Re: vendor collaboration [Re: Smalltalk for Web development?]

"A SERFer" <[hidden email]> wrote in message news:[hidden email]...
> "James A. Robertson" <[hidden email]> writes:
>
> >A SERFer wrote:
> >>
> >> "David Simmons" <[hidden email]> writes:
> >>
> >
> >> >
> >> >Gosh, I sound like I'm making some kind of a case for Microsoft. How
the fashionable
> and successful.
>
> Also, in exactly what market have MS beaten sun?  The internet runs
largely
> (and increasingly) on UNIX.  Which UNIX dominates this section?  Solaris.
> I'm not entirely sure this is healthy : I would prefer a wider range of
> Unixes, else Sun could well have excessive leverage (Although the grown
> of Linux is helping here).

<opinion-enabled>

A bit of history might again help here. The "unix" market was a total mess,
at risk of dying, and in chaos after the demise of the mini-computer
industry due to much cheaper, almost always better, and more diverse
available open-like software for the personal computers on the desktop
(something Microsoft played a key role in enabling to everyone's benefit if
you had witnessed the mini-computer era).

The commercial vendor "unix" market represents the survivors from the
mini-computer era and they absolutely needed some kind of standards just to
survive with their wildly divergent technologies and very hi-cost systems.
Without standards it was too risky to purchase and invest in developing and
using a given vendors technologies (and trapped with their support costs it
was a killer). Add to that mix the profile of their prime customer base
(which in large part you, as a US taxpayer, heavily subsidized through
questionable political processes and contract bidding systems).

Open standards, hah! Your Sun open standard's process was soley due to Sun
grabbing for a survival tactic and its mode of "open-standards" was nothing
like the GNU model or Linux driven evolution where the standards and the
ability to evolve them belonged to anyone who had net access.

This Sun fostered image revisionism regarding things like open-standards is
hooey to me, I was in both academic research institutions and the government
labs during the "historical periods" you are referencing and I lived it (and
worked with Sun and other vendors) first hand. I would never go back to that
world. I'll take Microsoft in a heartbeat compared to that costly and
painful mess.

</opinion-enabled>

-- Dave Simmons [www.qks.com / www.smallscript.com]
  "Effectively solving a problem begins with how you express it."
platform,

> > > James Robertson:
> > >
> > > The mistake is having the idea that Sun does this for altruistic
> > > reasons.  They do it because MS has beaten them in the market so badly.
> > > If the shoe were to reverse itself, you would see Sun start to rapidly
> > > lose interest in open systems.

Yes, they probably would have less interest in <Open Systems>, in the sense
James is referring to.

> >A serfer:
> > I'm sorry, but that is just total nonsense.  Sun have been strong
> > a supporter of open systems since they were a very small startup company.
> > I did not say it was for altruistic reasons!  It was, of course, a
> > business decision, ...

No, they would probably not have less interest in <Open Systems>, in the sense
that "a Serfer" is referring to.

> David Simmons:
> <opinion-enabled>
>
> A bit of history might again help here. The "unix" market was a total mess,
> at risk of dying, and in chaos

Agreed.

> after the demise of the mini-computer industry due to much cheaper,
> almost always better, and more diverse available open-like software
> for the personal computers on the desktop

Disagree - I think you need to split the history in two, here.

The Unix market was a total mess because there were two very strong, but
incompatible flavors of it, viz. ATT, and BSD, and not much happening to
promote interoperability.  Several attempts, but questionable results.
There were many vendors, reselling variously modified versions of these
two basic forms.  

Sun, and other players in the workstation arena played a strong role
in the demise of the mini-computer industry. Client/Server computing
had a much better 'bang for buck' than mini-computers, primarilly based
on hardware cost.  Mini-computers were just too costly to produce.

By focusing on <open systems> and manufacturing efficiency, Sun managed
to get a real foot in the door, against other seemingly entrenched vendors.

