If .net and j2ee are so great...

[43]

how come m$Office isn't written in .net?

how come Open Office isn't written in j2ee?

how come dbms systems aren't written in either?

how come browsers aren't written in either?

how come RealPlayer, MediaPlayer and all applications
that need speed are written in c++ ? (except for
LimeWire, which is a piece of shit ).

how come Linux and all Linux applications are written
in c/c++ ?

--
http://geeks4dean.com Dean '04
Linux - Where the Beat, Meet the Elite

[43]

Roedy Green:

On Thu, 21 Aug 2003 05:22:49 GMT, 43 <ja*****@earthlink.net> wrote or
quoted :

how come m$Office isn't written in .net?

how come Open Office isn't written in j2ee?

Because these are both server side technologies. Editing a document
is a solitary activity. It works best with instantaneous response to
each keystroke.

they are not 'server side'

web services are 'server side'

windows client and console applications are client side.

java applications and applets are both client side.

net beans can be either. rmi means strong clients and servers.

bottom line .net/j2ee can't write the p2p applications
that are going to dominate the next 10 years.

--
http://geeks4dean.com Dean '04
Linux - Where the Beat, Meet the Elite

[Chad Myers]

"Roedy Green" <ro***@mindprod.com> wrote in message
news:hm********************************@4ax.com...

On Thu, 21 Aug 2003 05:22:49 GMT, 43 <ja*****@earthlink.net> wrote or
quoted :

how come m$Office isn't written in .net?

how come Open Office isn't written in j2ee?

Because these are both server side technologies. Editing a document
is a solitary activity. It works best with instantaneous response to
each keystroke.

Actually,there is a good chunk of Office 2003 in .NET and it will
be moving to completely .NET switfly like most if not all MS products.

Windows 2003 Server depends heavily on the .NET Framework (whack your
c:\windows\assembly folder in Safe Mode and reboot and check the
Event Viewer).

I recall hearing a statistic that in Longhorn, large chunks of the
Win32 API are being moved to managed code. The API will go from
something
like 5x,xxx functions to 8,xxx functions with the difference existing
in new OO .NET assemblies.

Most of the new server products coming out require the Framework to
be installed. They are either entirely written in .NET (CMS, BizTalk),
or
partially (Commerce Server), or have strong ties (Win2K3 Server,
Exchange 2003, etc)

-c

[Chad Myers]

"43" <ja*****@earthlink.net> wrote in message
news:pa****************************@earthlink.net. ..

Roedy Green:

On Thu, 21 Aug 2003 05:22:49 GMT, 43 <ja*****@earthlink.net> wrote or quoted :

how come m$Office isn't written in .net?

how come Open Office isn't written in j2ee?

Because these are both server side technologies. Editing a document
is a solitary activity. It works best with instantaneous response to each keystroke.

they are not 'server side'

web services are 'server side'

windows client and console applications are client side.

java applications and applets are both client side.

net beans can be either. rmi means strong clients and servers.

bottom line .net/j2ee can't write the p2p applications
that are going to dominate the next 10 years.

..NET certainly can. I don't know about Java/J2EE, I won't comment
in that regard, but .NET certainly can.

In fact, MS Research has a rather large P2P project implemented
in the .NET Framework and it's pretty impressive.

-c

[Tim Tyler]

In comp.lang.java.advocacy 43 <ja*****@earthlink.net> wrote:
: 43 <ja*****@earthlink.net> wrote or quoted :

:>>how come m$Office isn't written in .net?
:>>
:>>how come Open Office isn't written in j2ee?
:>
:> Because these are both server side technologies. Editing a document
:> is a solitary activity. It works best with instantaneous response to
:> each keystroke.

: they are not 'server side'

J2EE is - as much as, say PHP is.
--
__________
|im |yler http://timtyler.org/ ti*@tt1.org

[Moshe Sayag]

..Not is not a technology but a marketing crap (AKA markitecture).
Converting M$Office to the .Not environment is easy... just start market
it as M$Office.Not
43 wrote:

how come m$Office isn't written in .net?

how come Open Office isn't written in j2ee?

how come dbms systems aren't written in either?

how come browsers aren't written in either?

how come RealPlayer, MediaPlayer and all applications
that need speed are written in c++ ? (except for
LimeWire, which is a piece of shit ).

how come Linux and all Linux applications are written
in c/c++ ?

[Mark Marsella]

Look what Moshe Sayag wrote in article <bi**********@new-usenet.uk.sun.com>,
on Thu, 21 Aug 2003 11:24:33 +0200

.Not is not a technology but a marketing crap (AKA markitecture).
Converting M$Office to the .Not environment is easy... just start market
it as M$Office.Not
43 wrote:

how come m$Office isn't written in .net?

how come Open Office isn't written in j2ee?

You show your lack of knowledge, .Net is a fully developed technology
designed by the best. C# for example is a programmers dream lang.
--
Thanks
Mark Marsella

[mlw]

43 wrote:

how come m$Office isn't written in .net?
I bet parts of it will be in the near future, but M$ office is so slow aqnd
disgusting that it doesn't make a difference. It will be crap, but people
will still get it on their systems and still have to pay for it.
how come Open Office isn't written in j2ee?
Because Java, and languages like it, including .NET, are interpreted
environments. They may play tricks with JITs "just in time" compilers, but
make no mistake, it is an interpreted environment.

