Dot Net on non-windows OS

[Guest]

I seems to me that in order for microsoft to survive past the windows os,
that they need to support DOT.NET on other platforms. Wasn't this the
rationel for DOT.NET in the first place, To break the dependancy on the
window os. I think they should take a direct role in bring dot net the mac
and linux. If Sun did one thing right, it was to port the java VM to all
popular platforms, insuring its wide spread use. Microsoft should do the
same. They should take an active role in this and not depend on the Mono
project to do it for them. Does anyone know if this is being address at
microsoft?

 

[Jon Skeet [C# MVP]]

<=?Utf-8?B?SmVmZnJleSBIYXVu?= <Jeffrey

I seems to me that in order for microsoft to survive past the windows os,
that they need to support DOT.NET on other platforms.

I don't think MS is trying to survive "past the Windows OS" - they're
trying to make sure that Windows keeps going.

Wasn't this the
rationel for DOT.NET in the first place, To break the dependancy on the
window os.

Not really, no. It helps to bridge the 32-bit/64-bit divide, of course,
but I don't think OS-independence was one of the major reasons behind
..NET.

I think they should take a direct role in bring dot net the mac
and linux. If Sun did one thing right, it was to port the java VM to all
popular platforms, insuring its wide spread use.

Sun obviously had to port Java to Windows - it would have been fatal
not to. It was a long time before they officially supported the Linux
port though - for years it came from blackdown.org.

Microsoft should do the
same. They should take an active role in this and not depend on the Mono
project to do it for them. Does anyone know if this is being address at
microsoft?

I doubt that it's being addressed very much, apart from by the Rotor
project. I haven't seen any evidence that MS cares very much about .NET
apps running on other operating systems.

 

[Davy Brion]

I seems to me that in order for microsoft to survive past the windows os,
that they need to support DOT.NET on other platforms. Wasn't this the
rationel for DOT.NET in the first place, To break the dependancy on the
window os. I think they should take a direct role in bring dot net the mac
and linux. If Sun did one thing right, it was to port the java VM to all
popular platforms, insuring its wide spread use. Microsoft should do the
same. They should take an active role in this and not depend on the Mono
project to do it for them. Does anyone know if this is being address at
microsoft?

http://www.mono-project.com

I've used Mono and some parts of it are very good, others aren't quite
there yet. As soon as they have a working System.Windows.Forms version
this could be great for windows developers that want to get their
applications running on Linux. Provided that Microsoft doesn't shut
them down...

 

[Tom Dacon]

I think Mono is always going to be kind of lame. The Mono people are always
going to be playing catch-up, and they're doing it without any access to the
MS development team's strategic roadmap for where they want to take .Net.

Miguel de Acaza is a real sharp guy, I have no doubt, and it's undoubtedly a
greatly challenging and ambitious goal to produce an open-source
multi-platform implementation of .Net. But for the rank and file people who
work on it, it's always going to be a case of following along in someone
else's shadow, doing what they do, but always being a year or more behind.
By the time you get something working according to the current release of
the MS .Net stuff, a new release from MS is on the horizon or has already
arrived. Myself, I think it'd be pretty demoralizing work.

I remember the hard work that went into the VB6-level Common Control
Replacement Project, by a bunch of top-flight VB developers some years ago.
When .Net came along, it kind of knocked the stuffing out of them. Just
think what's going to happen to Mono when Avalon and XAML come along.

Go to http://www.mono-project.com/about/index.html and follow the Blogs link
and do some reading about what the mono people are saying about the process.
Maybe it's just me, but it doesn't sound like a lot of fun.

Too bad...
Tom Dacon
Dacon Software Consulting

 

[Cor Ligthert]

Jeffrey,

I think that your message show only the importance of Microsoft platforms in
the world.

When Sun would not made there software fittable for that, we had a complete
different situation and probably Java was already long gone as something
that was nice trying.

However just my thought,

Cor

 

[Guest]

I don't think MS is trying to survive "past the Windows OS" - they're
trying to make sure that Windows keeps going.


