janssen@parc.xerox.com (Bill Janssen) (08/22/90)
msc@ramoth.esd.sgi.com (Mark Callow) writes:
Even PageView, which should be the premier NeWS application,
seems lacking. It has lost the random page access feature of psview.
While agreeing that NeWS is a 2nd-class citizen in OW 2.0, and that
pageview should have random page access, I don't think that pageview
should be seen as the premier NeWS application/demo. It is too easily
matched/overmatched by Display PostScript analogues on other X
servers.
On a different thread: I'm not sure that the 2nd-classness of NeWS is
Sun's fault. Seems to me that the world (or maybe "commercial
market") is the one putting NeWS in the back seat, and that Sun,
wanting to be a player in that world, is forced to go along.
Bill
--
Bill Janssen janssen@parc.xerox.com (415) 494-4763
Xerox Palo Alto Research Center
3333 Coyote Hill Road, Palo Alto, California 94304mlandau@bbn.com (Matthew Landau) (08/23/90)
janssen@parc.xerox.com (Bill Janssen) writes: >On a different thread: I'm not sure that the 2nd-classness of NeWS is >Sun's fault. Seems to me that the world (or maybe "commercial >market") is the one putting NeWS in the back seat, and that Sun, >wanting to be a player in that world, is forced to go along. This is a good point. Let's be realistic for a minute: as a vendor of worktations, Sun has to ask itself what the customer base is demanding, and at the moment, the customer base is demanding a good, high-performance implementation of X11. Yes, *we* all know that X pales in comparison to NeWS along any dimension you care to compare... except for the all-important dimension of market popularity. Given limited resources, I can't really fault Sun for deciding that they had to emphasize a good solid X11 implementation at this point. After all, where is NeWS going to be if Sun can't sell any machines? My hope is that, now that Sun HAS an implementation of X11 that's at least as good as anyone else's (and one that exhibits many fewer bugs than either DEC's or IBM's product offerings), and can satisfy that very immediate market demand, they'll have the freedom to redirect some of those resources back to the NeWS side of things, which is where the real added-value of OpenWindows (and Sun hardware) is going to come from in the future. -- Matt Landau Rebel without a clue. mlandau@bbn.com
ittai@shemesh.GBA.NYU.EDU (Ittai Hershman) (08/23/90)
In article <14471@diamond.BBN.COM>, mlandau@bbn.com (Matthew Landau) writes: > Let's be realistic for a minute: as a vendor of worktations, Sun has > to ask itself what the customer base is demanding, and at the moment, > the customer base is demanding a good, high-performance implementation > of X11. Yes, *we* all know that X pales in comparison to NeWS along > any dimension you care to compare... except for the all-important > dimension of market popularity. If Sun stopped its Open Look jihad and simply adopted the Motif look-n-feel, they would have a lot more latitude to develop new software. Let's not have a Motif vs. Open Look flame war -- see the latest Unix Today! editorial on the subject. I'm just pointing out that if Sun weren't expending so many resources on fighting Motif, they could develop better X11 and NeWS implementations and gain even more market share. -Ittai
tro@adiron.UUCP (Tom Olin) (08/24/90)
In article <4806@shemesh.GBA.NYU.EDU>, ittai@shemesh.GBA.NYU.EDU (Ittai Hershman) writes: > In article <14471@diamond.BBN.COM>, mlandau@bbn.com (Matthew Landau) writes: > > Let's be realistic for a minute: as a vendor of worktations, Sun has > > to ask itself what the customer base is demanding, and at the moment, > > the customer base is demanding a good, high-performance implementation > > of X11. Yes, *we* all know that X pales in comparison to NeWS along > > any dimension you care to compare... except for the all-important > > dimension of market popularity. > > If Sun stopped its Open Look jihad and simply adopted the Motif > look-n-feel, they would have a lot more latitude to develop new > software. Let's not have a Motif vs. Open Look flame war -- see the > latest Unix Today! editorial on the subject. I'm just pointing out > that if Sun weren't expending so many resources on fighting Motif, > they could develop better X11 and NeWS implementations and gain even > more market share. > > -Ittai I'm sure Sun appreciates the concern of both you and "UNIX Today!" for their well-being. However, whether or not Sun is doing the right thing, you, the "UNIX Today!" columnist, and many other people in the world seem to be missing a key point that has been discussed in previous issues of that same newspaper. The June 25 issue's lead article was on the June 11 Usenix conference held in Anaheim. At that conference, UNIX pioneer Dennis Ritchie of Bell Labs remarked, "Standards committees are not the best way to create a standard. Standards meetings and standards themselves are horribly political things.... One thing that people forget is that many standards are made by rather small groups of people. A few good people can really save the day, and a few idiots can really make it miserable for years to come." He also touched on some of the shortcomings of some of today's standards, including POSIX and X11. The July 23 issue's lead article was on a new experimental operating system being developed at Bell Labs. In that article, Ritchie's colleague Rob Pike was quoted discussing standards: Previously, standards were decided once a technology had been out in the market for some time. Now, he said, "we're standardizing before we build systems," with potentially dangerous consequences. "Standardization now usually precedes understanding," added Pike. "For instance, people are now trying to standardize distributed systems when they [commercial implementations] don't even exist yet." Don't misunderstand me. I think standards are great, if treated properly. But I also agree with what Ritchie and Pike (and others) are saying. Standards *are* being defined prematurely in many cases. The "UNIX Today!" editorial you mention states, "... most people say Open Look is technically superior to Motif." That is generally consistent with my experience with people who have used both. I have also noticed that people who know much about both X11 and NeWS generally consider NeWS to be technically superior to X11. The editorial also states, "Users, however, have overwhelmingly decided to use Motif." Perhaps. Some may doubt that the GUI war has been won, but there is little doubt that X11, whether good or bad, is in control of the window system battlefield. All of this raises some questions: 1. Assuming that NeWS is technically superior to X11 and that OpenLook is technically superior to Motif, did/do users really *want* to choose second-best solutions? 