C/C++ are not interpreted, they are compiled to machine code. Sure there are
a few theoretical examples of how a "JIT" enabled language may be able to
be faster than a compiled one, but for the most part both .NET and Java
have a huge amount of overhead and enforce an object model and programming
model that is very inefficient.

Say what you will, a well written C/C++ program will always be faster and
use less memory than a well written Java/.NET program because there is less
overhead.
how come dbms systems aren't written in either?
Similarly, a DBMS is a complex animal. Most DBMS designs require very
carefull control over memory, disk access, arrays, data types, etc. Items
in a database have to be as efficient as possible. A byte or two or eight,
dedicated to the programming language, would cause the system to choose a
disk based algorithm over a memory based one. At that point, you just lost
to your competitor.
how come browsers aren't written in either?
Anyone remember the "HotJava" browser? It was junk.
how come RealPlayer, MediaPlayer and all applications
that need speed are written in c++ ? (except for
LimeWire, which is a piece of shit ). Actually, I like LimeWire
how come Linux and all Linux applications are written
in c/c++ ?

Because all the systems come with these compilers, you have to get Java
elsewhere (usually)

Hey, Java and .NET, like VisualBasic and COBOL, are high level languages
intended for inexperienced developers or business applications. As such,
they are fine for things like LimeWire, Instant Message clients, and
perhaps some business logic stuff.

If you are going to write a system which needs 24x7 uptime, uses memory
extensively, needs to be fast, or needs to be efficient, those environments
just will not cut it. The VMs are huge to begin with and you have no real
control over memory or the specifics of your process. The only alternative
is C/C++.

[mlw]

Mark Marsella wrote:

Look what Moshe Sayag wrote in article
<bi**********@new-usenet.uk.sun.com>, on Thu, 21 Aug 2003 11:24:33 +0200

.Not is not a technology but a marketing crap (AKA markitecture).
Converting M$Office to the .Not environment is easy... just start market
it as M$Office.Not
43 wrote:

> how come m$Office isn't written in .net?
>
> how come Open Office isn't written in j2ee?
>

You show your lack of knowledge, .Net is a fully developed technology
designed by the best. C# for example is a programmers dream lang.

Having programmed and designed software and hardware for over 20 years, I
can honestly say .NET is a dreadful language and environment. Not only is
it a virtual copy of Java with just enough syntactic differences to be
annoying, its so much like Java that it is not usable for non-trivial
applications.

Interpreted languages will *NEVER* be sufficient for hard core performance
or 24x7x365 reliability. For these types of applications, you need serious
control over every aspect of the application, which these system do not
provide.

[Tim Tyler]

mlw <ml*@nospam.no> wrote:

: Having programmed and designed software and hardware for over 20 years, I
: can honestly say .NET is a dreadful language and environment. Not only is
: it a virtual copy of Java with just enough syntactic differences to be
: annoying, its so much like Java that it is not usable for non-trivial
: applications.

: Interpreted languages will *NEVER* be sufficient for hard core performance
: or 24x7x365 reliability. [...]

You may have a long track record - but if you think Java and .NET are
interpreted... I reckon you are probably in need of a refresher course.
--
__________
|im |yler http://timtyler.org/ ti*@tt1.org

[Tim Tyler]

mlw <ml*@nospam.no> wrote:

[Re: how come Open Office isn't written in j2ee?]

: Because Java, and languages like it, including .NET, are interpreted
: environments. They may play tricks with JITs "just in time" compilers, but
: make no mistake, it is an interpreted environment.

Uh - no they aren't. C'mon - get with the program here.

Java bytecodes were interpreted back in 1995. Things have moved on since then.
--
__________
|im |yler http://timtyler.org/ ti*@tt1.org

[asj]

j2ee is a SERVER-SIDE specification. it would be pretty hard to write
a desktop application using serverside implementations.

..net however, is pretty much EVERYTHING on microsoft's side, so they
are not actually directly comparable, and i do expect .net stuff to be
coming out on desktops in large numbers.
43 <ja*****@earthlink.net> wrote in message news:<pa****************************@earthlink.net >...

how come m$Office isn't written in .net?

how come Open Office isn't written in j2ee?

how come dbms systems aren't written in either?

how come browsers aren't written in either?

how come RealPlayer, MediaPlayer and all applications
that need speed are written in c++ ? (except for
LimeWire, which is a piece of shit ).

how come Linux and all Linux applications are written
in c/c++ ?

[mlw]

Tim Tyler wrote:

mlw <ml*@nospam.no> wrote:

[Re: how come Open Office isn't written in j2ee?]

: Because Java, and languages like it, including .NET, are interpreted
: environments. They may play tricks with JITs "just in time" compilers,
: but make no mistake, it is an interpreted environment.

Uh - no they aren't. C'mon - get with the program here.

Java bytecodes were interpreted back in 1995. Things have moved on since
then.

So, are you saying java programs are native binaries now? I can open one up
and see native x86 code? of course not, they are tokenized source. So much
so that you can dump out the originating source from a decompiler (sans
comments).

Not everything is JITed, and JITing takes time. It is a high level
interpretor, to claim otherwise is really being misinformed.

[John Straumann]

This is an idiotic question, and Erik's response is pretty close to teh
mark. .NET is a new technology and MS Office has been around for years. What
would the value to Microsoft be in committing teams of programmers to
rewrite MS Office in .NET? The same goes for SQL server or any other
exisiting product.

..NET is a fantastic platform for distributed development, far superior to
J2EE (and I've used both extensively)...

John.

