Some of the responses, if answered, would be restating the same things I already
said. So I'll skip some of it.

>With IBM, Oracle, Bea, etc. java is nothing and Sun knows it.

LOL! I'd say a lot of programmers have discovered the last couple of years
that Java is nothing. Sorry, I know that is not what you meant, but it was
so tempting.

>Second, depending on the application, it
>can be cost effective.

That I agree with. Total life cycle cost includes labor. If a development
staff is Java experienced, retraining and turnover costs may exceed the cost
savings of going Microsoft. That is what managers are supposed to figure
out.

>>The company that is porting it was mentioned in an article I read (not


>>from MS) where MS is cooperating.

>Which one?

Check out http://www.ximian.com/devzone/projects/mono.html and
http://www.ximian.com/about_us/press..._announce.html
and
http://go-mono.com/ and
http://go-mono.com/faq.html
It explicitly states that MS is not helping with the project, but they have
provided conversation back and forth with Ximian, and isn't opposing them
or suing them (as Sun is wont to do over Java). With no profit motive for
MS, that, to me, is MS cooperating.

>Then you don't understand life cycle. You want the life cycle of an
>application to be as long as possible. A life cycle is not just
>designing coding and implementing. You want the iterations to be short.

LOL! I guess you are the one who doesn't understand the common understanding
of life cycle. A life cycle is the time from requirements definition to
replacement of the system. Subcomponents (hardware & software) may have
their own lifecycles. The longer the life cycle, the more the cost is spread
out, reducing annual budgets. If you would like for me to explain the software
development life cycle in more depth, I'd be glad to.

>Benchmarks are usually rigged. Talk to IBM. The have a hand in all
>platforms.

IBM's only long term success has been in mainframes. They are not a major
player in PCs (especially after the bonehead PS2 move back in the 80s), in
UNIX servers, or Windows servers. And it would be hard to name any successful
non-mainframe software they make. The TPC benchmarks may not reflect everyone's
real-world use, but they are the same standards for all, and not designed
by MS. Yet, MS wins. Hands down.

>OO design and coding does take long than proceedural. But the design and
>coding iterations can be shorter. It is not an insult just realility.

Neither I, in using VB & VB.NET, nor my former Java colleagues, did procedural
programming. Maybe years ago in the VB3, Clipper, FORTRAN, and COBOL days,
but not in a long time. VB4, 5, and 6 were sufficiently OO that apps could
be developed using OO techniques in VB. In fact, at a former job, I could
take Rational Rose files for Java development, and use them to create VB
class files with identical interfaces. And vice versa. The only drawback,
that VB.NET fixes, is that inheritance was interface inheritance only. But,
using good code reuse techniques, that was easily overcome.

>How long ago was this Java developement? Java works and works well. The
>fact that there is good Java development going on leaves the only answer
>to be they weren't that experienced.

It was 3 years ago, and they are still doing it today. I keep in touch,
and they keep advancing their product line. I didn't say Java doesn't work
- I said it is more costly to develop in it. And that I stand by. I guarantee
you that whatever your Java apps are, I could develop them in VB in less
man-hours for less cost that perform better.

>I've used both (VB more than any Java) and I am more productive in my Java
>IDE.

Too bad someone won't finance putting that to the test. Give each of us
the same project, and see who finishes first and best.

>I was using 'experienced' the way you were. There are many VB
>programmers. The majority of them are not 'experienced'.

I'll agree there is a higher percentage of sloppy VB programmers, simply
because MS makes its tools more usable and available to the "average person".
But in corporate, professional IT teams, VB is the dominant language. Java
has been a disappointment. Failed in the browser, too dog slow in the middle
tier.

>Yes, bugs are bad. But usually programmer error not language.

Some languages are more difficult to get right, hence the higher incidence
of bugs during initial development.

>You said Java has greater platform requirements. Just proving it doesn't.

You didn't prove that. I am talking about real world Java apps that actually
do something. JVMs are fat. Always have been. Truly compiled (to machine
code) Java programs are much better, but then you lose the portability.

>By the way, they make good firewalls. And embedded computers.

Windows is taking a larger market share of embedded OSs than Linux. And
firewalls are made by the software, not the OS. But UNIX is a good use for
a dedicated firewall.

>Lots of 386 and 486 computers are in use today.

Not as servers in corporate America.

>But VB6 is dead.

Tell that to the millions of people using it, the numerous third party component
vendors that are still selling OCXs and DLLs and VB add-ins. There are still
sites using VB5, VB4, and even VB3. Probably on your 386s and 486s.

>VB.Net is totally different.

No, not totally. You can make 1 to 1 corelations between COM objects used
in VB6 and namespaces used in VB.NET. It's still using objects. But how
you access them, name them, and organize them is totally different. I work
with both, and it is no big deal. Most of the whiners I have heard are just
folks who don't want to change with the times. I am an old guy with many
years of experience, but one constant is that you have to be willing to try
new things and learn new technologies to remain relevant. I tried Java,
learned it, wrote a few small apps, assessed it versus MS technologies, and
chose MS. You obviously did the same, and chose Sun's Java. Nothing wrong
with that.