2. Assuming that NeWS is technically superior to X11 and that OpenLook is technically superior to Motif, is it mere coincidence that NeWS and OpenLook both came from Sun, the workstation market leader, while X11 and Motif both came from consortiums led by Sun's competitors in the workstation market? 3. Assuming X11 is the window system standard for the foreseeable future, does the fact that the whole world has decided that it is the standard mean that it will be the one and only standard forever? Other than incremental enhancements (R5, R6, ...), will it never get any better than this? My own feeling is that users who have decided to specify X11/Motif as their "standard" have done so, not necessarily because they believe that those products are the best answer to their needs, but because they believe that those products will be the most widely available from the industry. And while it is certainly important to avoid investing in products or vendors that might disappear in the near future, it is also important to understand that today's standards will not last forever. No matter how good X11 is, it will someday be replaced by something better. In fact, the same thing could even happen to - gasp! - UNIX. (See, for example, the aforementioned lead story in the July 23 issue of "UNIX Today!".) One might ask, if Sun is being stubborn on Motif, why aren't they being stubborn on NeWS? Simple. NeWS cannot currently replace X11. The product (NeWS) is not as mature as X11, and the whole world has a few years worth of investment in X11. Not so with Motif. The world could still switch to OpenLook without too much pain. So who's being stubborn, Sun or OSF? Perhaps, from a business standpoint, Sun should give up and go with the flow. From a technical standpoint, however, I admire them for sticking to their guns. And, as a programmer and a user, I prefer that *any* standards contest be decided on the basis of technical merit, not popularity. Then again, maybe I'm foolish for feeling that way. Please don't consider this an invitation for a Motif/OpenLook or even an X11/NeWS flame war. Please redirect all flames to /dev/null. If you don't like what I wrote, feel free to ignore it completely. Others probably will. -- ____ ____ | __ | /||___ | ||__||//|| __|| ||___//_||| __| || //__|| \\ Tom Olin ...!uunet!adiron!tro (315) 738-0600, Ext 638 || // || \\ PAR Technology Corp, 220 Seneca Tpke, New Hartford NY 13413
ittai@shemesh.GBA.NYU.EDU (Ittai Hershman) (08/24/90)
Interesting. I had an offline conversation with someone else on this topic, and I quoted the very same parts of the Unix Today! editorial as you did: ...and most people say Open Look is technically superior to Motif. Users, however, have overwhelmingly decided to use Motif. Whether these decisions are purely political, or because of Motif's link to DOS and OS/2 is irrelevant now; it's dead history. I think there is an important distinction being developed that is helpful to a rapprochement. The users, for whatever reason, like Motif; the programmers seem to feel that OpenLook is technically superior, and many in this group certainly feel that NeWS beats any X11 toolkit hands down. I see two different issues here, and there is no reason they are mutually exclusive. It seems simple: generate an XView and tNt which use Motif looking graphics. To the extent there is commonality, adopt the Motif styleguide, and where there is none, innovate. For example, a graphic designer can easily design a motif-looking pushpin. A case in point: Visix has a a "desktop manager" product called Looking Glass. Looking Glass is a Motif application, but it not written using the Motif toolkit: they wrote their own toolkit, because Motif did not have styleguide features they wanted (some of which are in OpenLook). Sun can forfeit a small skirmish and win the war. Once the users are happy, Sun can continue on its way to make NeWS or son-of-NeWS the successor to X11. And in the interim, make XView the toolkit of choice. Is it really worth fighting to the death about graphical design? Are roundtangles worth that much! C'mon, lets get on with it... -Ittai
uad1077@dircon.uucp (08/25/90)
[ Discussion about why NeWS isn't on every desktop...] There seems to be a significant sub-culture of people who liked NeWS so much they tried poting it (yeah, even way back in 1987). The ports worked real well, but unfortunately many of these people got so fed up of Sun's attitude to these people who were trying to help promote this product that they went away and did more rewarding things instead. To the best of my knowledge, Mark Callow is the only one of these people who has lasted the course. Now, if only Gosling had set up an independent software house when he left CMU, things might indeed have been different. Please stop blaming X. It may be bad, it may be political, but I think it is really just expanding into a vacuum. -- Ian D. Kemmish Tel. +44 767 601 361 18 Durham Close uad1077@dircon.UUCP Biggleswade ukc!dircon!uad1077 Beds SG18 8HZ United Kingd uad1077%dircon@ukc.ac.uk
montnaro@spyder.crd.ge.com (Skip Montanaro) (08/28/90)
In article <1990Aug24.181515.18866@dircon.uucp> uad1077@dircon.uucp writes:
There seems to be a significant sub-culture of people who liked
NeWS so much they tried poting it (yeah, even way back in 1987).