"43" <ja*****@earthlink.net> wrote in message
news:pa****************************@earthlink.net. ..

how come m$Office isn't written in .net?

how come Open Office isn't written in j2ee?

how come dbms systems aren't written in either?

how come browsers aren't written in either?

how come RealPlayer, MediaPlayer and all applications
that need speed are written in c++ ? (except for
LimeWire, which is a piece of shit ).

how come Linux and all Linux applications are written
in c/c++ ?

--
http://geeks4dean.com Dean '04
Linux - Where the Beat, Meet the Elite

[Michael A. Covington]

"Roedy Green" <ro***@mindprod.com> wrote in message
news:hm********************************@4ax.com...

On Thu, 21 Aug 2003 05:22:49 GMT, 43 <ja*****@earthlink.net> wrote or
quoted :

how come m$Office isn't written in .net?

how come Open Office isn't written in j2ee?

Because these are both server side technologies. Editing a document
is a solitary activity. It works best with instantaneous response to
each keystroke.

I think you are confusing .NET with ASP.NET. .NET is an object-oriented
API that has no particular association with servers. It's a great way to
write applications.

The reason MS Office isn't written in .NET has something to do with MS
Office being invented about 10 years before .NET was invented! Same reason
the battle reports of Julius Caesar aren't written in English...

[Gregory Gadow]

"Michael A. Covington" wrote:

"Roedy Green" <ro***@mindprod.com> wrote in message
news:hm********************************@4ax.com...

On Thu, 21 Aug 2003 05:22:49 GMT, 43 <ja*****@earthlink.net> wrote or
quoted :

how come m$Office isn't written in .net?

how come Open Office isn't written in j2ee?

Because these are both server side technologies. Editing a document
is a solitary activity. It works best with instantaneous response to
each keystroke.

I think you are confusing .NET with ASP.NET. .NET is an object-oriented
API that has no particular association with servers. It's a great way to
write applications.

The reason MS Office isn't written in .NET has something to do with MS
Office being invented about 10 years before .NET was invented! Same reason
the battle reports of Julius Caesar aren't written in English...

The new version of Office, recently unveiled by MS, is apparently written in
..NET.
--
Gregory Gadow
te******@serv.net
http://www.serv.net/~techbear

"If you make yourself a sheep, the wolves will eat you."
-- Benjamin Franklin

[xx]

Tor Iver Wilhelmsen wrote:

mlw <ml*@nospam.no> writes:

Or perhaps you don't know what you think you know. Byte codes are not
machine codes, they are interpreted or "JIT" compiled at runtime, either
way, a Java or .NET program is not in machine form.

The byte codes, whether targeting the CLR or JVM, are turned into
native code fractions of a second after the classes load - from that
point on,it's just as native as a "precompiled to native" C app.

Yes, "from that point on" is a common thing to say, but not very true.
Every time you run the program you have to re-interpret it into machine
code. It is an interpreted program, it matters not how it is interpreted.

If you update your JVM or CLR, it may "interpret" your byte codes
differently. Where as my binary application will execute the same
instructions until there is a new version. A very important distinction.

[Tor Iver Wilhelmsen]

xx <xx*@yyy.zzz> writes:

Yes, "from that point on" is a common thing to say, but not very
true. Every time you run the program you have to re-interpret it
into machine code. It is an interpreted program, it matters not how
it is interpreted.
Then you have a different "interpretation" of interpreted. In fact you
could claim all code is interpreted because the instructions can
trigger microcode which is the actual "code" that the processor
executes.
If you update your JVM or CLR, it may "interpret" your byte codes
differently.
Yes, you might get new runtime optimizations. Which is good.
Where as my binary application will execute the same instructions
until there is a new version. A very important distinction.

No, a 486 does not execute an x86 instruction the same way as a
Pentium IV.

[xx]

Tor Iver Wilhelmsen wrote:

xx <xx*@yyy.zzz> writes:

Yes, "from that point on" is a common thing to say, but not very
true. Every time you run the program you have to re-interpret it
into machine code. It is an interpreted program, it matters not how
it is interpreted.

Then you have a different "interpretation" of interpreted. In fact you
could claim all code is interpreted because the instructions can
trigger microcode which is the actual "code" that the processor
executes.

Ahh, yes, lose the argument, muddy the waters. Yup, it is clear that
Java is "interpreted" so try to make the case that the CPUs just
interpret. Hell, lets just get down to the gates and flipflops in the
CPU shall we? Maybe I'm cynical, but please, these are different issues
and if you don't know that, you ought too.

It is a bogus argument intended to distract from the real issue that
java byte codes are interpreted. Argue all you want, paint as many
shades of gray as you can, but it is still interpreted.

Java byte codes are not native machine codes. When interpreted by one
JVM it may work perfectly. When interpreted by another JVM it may not.
Just look at threading issues across various JVMs, especially older ones.

Why do different JVMs compete? Because there are differences. If your
code were compiled there would be no differences.

That is NOT the behavior of compiled code. Compiled code is translated
into the native instructions of the system and distributed as such.

[Jon Skeet]

xx <xx*@yyy.zzz> wrote:

Yes, "from that point on" is a common thing to say, but not very true.
Every time you run the program you have to re-interpret it into machine
code. It is an interpreted program, it matters not how it is interpreted.
Um, no. Interpreting is *not* the same as JIT compiling. For instance,
in an interpreted system, every time you go through the same loop you
have to do the same work, interpreting the original code and working
out what it means in the native system. With JIT compiling, that's done
*once* for the whole loop, and then it's native code.