They (Sun) were indeed strong supporters of <open systems>,
if we take <open systems> to mean:

        BSD Unix (developed at the 'open' UCB)
        and TCP/IP (developed by the 'open' IETF),

        as opposed to

        ATT Unix (developed at the 'closed' Bell Labs, [and licenced])
        and/or OSI (developed by the 'less open' ISO and 'mandated' by the U.S. Government)
                         
By the time 'desktop' machines started to have an impact
of the kind David describes, it wasn't the state of the 'Unix market',
but rather the cost of the hardware that made the difference,
(the same reason mini-computers had run into trouble).

With desktop hardware coming so much cheaper than workstations,
all it took was marginally functional software to fuel an intensely
competitive marketplace, rapidly improving the [very low quality] software
of the desktop.

Unix had little to do with it.

> (something Microsoft played a key role in enabling to everyone's benefit if
> you had witnessed the mini-computer era).

Eventually true, but perhaps overstated, at least early on.
Due to the progression of Smalltalk -> Mac -> Windows -> Windows 95
you finally had tools that began to compete, and then overtake, the Unix market.

Yes, Microsoft has helped fuel the industry, and helped bring
large quantities of useful new stuff to relative commodity level,
by their aggressive approach to integrating everything known to man
into the Windows platform (e.g. graphics, sound, video, device drivers up the wazoo...)

But it was mostly in response to the accidentally good decision
on IBM's part to make the hardware 'open', which was done in order
to get back in control of the destop onrush, as quickly as possible.
[Failed, of course, but a key role in enabling the 'everyone's benefit].
 
> The commercial vendor "unix" market represents the survivors from the
> mini-computer era and they absolutely needed some kind of standards just to
> survive with their wildly divergent technologies and very hi-cost systems.

Not really true.  Commercial Unix vendors are survivors of either
the mainframe era or the workstation era.  The only survivors of the
mini-computer era are Digital, and perhaps NCR.  The rest are gone,
or at least scrambled into completely unrecognizable form.

> Without standards it was too risky to purchase and invest in developing and
> using a given vendors technologies (and trapped with their support costs it
> was a killer). Add to that mix the profile of their prime customer base
> (which in large part you, as a US taxpayer, heavily subsidized through
> questionable political processes and contract bidding systems).

Bunk.  True, but still bunk.  And no worse for the customer than, say, OS360/70,
or some other proprietary system. Support costs were due to hardware cost and
market size.  As the server/workstation vendors hit the scene, hardware got cheaper.
This grew the market.  As did <open systems>, which really meant <tcp/ip network centered>.
 
> Open standards, hah! Your Sun open standard's process was soley due to Sun
> grabbing for a survival tactic and its mode of "open-standards" was nothing
> like the GNU model or Linux driven evolution where the standards and the
> ability to evolve them belonged to anyone who had net access.

Originally, Sun's open standards were BSD Unix and tcp/ip.
Not a survival tactic at all.  Sun *created* the business with this tactic.
And the standards belonged to anyone who contributed to BSD Unix, and/or anyone
who had access to the internet, read the RFC's, implemented the protocols,
and contributed to interoperability based on their results of using same.

The GNU model, or Linux model, is <open source>, a completely different beast.

Now that we have <open systems>, aka systems connected by tcp/ip running over an
OS which is ideally not completely controlled by a central licencing point, we are also
becoming interested in <open source>, which is itself not unlike BSD Unix.

Difference being that BSD Unix was developed *from* access to ATT Unix, and
controlled by the fact of needing an ATT licence in either case (to get source),
versus the GNU and Linux scenarios, wherein the base form is free of license
constraints, having been developed "from scratch", [although sometimes by people
who would not necessarilly qualify as "clean room" participants].

> This Sun fostered image revisionism regarding things like open-standards is
> hooey to me, I was in both academic research institutions and the government
> labs during the "historical periods" you are referencing and I lived it (and
> worked with Sun and other vendors) first hand. I would never go back to that
> world. I'll take Microsoft in a heartbeat compared to that costly and
> painful mess.

I lived it too.  Sun was a godsend, at the time.

Still, I'll take Microsoft on Intel in IBM-PC compatible boxes
compared to that costly, earlier time.