Not really, no. It helps to bridge the 32-bit/64-bit divide, of course,
but I don't think OS-independence was one of the major reasons behind
..NET.

That being said, then why bother with bytecode in the first place. the only
reason for putting up with the loss of proformance (compared to native) was
that you didn't need to recompile for run the came code on different
processors and OSes.

I doubt that it's being addressed very much, apart from by the Rotor
project. I haven't seen any evidence that MS cares very much about .NET
apps running on other operating systems.

Seems to me that is a loss of potential market space. There are still people
out there that run non windows OSes. (Solaris, Iris and other *nix variants)
Mostly in High performance applications. I worked for Nova Information
System, They are one of the largest credit card processors in the US and
while they ran the usual office app on Windows. They ran all their credit
card operations on 2 hugh Sun systems. Microsoft is not likely to convert
them over to windows anytime soon, but the advantages of the DOT NET platform
for inhouse developement might sell.

--

Just my humble opinion


Jeff

 

[UAError]

I seems to me that in order for microsoft to survive past the windows os,
that they need to support DOT.NET on other platforms. Wasn't this the
rationel for DOT.NET in the first place, To break the dependancy on the
window os. I think they should take a direct role in bring dot net the mac
and linux. If Sun did one thing right, it was to port the java VM to all
popular platforms, insuring its wide spread use. Microsoft should do the
same. They should take an active role in this and not depend on the Mono
project to do it for them. Does anyone know if this is being address at
microsoft?

Has it ever occured to you that you may not be thinking big
enough?

DotNet isn't divorcing itself from Windows, however each
successive release of the Framework will remove itself
further from the Win API.

Yes, the Alpha Win NT effort wasn't widely adopted but that
doesn't mean Microsoft has, in the long run, abandonded the
goal of Windows as a multi hardware-platform OS.

Who is to say that DotNet and .NET Server 2003 are not only
the first tiny steps in a long term effort to minimize the
footprint of the hardware dependent Windows kernel?

Once that is achieved, "Windows" (especially for the server
versions) may be in a better position to move to alternate
hardware platforms. Once a certain acceptance is achieved
with a particular platform, alternate, more elaborate
kernels may be used to improve performance to consolidate
the foothold.

If that is Mirosoft's ultimate goal, why would they bother
diluting their efforts with multi-OS versions of their
virtual machine and framework?

Could it work? Maybe, but there are a lot of hurdles along
the way.

 

[Jon Skeet [C# MVP]]

That being said, then why bother with bytecode in the first place. the only
reason for putting up with the loss of proformance (compared to native) was
that you didn't need to recompile for run the came code on different
processors and OSes.

No, that's not the only reason:

1) MSIL allows for verification of code
2) An intermediate bytecode form allows JITs to improve over time, to
achieve *better* than native but compiled for lowest common denominator
code
3) Running within a VM effectively allows all kinds of extra safety
checks to be automatically inserted at runtime

That's just 3. Anyway, you're tying OS and processor independence
together - I suggested that processor independence was probably fairly
important, given the move to 64 bit processors, but that OS
independence isn't nearly as important.

Seems to me that is a loss of potential market space. There are still people
out there that run non windows OSes. (Solaris, Iris and other *nix variants)
Mostly in High performance applications. I worked for Nova Information
System, They are one of the largest credit card processors in the US and
while they ran the usual office app on Windows. They ran all their credit
card operations on 2 hugh Sun systems. Microsoft is not likely to convert
them over to windows anytime soon, but the advantages of the DOT NET platform
for inhouse developement might sell.

Sell what though? I don't believe MS make much money on development -
they make money on people who *do* convert over to Windows. Although
some people obviously won't, others may well.

 

[Erik Vandamme]

Tom,
Please do a Google for myXAML, a re-think
Erik Vandamme

 

[Tom Dacon]

Thanks for the tip. I'll take a peek.

Tom

 

[Guest]

UAError, you've hit the nail on the head there.

..Net is Microsoft's next generation stuff, design to keep them in control.
i would say just forget about .Net ever being multi-platform, however
something to look forward to is the opening up of the Windows source.
European courts have passed it as part of thier monopoly stuff.