Compiling and interpreting are very, very different things - and until
you accept that, you *will* find people disagreeing with you all over
the place just as they are in this thread.
If you update your JVM or CLR, it may "interpret" your byte codes
differently.
You mean it may compile your byte code differently - indeed it might.
Where as my binary application will execute the same
instructions until there is a new version. A very important distinction.

Sure, but not one which means that Java is interpreted these days.

--
Jon Skeet - <sk***@pobox.com>
http://www.pobox.com/~skeet/
If replying to the group, please do not mail me too

[43]

nhoel:

that doesn't mean it's a bad product. Although, I wish they can quickly
come up with the UNIX/LINUX/OS390 versions of the .NET soon so I can program
for these other operating systems quick. But, it's a promise from

here you go:

http://www.dotgnu.org

c# for linux; i installed it months ago.
--
http://geeks4dean.com Dean '04
Linux - Where the Beat, Meet the Elite

[nhoel]

Thanks for the info...

I'll try out my applications if it works there... Thanks again...

-Nhoel-

"43" <ja*****@earthlink.net> wrote in message
news:pa****************************@earthlink.net. ..

nhoel:

that doesn't mean it's a bad product. Although, I wish they can quickly
come up with the UNIX/LINUX/OS390 versions of the .NET soon so I can program for these other operating systems quick. But, it's a promise from

here you go:

http://www.dotgnu.org

c# for linux; i installed it months ago.
--
http://geeks4dean.com Dean '04
Linux - Where the Beat, Meet the Elite

[Simon Cooke]

On Thu, 21 Aug 2003 01:17:54 -0500, Chad Myers wrote:

"Roedy Green" <ro***@mindprod.com> wrote in message
news:hm********************************@4ax.com...

On Thu, 21 Aug 2003 05:22:49 GMT, 43 <ja*****@earthlink.net> wrote or
quoted :

how come m$Office isn't written in .net?

how come Open Office isn't written in j2ee?

Because these are both server side technologies. Editing a document
is a solitary activity. It works best with instantaneous response to
each keystroke.

Actually,there is a good chunk of Office 2003 in .NET and it will
be moving to completely .NET switfly like most if not all MS products.

Windows 2003 Server depends heavily on the .NET Framework (whack your
c:\windows\assembly folder in Safe Mode and reboot and check the
Event Viewer).

I recall hearing a statistic that in Longhorn, large chunks of the
Win32 API are being moved to managed code. The API will go from
something
like 5x,xxx functions to 8,xxx functions with the difference existing
in new OO .NET assemblies.

.... at which point, if the Win32(64) API is not longer native, and they've
not fixed their obvious and blatant performance issues with .NET startup
(plus the fact that the GC is nondeterministic and can get in the way of UI
handling), I'll switch to programming Linux.

I stopped programming in Java because I couldn't stand the amount of memory
it was taking up to run simple apps, or the perf hit. It's the same reason
I won't use C#.

That, and the lack of templates.

Simon

[Tom Shelton]

Simon Cooke wrote:

On Thu, 21 Aug 2003 01:17:54 -0500, Chad Myers wrote:

"Roedy Green" <ro***@mindprod.com> wrote in message
news:hm********************************@4ax.com. ..

On Thu, 21 Aug 2003 05:22:49 GMT, 43 <ja*****@earthlink.net> wrote or
quoted :
how come m$Office isn't written in .net?

how come Open Office isn't written in j2ee?

Because these are both server side technologies. Editing a document
is a solitary activity. It works best with instantaneous response to
each keystroke.
Actually,there is a good chunk of Office 2003 in .NET and it will
be moving to completely .NET switfly like most if not all MS products.

Windows 2003 Server depends heavily on the .NET Framework (whack your
c:\windows\assembly folder in Safe Mode and reboot and check the
Event Viewer).

I recall hearing a statistic that in Longhorn, large chunks of the
Win32 API are being moved to managed code. The API will go from
something
like 5x,xxx functions to 8,xxx functions with the difference existing
in new OO .NET assemblies.

... at which point, if the Win32(64) API is not longer native, and they've
not fixed their obvious and blatant performance issues with .NET startup
(plus the fact that the GC is nondeterministic and can get in the way of UI
handling), I'll switch to programming Linux.

The native interfaces aren't going away, nor the API. Chad seems to be
a little confused here. They are writting managed wrappers around most
of the native interface. By the way, while startup times are not what I
would call stellar - they have improved significantly in .NET 1.1.
Actually, 1.0 wasn't awful - except for that stupid authentication bug
that would cause applications to start really, really slowly if the
ASP.NET user account didn't exist (this might have been fixed in one of
the service packs)....

As for GC... Well, it certainly isn't right for every application - but
the majority of applications do benifit from lazy memory management.
I stopped programming in Java because I couldn't stand the amount of memory
it was taking up to run simple apps, or the perf hit. It's the same reason
I won't use C#.
I don't know if that will ever change... It's the price you pay for all
the extra security checks and the application framework.
That, and the lack of templates.

While not templates, C# is getting generics in v2.0 and Java in 1.5.

Tom Shelton

[Roedy Green]

On Thu, 21 Aug 2003 18:07:35 GMT, Simon Cooke
<si********@eaSPAMMAGErthNOSPAMlink.net> wrote or quoted :

I stopped programming in Java because I couldn't stand the amount of memory
it was taking up to run simple apps, or the perf hit. It's the same reason
I won't use C#.