Sun's revisionism regarding Java is attrocious, intellectually dishonest,
and has been in my way for several years now.  It is also, alas, just business.

As to whether we've escaped the painful mess,
all I can say is, it's still there, but I can (mostly) ignore it.

t.g.f.s.!

> > >>(A serfer):
> > >>
> > >> Sun have always been a supporter of open systems, allowing easy
> > >> porting of software and the implementation of standard systems.

With terms defined as above, yes.  
Easy porting?  Define easy.
Standard Systems? Define Standard.  Define System.

> > >> Even when sun design a processor (SPARC) the release the specs and
> > >> allow others to licence it.

Just like Intel.

Because you have to, at least when you first show up.
Nobody bets the company on hardware that isn't second sourced.

> > >> Java can be implemented by anyone and does not contain hooks to
> > >> bind you to any one supplier or platform.

Java cannot be implemented by anyone, but that's beside the point.

Any one supplier? Maybe...
Any one platform? Define platform.

"David Simmons" <[hidden email]> writes:

>I dealt a great deal with Sun in the 80's and they are no saints, just a
>business like the rest playing in the cutthroat hi-stakes technology game.
>What I can't really fathom is why everyone thinks these other Microsoft
>competitor companies are such good guys; I can easily understand why people
>recognize Microsoft as a problem.

I'm not saying the others are good guys *at all*.  I'm saying that they
are *less bad* guys.  That is a very different thing to say!

I'm sure that Sun are only doing what they are doing for sheet profit.
But, what I feel is that their general approach is better for the end
user and the industry.  

It has been a tiresome and frustrating experience to have to deal with
MS products.  It has almost always been a case of theirbeing a move towards
a common standard in areas of IT, but with MS going their own way.
For what I'm sure are the meanest, most money-grabbing motives Sun
have tended to follow the common standards.  Its not a case of
good or bad, its that common standards are good as they allow
portability and competition.  

>
>As to Microsoft and the risk for QKS, I learned some incredibly costly and
>painful lessons about surviving after being misused by Apple behavior.
>Microsoft has, at least over seven or so years of business interaction with
>them, had the virtue that they didn't mislead me up front about how we might
>get used (of course you have to be smart enough to ask the right questions
>but that is always your responsibility in any business deal).

Again, I'm not saying that MS are evil, or are misleading anyone.
Its just that they have an (enviable) ability to change direction
rapdily.  There have been many firms that have suffered from abandonded
technologies.

I always thought it was more the academic side that pushed open
technologies, but this has always been a matter of historical
debate.

Indeed.  But I guess its the corporate culture that we see.  I think
MS is its own worst enemy in this respect: perhaps it should publicise
other aspecs of its work.  Instead we get Bill Gates making a fool of
himself in Court.

True.

jas <[hidden email]> writes:


'Open Systems' is not ambiguous.  It refers to common public protocols
and interfaces.  POSIX, TCP/IP, POP, NNTP, FTP, NFS, etc.
I haven't a clue what James means.

Actually it was System V and BSD.

TCP/IP is not 'opposed to' OSI. They aren't compenting technologies.

AT&T (System V) Unix was never 'Open'.  It was the APIS that were
available that were 'open'.

No.  You are talking about completely different things.  Imagine the UNIX
market as trucks, and the Windows market as cars.  Its different roles.
Windows has hardly impacted the Technical Workstation or mid-range server
market of UNIX.  

Windows has competed extremely well on the low-end desktop and small-scale
servers:  it has vanquished Netware, for example.

>>
>Still, I'll take Microsoft on Intel in IBM-PC compatible boxes
>compared to that costly, earlier time.

Agsin. you are talking about different things, different markets.
>> > >> Even when sun design a processor (SPARC) the release the specs and
>> > >> allow others to licence it.
>
>Just like Intel.

Yes, so?
>
>Because you have to, at least when you first show up.
>Nobody bets the company on hardware that isn't second sourced.
>
>> > >> Java can be implemented by anyone and does not contain hooks to
>> > >> bind you to any one supplier or platform.
>
>Java cannot be implemented by anyone, but that's beside the point.

