OT: .Net Future

I thought this was rather interesting:

http://www.microsoft-watch.com/artic...1820607,00.asp

David Guest wrote:

> I thought this was rather interesting:
>
> http://www.microsoft-watch.com/artic...1820607,00.asp

David, just to put that article into correct light.

I have never heard before that Microsoft claimed Longhorn is built on .NET.
That's simply a wrong claim, and since the whole article is based on this
claim, the article is nil.

Just my opinion about it, no pun intended David

--

Hannes Danzl [NexusDB Developer]
Newsgroup archive at http://www.tamaracka.com/search.htm

Hannes,

> I have never heard before that Microsoft claimed Longhorn is built on
> .NET.
> That's simply a wrong claim, and since the whole article is based on this
> claim, the article is nil.

http://msdn.microsoft.com/longhorn/d...n/default.aspx

....OK, so it doesn't directly say ".NYET" ....just "managed code" ...the
intention is clear though. This is just another back-pedal after the
postponment of WinFS.

Eryk

Eryk Bottomley wrote:

> Hannes,
>
> > I have never heard before that Microsoft claimed Longhorn is built on
> > .NET.
> > That's simply a wrong claim, and since the whole article is based on this
> > claim, the article is nil.
>
> http://msdn.microsoft.com/longhorn/d...ssues/04/01/De
> velopingAppsforLonghorn/default.aspx
>
> ...OK, so it doesn't directly say ".NYET" ....just "managed code" ...the
> intention is clear though. This is just another back-pedal after the
> postponment of WinFS.

i agree on that, but:

"It is the first operating system built with managed code ..."

that's a long way from "built on top of the .Net Framework". I read it as
"makes (heavy) use of managed code" which it still will do. There was always
the plan to have heavily optimized native core services (system, data, ui,
....) that are the basis of the whole operating system and .net as the access
layer on top of it replacing the native api. thus applications and thus most
parts of the OS are written using managed code. After all the core of the OS
itself is not very large, maybe some 10 to 15 megs native code (without the
drivers)

--

Hannes Danzl [NexusDB Developer]
Newsgroup archive at http://www.tamaracka.com/search.htm

> >
http://msdn.microsoft.com/longhorn/d...ssues/04/01/De
> > velopingAppsforLonghorn/default.aspx
> >
> > ...OK, so it doesn't directly say ".NYET" ....just "managed code" ...the
> > intention is clear though. This is just another back-pedal after the
> > postponment of WinFS.
>
> i agree on that, but:
>
> "It is the first operating system built with managed code ..."
>
> that's a long way from "built on top of the .Net Framework". I read it as
> "makes (heavy) use of managed code" which it still will do. There was
always
> the plan to have heavily optimized native core services (system, data, ui,
> ...) that are the basis of the whole operating system and .net as the
access
> layer on top of it replacing the native api. thus applications and thus
most
> parts of the OS are written using managed code. After all the core of the
OS
> itself is not very large, maybe some 10 to 15 megs native code (without
the
> drivers)

But reading the rest of the article makes it pretty clear that making .net
central to it was a pretty clear intention, especially for security
purposes.

That's the way I read it, anyway.

Lauchlan M

Hannes,

>> After all the core of the OS
>> itself is not very large, maybe some 10 to 15 megs native code
That's the biggest "virus" in the world ... <g>

Bert.

"Hannes Danzl[NDD]" <hannes@nexusdb.dbnexus.com> wrote in message
news:xn0e2qy4a8ddrva035@news.nexusdb.com...
> Eryk Bottomley wrote:
>
>> Hannes,
>>
>> > I have never heard before that Microsoft claimed Longhorn is built on
>> > .NET.
>> > That's simply a wrong claim, and since the whole article is based on
>> > this
>> > claim, the article is nil.
>>
>> http://msdn.microsoft.com/longhorn/d...ssues/04/01/De
>> velopingAppsforLonghorn/default.aspx
>>
>> ...OK, so it doesn't directly say ".NYET" ....just "managed code" ...the
>> intention is clear though. This is just another back-pedal after the
>> postponment of WinFS.
>
> i agree on that, but:
>
> "It is the first operating system built with managed code ..."
>
> that's a long way from "built on top of the .Net Framework". I read it as
> "makes (heavy) use of managed code" which it still will do. There was
> always
> the plan to have heavily optimized native core services (system, data, ui,
> ...) that are the basis of the whole operating system and .net as the
> access
> layer on top of it replacing the native api. thus applications and thus
> most
> parts of the OS are written using managed code. After all the core of the
> OS
> itself is not very large, maybe some 10 to 15 megs native code (without
> the
> drivers)
>
> --
>
> Hannes Danzl [NexusDB Developer]
> Newsgroup archive at http://www.tamaracka.com/search.htm