That can be easily fixed. see http://mindprod.com/jgloss/jet.html
--
Canadian Mind Products, Roedy Green.
Coaching, problem solving, economical contract programming.
See http://mindprod.com/jgloss/jgloss.html for The Java Glossary.

[Roedy Green]

On Thu, 21 Aug 2003 15:51:30 GMT, xx <xx*@yyy.zzz> wrote or quoted :

If you update your JVM or CLR, it may "interpret" your byte codes
differently. Where as my binary application will execute the same
instructions until there is a new version. A very important distinction.

Usually later versions have fewer bugs. The nice thing is you don't
have to recompile anything to get the advantage of all the new bug
fixes and optimisations.

On the other hand, you can see the premature male pattern baldness
coming from programmers dealing with clients who insist on using a
different JVM (Usually the old MS 1.1.4 JVM) than they do.

--
Canadian Mind Products, Roedy Green.
Coaching, problem solving, economical contract programming.
See http://mindprod.com/jgloss/jgloss.html for The Java Glossary.

[Roedy Green]

On Thu, 21 Aug 2003 09:53:55 -0700, "nhoel" <no*****@nomail.com> wrote
or quoted :

One major
disadvantage of a "pure" machine code application is that the entire code is
loaded into memory even if you only have to use 10 of the 50,000 functions
it has.

It depends on how you design your machine code. If you put your
functionality in different DLLs, they won't be loaded until used.

In some operating systems, you can load code merely my mapping the
program into virtual address space. It takes no disk i/o at all.
Only the pages of the program needed are loaded in to real RAM as they
are needed.

--
Canadian Mind Products, Roedy Green.
Coaching, problem solving, economical contract programming.
See http://mindprod.com/jgloss/jgloss.html for The Java Glossary.

[Roedy Green]

On Thu, 21 Aug 2003 09:53:55 -0700, "nhoel" <no*****@nomail.com> wrote
or quoted :

Another problem with machine code applications (but not true with .NET) is
it's easy to be infected by viruses or modified by a malicious program.
They don't have the idea of a "digital signature" that the CLR can use to
validate whether an assembly has been tampered with or not.

Digitally signing can be applied to machine code, Java jar files, or
CLR. Microsoft itself uses it for digitally signed machine code
device drivers to indicate it approves of them.

The overhead to do the checks is becoming less and less onerous with
faster cpus, and the advantages more and more apparent.

Eventually I expect every transmission and file to be routinely
encrypted and signed.

--
Canadian Mind Products, Roedy Green.
Coaching, problem solving, economical contract programming.
See http://mindprod.com/jgloss/jgloss.html for The Java Glossary.

[42]

Gee maybe because .net and j2ee were even a twinkle is your bleary eyes when
MSOffice, Open Office, most dbms systems and most browers were invented long
before.

Office 2003 currently in beta has large of portions of .net code. Not sure
and don't care if the whole app is .net or not.

Duh?, maybe cut back on 40's and up the meds.

"43" <ja*****@earthlink.net> wrote in message
news:pa****************************@earthlink.net. ..

how come m$Office isn't written in .net?

how come Open Office isn't written in j2ee?

how come dbms systems aren't written in either?

how come browsers aren't written in either?

how come RealPlayer, MediaPlayer and all applications
that need speed are written in c++ ? (except for
LimeWire, which is a piece of shit ).

how come Linux and all Linux applications are written
in c/c++ ?

--
http://geeks4dean.com Dean '04
Linux - Where the Beat, Meet the Elite

[43]

42:

Office 2003 currently in beta has large of portions of .net code. Not sure
and don't care if the whole app is .net or not.

ok. but is this a true IAYF implementation?

when these products were a 'twinkle in my eye'
( or a Twinkie in your case ) the goal was to
replace monolithic apps with a set of services.

a universal spell checker. a universal function
eval. and so on.

is this happening? or is it just a camouflage to
re-work monolithic apps in the .Net clothing?

..Net would be the perfect way to do IAYF. Java
would be the perfect way to do IAYF.

where is it?
--
http://geeks4dean.com Dean '04
Linux - Where the Beat, Meet the Elite

[Chad Myers]

"xx" <xx*@yyy.zzz> wrote in message news:3F**************@yyy.zzz...

Tor Iver Wilhelmsen wrote:

mlw <ml*@nospam.no> writes:

Or perhaps you don't know what you think you know. Byte codes are notmachine codes, they are interpreted or "JIT" compiled at runtime, eitherway, a Java or .NET program is not in machine form.

The byte codes, whether targeting the CLR or JVM, are turned into
native code fractions of a second after the classes load - from that
point on,it's just as native as a "precompiled to native" C app.

Yes, "from that point on" is a common thing to say, but not very true.
Every time you run the program you have to re-interpret it into

machine code. It is an interpreted program, it matters not how it is interpreted.

Actually, .NET compilations are usually cached, so they're not
compiled every time, just the once.

..NET is interpreted in the same way that C++ is interpreted into
machine code :)
If you update your JVM or CLR, it may "interpret" your byte codes
differently. Where as my binary application will execute the same
instructions until there is a new version. A very important

distinction.

Really? So all those DLLs you call will never change? You have
external dependencies on the system and the runtime environment.
It just happens that .NET compile code's runtime environment
is more compact and consistent across platforms.

-c

[Chad Myers]

"Tom Shelton" <to*@mtogden.com> wrote in message
news:h4****************@news.uswest.net...

Simon Cooke wrote:

On Thu, 21 Aug 2003 01:17:54 -0500, Chad Myers wrote:

"Roedy Green" <ro***@mindprod.com> wrote in message
news:hm********************************@4ax.com. ..