The ports worked real well, but unfortunately many of these people
got so fed up of Sun's attitude to these people who were trying to
help promote this product that they went away and did more rewarding
things instead. To the best of my knowledge, Mark Callow is the only
one of these people who has lasted the course.
Scott McNealy spoke recently at a local Sun symposium we hold each year. He
said (paraphrased):
"The biggest mistake I ever made was rolling over on the X versus
NeWS issue."
He made two other statements that reflect on the X vs. NeWS and Open Look
vs. Motif things:
"We didn't set out to design a new graphical user interface. Bill
Joy and I first talked with Bill Gates. He wanted too high a royalty
for Windows. [Sun wanted to pay $5 per seat, Gates wanted $80 and
wouldn't go below $50.] Then we went to John Scully. He was too busy
promoting a new book. Then we went to Steve Jobs. He said, 'I'll get
back to you.' and never did, so AT&T, Xerox, and Sun teamed up to do
Open Look."
"The royalties for Motif 2.0 will be double that for 1.0 [$80 vs.
$40]. To get retail dollars, multiply that by 5. It's the only
revenue OSF has, so they're milking it."
--
Skip (montanaro@crdgw1.ge.com)tro@adiron.UUCP (Tom Olin) (09/05/90)
I'm afraid my first article did not clearly make my points. Let me try one more time. In article <4845@shemesh.GBA.NYU.EDU>, ittai@shemesh.GBA.NYU.EDU (Ittai Hershman) writes: > The users, for whatever reason, like > Motif... See point 2 below. > It seems simple: generate an XView and tNt which > use Motif looking graphics. To the extent there is commonality, adopt > the Motif styleguide, and where there is none, innovate. For example, > a graphic designer can easily design a motif-looking pushpin. There is more to OPEN LOOK's pushpin than just its appearance. It plays a very specific role in OL's look-and-feel. As far as I know, that specific role is not defined in Motif. And I believe this whole Motif vs. OPEN LOOK debate is primarily over look-and-feel, isn't it? > Sun can forfeit a small skirmish and win the war. Once the users > are happy, Sun can continue on its way to make NeWS or son-of-NeWS > the successor to X11. And in the interim, make XView the toolkit > of choice. See point 3 below. > Is it really worth fighting to the death about graphical design? > Are roundtangles worth that much! C'mon, lets get on with it... See point 3 below. > > -Ittai A quick recap of the points I was trying to make: 1. I question the wisdom of deciding on standards before they have been implemented and/or fully tested. And I question the durability of standards and the importance with which people seem to view them. Standards do get replaced (e.g., CORE graphics, GKS, etc.). While I support the proper development of standards, I often feel that too many people have gone overboard (i.e., off the deep end) on the idea. 2. I believe most users in the Motif camp have chosen Motif, not because they particularly like it, but because they see it as the safe business choice. There is a difference. 3. If, as you and others claim, the world has decided on Motif, then it shouldn't matter whether Sun gives up on OPEN LOOK or not. So why is everybody still wringing their hands and begging Sun to give up the fight? Maybe the world hasn't really decided? Personally, I can live with X11 instead of NeWS, and I can live with Motif instead of OPEN LOOK. In both cases, the job gets done. But I don't have to like the way some people make their decisions. I'll stop kicking this horse now. My foot hurts. -- ____ ____ | __ | /||___ | ||__||//|| __|| ||___//_||| __| || //__|| \\ Tom Olin ...!uunet!adiron!tro (315) 738-0600, Ext 638 || // || \\ PAR Technology Corp, 220 Seneca Tpke, New Hartford NY 13413