Hannes Danzl[NDD] wrote:
>
> i agree on that, but:
>
> "It is the first operating system built with managed code ..."
>
> that's a long way from "built on top of the .Net Framework". I read it as
> "makes (heavy) use of managed code" which it still will do. There was always
> the plan to have heavily optimized native core services (system, data, ui,
> ...) that are the basis of the whole operating system and .net as the access
> layer on top of it replacing the native api. thus applications and thus most
> parts of the OS are written using managed code. After all the core of the OS
> itself is not very large, maybe some 10 to 15 megs native code (without the
> drivers)
>

It was hinted it would be computability/legacy layer providing
the Win32 API while underneath the hood it would be all .NET
managed code. So Win32 would in effect become a second class
citizen and .NET code would be more 'native' to the OS.

Now it's pretty obvious .NET will continue to be a layer above
the Win32 API in most cases. Over time various APIs will get
pure .NET managed versions. Also some new APIs might only
come out on the .NET side leaving Win32 to need call into
..NET code for that functionality. Long term this will lead
to Win32 being less preferred and .NET underpinning a lot
of functionality. Thing is some thought this would all
happen with the next release of Windows. Reality seems to
be give it 5 to 10 years.

A lot of supporting apps probably won't be rewritten to .NET
just for the sake of doing so as they already have enough
real work to do. The Win32 API itself will be supported
in some way for at least another 50 years and most likely
longer even if in some form of emulation or virtualization.

After getting sucked on to the 'everything will be Java'
train years ago I sure am not going to jump on a
'everything will be .NET' train now. This time I'll
wait till the train goes somewhere and ignore how
shiny and filled with potential it is. Java stations
and Java office suites were all shiny new and filled
with potential to revolutionize things but both never
never went anywhere in mainstream use.

Brian Evans [NDX]

Brian Evans [NDX] wrote:

> Hannes Danzl[NDD] wrote:
>
>>
>> i agree on that, but:
>>
>> "It is the first operating system built with managed code ..."
>>
>> that's a long way from "built on top of the .Net Framework". I read
>> it as
>> "makes (heavy) use of managed code" which it still will do. There was
>> always
>> the plan to have heavily optimized native core services (system,
>> data, ui,
>> ...) that are the basis of the whole operating system and .net as the
>> access
>> layer on top of it replacing the native api. thus applications and
>> thus most
>> parts of the OS are written using managed code. After all the core of
>> the OS
>> itself is not very large, maybe some 10 to 15 megs native code
>> (without the
>> drivers)
>>
>
> It was hinted it would be computability/legacy layer providing
> the Win32 API while underneath the hood it would be all .NET
> managed code. So Win32 would in effect become a second class
> citizen and .NET code would be more 'native' to the OS.
>
> Now it's pretty obvious .NET will continue to be a layer above
> the Win32 API in most cases. Over time various APIs will get
> pure .NET managed versions. Also some new APIs might only
> come out on the .NET side leaving Win32 to need call into
> .NET code for that functionality. Long term this will lead
> to Win32 being less preferred and .NET underpinning a lot
> of functionality. Thing is some thought this would all
> happen with the next release of Windows. Reality seems to
> be give it 5 to 10 years.
>
> A lot of supporting apps probably won't be rewritten to .NET
> just for the sake of doing so as they already have enough
> real work to do. The Win32 API itself will be supported
> in some way for at least another 50 years and most likely
> longer even if in some form of emulation or virtualization.
>
> After getting sucked on to the 'everything will be Java'
> train years ago I sure am not going to jump on a
> 'everything will be .NET' train now. This time I'll
> wait till the train goes somewhere and ignore how
> shiny and filled with potential it is. Java stations
> and Java office suites were all shiny new and filled
> with potential to revolutionize things but both never
> never went anywhere in mainstream use.
>
> Brian Evans [NDX]

I have seen some evidence of this already. The Ink Aware components
which are available for writing applications for the tablet PC are only
available to run in a .NET environment. Thus they can't be used in
win32 applications...