Yes it can.  Sun.  IBM.  Kaffe.  TowerJ.  GNU's javac.
What you can't do is use Sun's code.

David Simmons <[hidden email]> said:
>A clean room java is relatively useless without Sun's proprietary controlled
>and Sun business centric frameworks. And, Sun has made sure that on the
>hi-volume platforms they provide a sufficiently performing version to
>prevent any self supporting financial/business basis for any such
>competition.
>
While I disagree on the value of a clean room Java (relative to the value
of Java at all, of course ;-)), I do agree that I see Sun going into the
wrong direction here for the last couple of years.

It used to be that I said to everyone: "Sun is easy to predict. Analyze their
behaviour in terms of how they can sell more boxes and you're done". The good
thing here, and something I think that has helped making Sun what it is today,
is that you could rely on them to stay a hardware-supplying partner, leaving
the software and integration business to you. Now, with their own appserver
offerings and other "extended Java" stuff, they seem to be getting out of that
corner.

>Actually, a large part of all software produced and used in the World is on
>Windows based systems. What are we really saying?
>
Nothing, except that it suddenly has become "no problem" for governments
etcetera to have a single source supplier. I don't have anything against
Windows as a technical system (it's not the best stuff ever made, but it makes
a decent gaming platform if your requirements aren't too high), but against
the fact that it is owned by a single entity that has way too much power, and
has proven to be able and willing to abuse that power.

Contrast this to the Linux market. One system, lots of vendors. The advantage:
competition; the problem: compatibility. So far, the advantages seem to
outweight the problems by far.

>> Microsoft generally operate closed systems, which use undocumented
>> APIs and protcols.
>
>Which, practically speaking is a real problem how?  For their really big and
>important customers, one can obtain source and support for almost anything.

And aginst their really big and important competitors, they abuse this
to the max by /not/ disclosing how they get things done faster/better/... than
would be possible through the published API.


--
Cees de Groot               http://www.cdegroot.com     <[hidden email]>
GnuPG 1024D/E0989E8B 0016 F679 F38D 5946 4ECD  1986 F303 937F E098 9E8B

"A SERFer" <[hidden email]> wrote in message news:[hidden email]...
> >Sun, IBM, HP and all the rest of the Unix vendors have bleated on about
> >"open" systems for 20 years but never settled on anything. Why? Because
they
> >all want their own proprietary technologies to become "standards", just
like
> >everyone els.e
>
> Er... what?
> So I suppose you would not call systems such as posix, X11, TCP/IP, etc.
> 'not successful'?????
>
> Of course vendors have been adding their own extensions, and that has been
> a problem, but to call open systems 'unsuccessful' is to show a
fundamental
> lack of knowledge of the computing industry and its history.

"Open Systems" is and always was a marketing catchphrase for Unix vendors to
deal with IBM and DEC dominance. Please spare us the history lesson.
Crediting Sun with the development of the standards you cite is like
crediting Dell with the development of Lotus Notes or Windows.

Sun is a hardware vendor that licensed Unix from Bell Labs. Comparing them
to Microsoft shows a "complete lack of knowledge of the computing industry
and its history".

> >I hate to say it, but Microsoft has done more to impose quality software
on
> >the market than any other vendor. They got into big companies without
> >thousands of sales droids; that has to be a good thing. What they've done
> >with their resulting power is another issue.
>
> If you think Microsoft has imposed 'quality software' on the market you
> obviously have a very strange definition of the word 'quality'!

Have you been sleeping under a rock for the past 10 years? Or do you think
that office workers should be using Emacs and X-Windows?

> And I forget why I'm bothering to reply to posts that are based on
> wishful thinking and ignorance about the computing industry!
>
> I think I'll stop posting here and get on with actual coding for a while
:)

Based on the volume of your off-topic posts I find that hard to imagine.

"Steve Wart" <[hidden email]> writes:

I didn't credit Sun with developing the standards.  I credited Sun with
spporting them.