On Thu, 21 Aug 2003 05:22:49 GMT, 43 <ja*****@earthlink.net> wrote orquoted :
>how come m$Office isn't written in .net?
>
>how come Open Office isn't written in j2ee?

Because these are both server side technologies. Editing a documentis a solitary activity. It works best with instantaneous response toeach keystroke.

Actually,there is a good chunk of Office 2003 in .NET and it will
be moving to completely .NET switfly like most if not all MS products.
Windows 2003 Server depends heavily on the .NET Framework (whack yourc:\windows\assembly folder in Safe Mode and reboot and check the
Event Viewer).

I recall hearing a statistic that in Longhorn, large chunks of the
Win32 API are being moved to managed code. The API will go from
something
like 5x,xxx functions to 8,xxx functions with the difference existingin new OO .NET assemblies.

... at which point, if the Win32(64) API is not longer native, and they've not fixed their obvious and blatant performance issues with .NET startup (plus the fact that the GC is nondeterministic and can get in the way of UI handling), I'll switch to programming Linux.

(I'm replying through Tom, I hope you don't mind)

First, the startup time is fixed in Win2K3 because the the first
time the CLR starts up in the OS, there are several initialization
things that must occur system-wide. Since Win2K3 relies so heavily
on the .NET framework, the CLR is initialized during OS startup.

Longhorn will have the same effect.

As far as the GC, I don't think you understand what non-deterministic
means in this context. It's related to memory management, not
the order in which events occur and such. GC has no effect on
UI development. Have you (Simon) even used .NET for anything? I
doubt it.

The native interfaces aren't going away, nor the API. Chad seems to be a little confused here.

http://www.develop4.net/news.html

"First, the software giant aims to slash the number of API calls in
the Win32 API set from more than 70,000 to fewer than 10,000 to help
developers better exploit the next-generation Windows shell, user
interface (code-named Aero) and .Net framework components in Longhorn,
according to sources familiar with the Longhorn plans.

"Win32 has like 76,000 APIs, and they're taking it down to 8,000 with
Longhorn technology," said one source familiar with the plans.

"Microsoft is expected to release additional details about Longhorn
at the Windows Hardware Engineering (WinHEC) trade show here this
week."

I'm not confused, it's a fact. Some of those APIs will just disappear,
some of them will be manifest in new .NET APIs.

-c

[Chad Myers]

"GreyCloud" <cu*****@mist.com> wrote in message
news:3F***************@mist.com...

Simon Cooke wrote:

<snip>

I stopped programming in Java because I couldn't stand the amount of memory it was taking up to run simple apps, or the perf hit. It's the same reason I won't use C#.

That, and the lack of templates.

Same here... I didn't really like the anonymous handler
style... to me it was rather cumbersome and vague at first
glance.

Uh oh, anonymous methods are coming to C#. I personally
disagree with the reasons why seeing their abuse in Java,
but whatever.

I'm still not sure about partial types either, it seems
like it'll just lead to confusion and abuse as well,
with very little benefit

-c

[Tom Shelton]

Chad Myers wrote:

"Tom Shelton" <to*@mtogden.com> wrote in message
news:h4****************@news.uswest.net...

Simon Cooke wrote:

On Thu, 21 Aug 2003 01:17:54 -0500, Chad Myers wrote:

"Roedy Green" <ro***@mindprod.com> wrote in message
news:hm********************************@4ax.co m...
>On Thu, 21 Aug 2003 05:22:49 GMT, 43 <ja*****@earthlink.net> wrote
or
quoted :
>
>
>
>>how come m$Office isn't written in .net?
>>
>>how come Open Office isn't written in j2ee?
>
>Because these are both server side technologies. Editing a
document
is a solitary activity. It works best with instantaneous response
to
each keystroke.

Actually,there is a good chunk of Office 2003 in .NET and it will
be moving to completely .NET switfly like most if not all MS
products.
Windows 2003 Server depends heavily on the .NET Framework (whack
your
c:\windows\assembly folder in Safe Mode and reboot and check the
Event Viewer).

I recall hearing a statistic that in Longhorn, large chunks of the
Win32 API are being moved to managed code. The API will go from
something
like 5x,xxx functions to 8,xxx functions with the difference
existing
in new OO .NET assemblies.
... at which point, if the Win32(64) API is not longer native, and
they've
not fixed their obvious and blatant performance issues with .NET
startup
(plus the fact that the GC is nondeterministic and can get in the
way of UI
handling), I'll switch to programming Linux.

(I'm replying through Tom, I hope you don't mind)

First, the startup time is fixed in Win2K3 because the the first
time the CLR starts up in the OS, there are several initialization
things that must occur system-wide. Since Win2K3 relies so heavily
on the .NET framework, the CLR is initialized during OS startup.

Longhorn will have the same effect.

As far as the GC, I don't think you understand what non-deterministic
means in this context. It's related to memory management, not
the order in which events occur and such. GC has no effect on
UI development. Have you (Simon) even used .NET for anything? I
doubt it.

The native interfaces aren't going away, nor the API. Chad seems to

be

a little confused here.

http://www.develop4.net/news.html

"First, the software giant aims to slash the number of API calls in
the Win32 API set from more than 70,000 to fewer than 10,000 to help
developers better exploit the next-generation Windows shell, user
interface (code-named Aero) and .Net framework components in Longhorn,
according to sources familiar with the Longhorn plans.