>And it and C# resemble Java

Well, C# sure does! But VB.NET is way more like VB6 than like Java. The
CLR certainly learned some valuable lessons from Java. Just another example
of MS being able to provide better tools than Sun, give the same concepts.
MS never met a good idea they were afraid to steal.

>>Come on, name a company who has extended the Java language or made a Java
>>VM without Sun's permission.

>IBM.

Oh, please. You think IBM did *anything* to Java without Sun's permission?
Remember, Java was rejected by the standards agencies in the US and Europe
because Sun wouldn't let go. MS didn't make that mistake with the ECMA and
C#.

>>>Could you make mine run? It crashes alot.

>>Probably. Are you running NT, Win2K, or XP? Whose JVM do you use? Are
>>you sure your drivers were certified for the OS? Have you checked for


>>memory leaks in non-MS apps? I've had NT servers than run for months
>>before being rebooted, and only then to upgrade something. Win2K and XP


>>do the same.

>It has nothing to do with Java. It is Windows. You are lucky. Must not
>be doing anything complicated.

Cute. If you knew Windows servers very well, you would know most Windows
problems are untested video and other drivers, poorly made third party DLLs,
etc. I have never found a Windows server that was crashing that I couldn't
trace the problem to third party software that was buggy. Now, if you'd
like to discuss the numerous things that break when applying minor UNIX patches,
let me know. I saw that all the time with Solaris servers and third party
UNIX apps. Never had that problem with Windows service packs, which are
less numerous than UNIX patches.

>No people don't 'want' Windows. The are ill informed or aren't given a


>choice.

OK. Blame the customer. What they really want is a cryptic OS, for which
there is very little commercially successful software. Like Linux. In this
country, what the consumer wants, they get.

> Go to Best buy and find a computer with anything besides Windows. Find
>one without any OS. You won't.

Of course. Why would they make (meaning provide, support, and advertise)
PCs with an OS that customers don't want? Linux is publicized everywhere.
Yet, those stupid consumers - they are just too ignorant to know they really
want Linux.

>> Like I said, Windows and UNIX is the vast majority of server OSs.

>Are they? The major of server applications are in COBOL and they run on
>Mainframes.

I'll agree most or at least a significant percentage of COBOL apps run on
mainframes. A sizable amount of them run on minis, which run UNIX or some
proprietary non-UNIX OS. But COBOL apps no longer constitute the majority
of existing apps. Lines of code maybe, since 4GL languages have less lines
of code for the same app. But there are more VB apps than COBOL. And more
VB server apps than Java server apps in corporate America. Why? because
it works best and it is chosen because it produces results.

>Good as gone. I guess that means OS2 isn't gone either.

A lot more VB6 currently in use than OS2, which is about as rare as a 5 leaf
clover. OS2 is the perfect example of why IBM is a software midget outside
the mainframe world.

>Try doing MVC.

What is MVC? Microsoft Visual C? Define what *application* you would suggest
cannot be done in VB, but can be in Java.

>If OS2 is going like gangbusters then yes.

And we know it's not. The main use for OS2 machines today is boat anchors.

>I didn't mention any Java beans. Obviously you are ill informed in Java
>and OpenSource.

OpenSource is useless. As for Java, take J2EE, JDO, or whatever the Java
emphasis of the day is. It is still a poor performer. Not because it can't
be, but because Sun has done such a poor job with it.

>>I would recommend taking a look at http://www.objectwatch.com/, run by


>>one of CORBA's founders, Roger Sessions. He is an expert on both sides,


>>and gives good advice on both.

>I don't use CORBA. Maybe that is the problem.

I was talking about Java. Had you checked out his web site, you would see
a numebr of good articles he has done on Java. So if you don't use CORBA,
what do you use, Java RMI?

>Ok. I am constant looking at trends. Linux use is on the rise. So is


>Java.

Not in the US. For example, if something has 1% of market share, and rises
to 3% of market share, that is accurately said to be a 200% increase. If
the major competitor in the field has 45% of market share, and gains 3% to
48% (1.35% growth), it can accurately be said the former grew faster than
the latter. But given the actual market share, it is irrelevant. In general,
where profits matter, MS is the technology of choice. Where profits don't
matter, it's Java, UNIX, etc.

Personally, I think DevX, or somebody like Fawcette Publications that is
neutral to MS and Sun, could get a lot of publicity having a VB.NET versus
Java contest (employees or contractors of MS and SUN would be excluded).
The top 5 or 10 winners on each side could be brought to Comdex or some
such event to display the apps, and then name a winner. Of course, with
Sun's posse of lawyers, there are legal considerations when Java loses.