>
>Sun is a hardware vendor that licensed Unix from Bell Labs. Comparing them
>to Microsoft shows a "complete lack of knowledge of the computing industry
>and its history".

Not at all.  Both are now competing in the area of software tools.
Both supply operating systems.

Have you *tried* X-windows?  Have you seen GNOME or enlightenment?  
You seem to confure popularity with quality.  Visual Basic is popular-
would you say it was a quality language?

Steve

[hidden email] (A SERFer) wrote:

>Have you *tried* X-windows?  Have you seen GNOME or enlightenment?  

I sure have, and IMO, I prefer the Windows GUI.

>You seem to confure popularity with quality.  Visual Basic is popular-
>would you say it was a quality language?

I definately wouldn't say Visual Basic is a quality language, but I would
say for the kinds of applications I would be building, that it would be a
much more useful (and rapid) tool than Java.

Ian

> Have you *tried* X-windows?  Have you seen GNOME or enlightenment?
> You seem to confure popularity with quality.  Visual Basic is popular-
> would you say it was a quality language?

Yup. Unfortunately, I use X Windows almost every day. I must say that VW's
UI model is better suited to X than it is to MS-Windows, which IMO has been
a big factor inhibiting the acceptance of Smalltalk. I'm not sure about
other Smalltalks, although it is probably true of Squeak. (With apologies to
Dolphin users, of course, which appears to conform well to the Windows UI,
although at the expense of portability.)

I've never used VB, but from what I've seen of it, I would say that it is of
extremely high-quality. The environment is rock-solid, hundreds of millions
of developers use it every day, and to somehow conclude that it is of "low
quality" is either wishful thinking or misguided elitism.

I would love to see a Smalltalk with a UI builder whose "quality" is
somewhere near that of VB (QKS seems to be on the right track; I'm not sure
about Dolphin), but which maintains true binary cross-platform portability.
My concern is that as long as VW developers are using X11 on Unix as their
primary development environment, they are missing the mark on what it will
take to gain widespread adoption.

"Steve Wart" <[hidden email]> wrote:

>I would love to see a Smalltalk with a UI builder whose "quality" is
>somewhere near that of VB (QKS seems to be on the right track; I'm not sure
>about Dolphin), but which maintains true binary cross-platform portability.
>My concern is that as long as VW developers are using X11 on Unix as their
>primary development environment, they are missing the mark on what it will
>take to gain widespread adoption.

IMO, WindowBuilder on VisualAge goes quite a long ways to being as good as
VB, and in addition is cross-platfrom, and yet still retains the windows
look and feel.

Ian

"Steve Wart" <[hidden email]> writes:

I don't mean to sound elitist.  I believe VB to be low quality because
of its design.  The implementation/packaging/marketing of the product is
very good.  The reason that I think its low quality is because it is
has always missed out on features that trained software engineers find
important.  The most serious of these is OOP.  The component-based nature
of VB over most of its history required a separation of object users
(VB coders) from object creators (C/C++).  If you wanted to create
VBX/OCX/ActiveX controls you could not (until recently) use VB itself.
Even when you use VB to create objects, aspects of OOP such as inheritance
and polymorphism are either absent or very poorly implemented.

Bill Gates had seen Smalltalk and had said good things about it, so why
he ended up delivering the non-standard mess of VB is a puzzle to me.
If they wanted to develop a good object-based rather than fully
object-oriented language (if they felt that OOP was too complicated
for MS to implement, or to difficult for the masses to use) there were
plent of good examples of how to implement this, such as Oberon.

Other companies managed to produce full OOP systems for Windows: the most
successful being Turbo Pascal for Windows, which morphed into Delphi.

My assessment of VB as 'low quality', is in the same way that I would
call 'tabloid  journalism' as low quality:  The packaging may be good,
and the pages may be glossy, but the overall design and level at which it
is pitched and the content are not of a high standard.

>
>I would love to see a Smalltalk with a UI builder whose "quality" is
>somewhere near that of VB (QKS seems to be on the right track; I'm not sure
>about Dolphin), but which maintains true binary cross-platform portability.
>My concern is that as long as VW developers are using X11 on Unix as their
>primary development environment, they are missing the mark on what it will
>take to gain widespread adoption.