"Win32 has like 76,000 APIs, and they're taking it down to 8,000 with
Longhorn technology," said one source familiar with the plans.

"Microsoft is expected to release additional details about Longhorn
at the Windows Hardware Engineering (WinHEC) trade show here this
week."

I'm not confused, it's a fact. Some of those APIs will just disappear,
some of them will be manifest in new .NET APIs.

-c

I'm looking at the article now. Thanks. I hadn't seen this before. I
was under the impression that they were simply wrapping the native
interface more completly. By the way, isn't Longhorn going to be a
64-bit OS?

Tom Shelton

[Jeff Relf]

Hi 43 , you say :
" How come Linux and all Linux applications
are written in c/c++ ? "
Scripts are for quick and dirty jobs . Simple Jobs .

C++ is for Large serious jobs . Complex Jobs .

[Roedy Green]

On Fri, 22 Aug 2003 00:05:47 -0500, "Chad Myers"
<cm****@N0.SP.AM.austin.rr.com> wrote or quoted :

Uh oh, anonymous methods are coming to C#. I personally
disagree with the reasons why seeing their abuse in Java,
but whatever.

there are no anonymous methods in Java. There are anonymous inner
classes however. Are you sure you feel qualified to judge something
you so obviously have never used?

--
Canadian Mind Products, Roedy Green.
Coaching, problem solving, economical contract programming.
See http://mindprod.com/jgloss/jgloss.html for The Java Glossary.

[Jeff Relf]

Oops , I said :
" P2P devices , like the Blueberry ... "

The Blueberry requires a network hierarchy .

A CB is more like P2P ... and even that requires a license .

Are walkie talkies the main P2P devices today ?

[43]

Jeff Relf:

Oops , I said :
" P2P devices , like the Blueberry ... "

The Blueberry requires a network hierarchy .

A CB is more like P2P ... and even that requires a license .

Are walkie talkies the main P2P devices today ?

walkie talkies are a P2P network where n=2.

if you're walkie talkie could allow other walkie talkies
to 'island hop' their signal on part of your
system, then n would be higher.

[Jon Skeet]

Chad Myers <cm****@N0.SP.AM.austin.rr.com> wrote:

Uh oh, anonymous methods are coming to C#. I personally
disagree with the reasons why seeing their abuse in Java,
but whatever.
Anonymous *classes* in Java are indeed a pain (IMO) - but I think
anonymous delegates will be somewhat better. There's less scope for
abuse, I believe. We'll see though :)
I'm still not sure about partial types either, it seems
like it'll just lead to confusion and abuse as well,
with very little benefit

I don't know - I think there's clear benefit in the situation where
some code is machine generated and some isn't. That's the *only* time I
would use them though.

--
Jon Skeet - <sk***@pobox.com>
http://www.pobox.com/~skeet/
If replying to the group, please do not mail me too

[Tim Tyler]

In comp.lang.java.advocacy Chad Myers <cm****@n0.sp.am.austin.rr.com> wrote:

: It just happens that .NET compile code's runtime environment
: is more compact and consistent across platforms.

It more consistently doesn't exist if that's what you mean:

Platform .NET
Windows Y
Linux N
Solaris N
FreeBSD N
Mac OS N

"More compact" than what? It's a /lot/ bigger than the JRE.
--
__________
|im |yler http://timtyler.org/ ti*@tt1.org

[asj]

mlw <ml*@nospam.no> wrote in message news:<5g11b.216055$uu5.40024@sccrnsc04>...

43 wrote:
Hey, Java and .NET, like VisualBasic and COBOL, are high level languages
intended for inexperienced developers or business applications. As such,
they are fine for things like LimeWire, Instant Message clients, and
perhaps some business logic stuff.

"perhaps some business logic stuff"?

perhaps you should tell this to eBay, which dumped microsoft and will
be running all of its transactions using J2EE. right now, it handles
more than 400 million transactions a day, but the thing is architected
to handle up to a billion dynamic page views a day by 2004.

or perhaps you should tell this to the thousands of large enterprises
that run IBM Websphere, BEA Weblogic, SunONE, Oracle 9i, etc.

BEA WEblogic's J2EE app server is used in 30,000 customers worldwide,
while IBM's J2EE app server Websphere provides service to:

65% of the Fortune 500 companies
80% of the top US healthcare companies
75% of commercial banks worldwide
90% of the top commercial banks in the US
67% of the world's largest banks use IBM messaging servers
15 of the top Wall Street brokerage firms
7 of the 8 largest US telecommunications companies

yeah, "simple apps" indeed.

[xx]

asj wrote:

mlw <ml*@nospam.no> wrote in message news:<5g11b.216055$uu5.40024@sccrnsc04>...

43 wrote:
Hey, Java and .NET, like VisualBasic and COBOL, are high level languages
intended for inexperienced developers or business applications. As such,
they are fine for things like LimeWire, Instant Message clients, and
perhaps some business logic stuff.

"perhaps some business logic stuff"?

perhaps you should tell this to eBay, which dumped microsoft and will
be running all of its transactions using J2EE. right now, it handles
more than 400 million transactions a day, but the thing is architected
to handle up to a billion dynamic page views a day by 2004.

or perhaps you should tell this to the thousands of large enterprises
that run IBM Websphere, BEA Weblogic, SunONE, Oracle 9i, etc.

