and prefer it to the OLIT style of
-
- XView is easy, painless and fun. Yes there are warts, but
- if the toolkit was fixed (a large re-write of the panel and
- panel-items is in order), incorporate slingshots in the
- mainstream, give us a bonafide (and new) interface to
- postscript and give us a *REAL* C++ interface (devguide's C++
- output is... ahem... non-optimal) then we're cooking.
-
- Someone from Sun told me that they've noticed that there
- are not too many people posting to the Xview-ish net groups
- in comparison to the Xt group which they believe
- indicates a lack of interest in Xview. Maybe it belies
- the fact that Xview is so easy, folks do not need to ask
- that many questions?
-
- With apologies to many of my good net friends, Xt-style of
- programming is a poor example of good software engineering
- (imho). Yes, you can do it if you hold your nose. But in
- today's commercial environment especially with strong
- competition from the Microsoft's of the world, your
- applications must be functional, attractive and fast. Xt,
- again imho, does not easy facilitate such applications; OLIT
- and MoOLIT do not cut the mustard in a modern workstation
- environment. The application programmers of the world (system
- types can deal with anything... right?) don't deserve Xt.
-
- > programming. Why is it that when a few people in Sun decide they
- > don't like something, they make these decisions that affects
- > thousands of their customers. They need to remember that code developed
- > by just a few developers can result in many workstation sales. Some
- > of the tools I'm working on in XView will translate to perhaps $100
- > million in workstation sales over the next five years or so.
-
- Again, that's because many of the folks inside of Sun *NEVER*
- worked for an applications (ie, non-computer-vendor) group
- before. Not a criticism per se, merely an observation.
- As you know, I am very much a believer in Sun (and
- subsidiaries) and have been for a long, long while. But I
- believe that Sun's current software strategy is become
- extremely fragmented and ill-chosen.
-
- Now think of how all-encompassing DOE is supposed to be. Now
- think of the way that TNT/NeWS has been handled. Think of the
- way this XView situation is being handled. Think of the
- performance of OpenWindow's X/NeWS server. Think of the
- floating point performance of the SPARC v.s. the latest HP
- boxes. Think of the XView panel-list object. Think of the way
- XView fonts (esp in panels and panel items) and colors are
- handled. Think about cmdtool/shelltool (not too long, you'll
- get sick) and the underlying TERMSW/TTYSW. Think about the
- lack of a table object in XView. Think of the relationship
- between OLIT and Devguide. Think of the performance of
- the GX wrt OpenWindows; shouldn't things be much faster?.
- Think of the lack of distributed application tools for RPC
- programming.
-
- I could go on, but the idea is that if Sun wants people to
- use DOE in the future, they are not relaying a feeling of
- confidence to the current technical customer base.
-
- I wonder what Sun's internal IS staff (CIO Radushel's folks)
- are using for GUI tools.
-
- > Sun is becoming increasingly unresponsive to the people that made them
- > successful: the Unix hackers. We are the ones who put all the wood behind
- > one arrow, as Scott loves to say. Why is it that now, I feel like Sun
- > keeps the arrowhead, and we get the shaft?
-
- Boy do you have that right! And they expect us to keep the Unix
- torch burning while MS NT rolls in?
-
- It's broke. Start fixing it.
-
- Frank G.
-
- - "Remember, you're the boss and they run DOS"
- =-=-=-=-=-=-=-=-=-=
-