I'm not sure about this.  There *has* to be something better than Windows,
and I think that the total customizability of X window managers allows
people to see what can really be done with UIs.  On the other hand, perhaps
Windows has become like the QWERTY keyboard - a bad design that is now
so widespread that nothing else stands a chance.

Steve
>
>

 Actually if your product is *good* in a domain sense, you will be
surprised to see, how the people can (and are willing) to use these
applications - even if they do not look 100% like MS-software.

 You can create programs with all dialects of Smalltalk, which do not
look like 100% windows, but which look quite smart !

 If you write a program, which has been programmed already 1000 times
and the only way to be better than these programs is to do a much nicer
GUI ... ok, that I will admit, that it would be not my preferred way
to use Smalltalk.

 On the other hand I still believe, that most of the VW programmers DO
program under Windows.


 Marten

Hi Ian,

I've heard great things about WindowBuilder and VAST (as Mr. Gale reminds me
every day), but when I say "cross-platform", I really am referring to MS
Windows, Linux and the Mac OS. VAST is cross-platform between Windows and
most commercial Unixes. As I implied earlier in this thread, I personally
find commercial Unix environments to be way overpriced and mostly unpleasant
to use.

The Mac and Linux are mass-market operating systems that are way too
important to ignore. Kudos to the VW folks for supporting these platforms,
fooey on IBM.

Cheers,
Steve

"Ian Upright" <[hidden email]> wrote in message
news:[hidden email]...
> "Steve Wart" <[hidden email]> wrote:
>
> >I would love to see a Smalltalk with a UI builder whose "quality" is
> >somewhere near that of VB (QKS seems to be on the right track; I'm not
sure
> >about Dolphin), but which maintains true binary cross-platform
portability.
> >My concern is that as long as VW developers are using X11 on Unix as
their
> >primary development environment, they are missing the mark on what it
will
> >take to gain widespread adoption.
>
> IMO, WindowBuilder on VisualAge goes quite a long ways to being as good as
> VB, and in addition is cross-platfrom, and yet still retains the windows
> look and feel.
>
> Ian
>

"Steve Wart" <[hidden email]> wrote in message
news:J3OX5.767241$[hidden email]...
>
> I've heard great things about WindowBuilder and VAST

BTW, if you want to read about WBPro/VAST and see lots of screen shots of
its features, you can do so here:

    http://www.instantiations.com/sts/wbp.htm

-Eric Clayberg
 Sr. Vice President of Product Development
 Instantiations, Inc.
 mailto:[hidden email]
 http://www.instantiations.com

"Steve Wart" <[hidden email]> wrote:

Hi Steve..

Yeah, I totally agree with you there.  It badly needs Linux & Mac support.
Maybe we'll see it someday...  (I'm not overly hopeful)

Ian

[hidden email] (A SERFer) wrote:

>I don't mean to sound elitist.  I believe VB to be low quality because
>of its design.  The implementation/packaging/marketing of the product is
>very good.  The reason that I think its low quality is because it is
>has always missed out on features that trained software engineers find
>important.  The most serious of these is OOP.

Are you saying OOP is a requirement to make something "high quality"?  Does
that automatically make several flavours of UNIX or other operating systems
"low quality" because it's kernel wasn't written in an OOP language?  Does
that mean that there has never been any high quality software delivered that
was built upon non-OOP tools?

Don't get me wrong, I really like OOP, but I see it as a different
*approach* to software development.  Yes, IMO, it's a better approach, but
the other approaches have also proven to work well and create high quality
end-results (which in the end, is all that really matters).  Of course,
maybe it would take more manpower in the end to achieve this result, but I
don't see anything by this that can be regarded as "low quality".

Low quality to me, isn't a poor design, but something that is so buggy or
implemented so poorly that one could never achieve the desired end result.
Given the huge number of delivered VB applications out there, and given that
some of them are more stable than those created in C++, and as much as I
hate VB's design, it can't be considered "low quality".

Ian

Ian Upright <[hidden email]> writes:


No.  Its the way that OOP is 'half implemented' in VB is that is poor
quality.  Its as if MS are saying that you can go so far, but no
further with the language.  Its like a car with only 2 gears.
>
>Low quality to me, isn't a poor design, but something that is so buggy or
>implemented so poorly that one could never achieve the desired end result.
>Given the huge number of delivered VB applications out there, and given that
>some of them are more stable than those created in C++, and as much as I
>hate VB's design, it can't be considered "low quality".

I think the language is a poor implementation of ideas that *does*
restruct what you can do, and what you can achieve. (Or at least makes it
far more difficult than necessary).

>
>Ian
>

I heard today that WindowBuilder has been ported to VW, but there are
licensing issues. I would like to write to an "API" that is portable between
Smalltalks (because once you've achieved operating system independence, then
you have to deal with Smalltalk-vendor independence!).

Is WindowBuilder available for VW? If not, what would it take to make it
available?

"Eric Clayberg" <[hidden email]> wrote in message
news:90obb9$5t7$[hidden email]...

"Steve Wart" <[hidden email]> wrote in message
news:BWYX5.768966$[hidden email]...
> I heard today that WindowBuilder has been ported to VW, but there are
> licensing issues.

That's news to me. AFAIK, WB has not been ported to VW (although that has
been discussed many times in the past).

> I would like to write to an "API" that is portable between
> Smalltalks (because once you've achieved operating system independence,
then
> you have to deal with Smalltalk-vendor independence!).

WB itself does not present a portable API. The two WB versions (for VSE and
VAST) each generate code to their respective Smalltalk GUI APIs (rather than
to a common, portable WB API). WB/VSE code can be run in VAST with an
available emulation layer but not vice versa. Creation of a portable GUI
framework has been discussed as a potential Camp Smalltalk projects, so
maybe it will happen one of these days.

> Is WindowBuilder available for VW?

Not that I know of (although Cincom could certainly create it for VW, if
they wished).

> If not, what would it take to make it available?

If I were the one doing the work, it would take about $200K (and about three
months). Know anyone who wants to prepay for 400 licenses at $500 a pop? ;-)

-Eric

>A SERFer
>
> >jas <[hidden email]>
>
...snip...
That would be ATT System V, no?
If I recall correctly, the problem existed earlier than System V,
which was itself an attempt to create a combined version,
i.e. both ATT and BSD compatible features.

> >...snip...
> >
> TCP/IP is not 'opposed to' OSI. They aren't compenting technologies.

How do you figure? The OSI/ISO standard, as expressed in (FIPS/GOSSIP),
was a directly competing network technology.
 
> AT&T (System V) Unix was never 'Open'.  It was the APIS that were
> available that were 'open'.

Agreed.
 
Windows has impacted the Technical Workstation market of Unix,
though much less so the mid-range server market.

Used to be rare to do development on a Windows based machine.
Business users had Windows machines, technical developers had Unix workstations.

It is now much more common to see developers using a Windows based machine,
even when the eventual target is a Unix server.

In VW, for instance...

> Windows has competed extremely well on the low-end desktop and small-scale
> servers:  it has vanquished Netware, for example.

Yes.

> >...snip...
> >
> >Still, I'll take Microsoft on Intel in IBM-PC compatible boxes
> >compared to that costly, earlier time.
>
> Agsin. you are talking about different things, different markets.

Somewhat.
The point stands.
Microsoft on Intel is cheaper than Unix on (various non-intel) workstations,
with generally speaking, more features.  It's a good deal - even though the
quality, especially of the underlying OS, is much lower, indeed quite low.

> >...snip...
> >
> >> > >> Java can be implemented by anyone and does not contain hooks to
> >> > >> bind you to any one supplier or platform.
> >
> >Java cannot be implemented by anyone, but that's beside the point.

Sarcasm alert ;-)

> Yes it can.  Sun.  IBM.  Kaffe.  TowerJ.  GNU's javac.
> What you can't do is use Sun's code.

Agreed.

-jim