BEA WEblogic's J2EE app server is used in 30,000 customers worldwide,
while IBM's J2EE app server Websphere provides service to:

65% of the Fortune 500 companies
80% of the top US healthcare companies
75% of commercial banks worldwide
90% of the top commercial banks in the US
67% of the world's largest banks use IBM messaging servers
15 of the top Wall Street brokerage firms
7 of the 8 largest US telecommunications companies

yeah, "simple apps" indeed.

Actually, web applications and business logic applications are *very*
simple indeed. It is the infrastructure behind the J2EE environments is
very heavy duty and complex. Clustered SQL databases (obviously not
written in Java), caching systems, and so on.

The front end, load balanced, redundent, J2EE environments doing the
business logic are very simple. Have you never seen and enterprise system?

Now, what was your point?

[xx]

asj wrote:

mlw <ml*@nospam.no> wrote in message news:<5g11b.216055$uu5.40024@sccrnsc04>...

43 wrote:
Hey, Java and .NET, like VisualBasic and COBOL, are high level languages
intended for inexperienced developers or business applications. As such,
they are fine for things like LimeWire, Instant Message clients, and
perhaps some business logic stuff.

"perhaps some business logic stuff"?

perhaps you should tell this to eBay, which dumped microsoft and will
be running all of its transactions using J2EE. right now, it handles
more than 400 million transactions a day, but the thing is architected
to handle up to a billion dynamic page views a day by 2004.

The Mohawk Software "Msession" manager for PHP can handle a regularly
distributed "400 million" page/session operations in a day on one 800MHZ
dual PIII Linux box.

This comes to a little more that 4600 pages views a second. The msession
daemon can handle over 5000 in the afore mention configuration. And, no,
it is not written in Java. It is written in C++.

[Jeff Relf]

Hi 101011 , You say :
" Walkie talkies are a P2P network where n = 2 .
If your walkie talkie could allow other walkie talkies
to ' island hop ' their signal
on part of your system , then n would be higher . "

I can't think of a P2P network that
doesn't fall under some hierarchy .

Even Napster was under
the hierarchy of the internet and the government .

Are you dreaming again John ?

[xx]

Jeff Relf wrote:

Hi 43 , you say :
" How come Linux and all Linux applications
are written in c/c++ ? "
Scripts are for quick and dirty jobs . Simple Jobs .

C++ is for Large serious jobs . Complex Jobs .

And Java is for?

[43]

Jeff Relf:

Hi 101011 , You say :
" Walkie talkies are a P2P network where n = 2 .
If your walkie talkie could allow other walkie talkies
to ' island hop ' their signal
on part of your system , then n would be higher . "

I can't think of a P2P network that
doesn't fall under some hierarchy .

Even Napster was under
the hierarchy of the internet and the government .

you are somewhat correct:

the Internet is a hierarchy all the way up to the
class A DNS servers. however, no one said the
Internet is the be all and end all. with enough
pc's closely linked, the grid can be networks of
wireless workstations, sans an external 'network'

in that case, a p2p running on such a collection
of workstations that make up the network.

the current 'internet' is like the 'aether' of
19th century physics. once people realize
that it's not really necessary, it will be
abandoned.

[Heinz Getzler]

Gregory Gadow <te******@serv.net> wrote in message news:<3F**************@serv.net>...

"Michael A. Covington" wrote:

"Roedy Green" <ro***@mindprod.com> wrote in message
news:hm********************************@4ax.com...

On Thu, 21 Aug 2003 05:22:49 GMT, 43 <ja*****@earthlink.net> wrote or
quoted :

>how come m$Office isn't written in .net?
>
>how come Open Office isn't written in j2ee?

Because these are both server side technologies. Editing a document
is a solitary activity. It works best with instantaneous response to
each keystroke.

I think you are confusing .NET with ASP.NET. .NET is an object-oriented
API that has no particular association with servers. It's a great way to
write applications.

The reason MS Office isn't written in .NET has something to do with MS
Office being invented about 10 years before .NET was invented! Same reason
the battle reports of Julius Caesar aren't written in English...

The new version of Office, recently unveiled by MS, is apparently written in
.NET.

okay so how would a office .NET version be functionally different??
Would a person be able to compose word documents in a browser??? Or
would Word become a distributed Word processing application??

[Chad Myers]

"Jeff Relf" <__**********@NCPlus.NET> wrote in message
news:f3**************@Jeff.Relf...

Hi 43 , you say :
" How come Linux and all Linux applications
are written in c/c++ ? "
Scripts are for quick and dirty jobs . Simple Jobs .

C++ is for Large serious jobs . Complex Jobs .

Is there a lot of C++ in Linux? It seems most
apps I've downloaded/compiled/whatever are either
Perl or C.

-c

[Chad Myers]

"Roedy Green" <ro***@mindprod.com> wrote in message
news:ag********************************@4ax.com...

On Fri, 22 Aug 2003 00:05:47 -0500, "Chad Myers"
<cm****@N0.SP.AM.austin.rr.com> wrote or quoted :

Uh oh, anonymous methods are coming to C#. I personally
disagree with the reasons why seeing their abuse in Java,
but whatever.

there are no anonymous methods in Java. There are anonymous inner
classes however. Are you sure you feel qualified to judge something
you so obviously have never used?

I know what I meant :) And yes, I've used them.

I haven't done a lot of Swing development, but I've had
to do a little maintenance on already-existing apps and
I'm refering to the (pardon me, it's been awhile)
code like:

someWidget.onSomeEvent = new{ .... }

That's essentially what'll happen in .NET now. I know
they're anon. inner classes in .NET, but in Swing,
those A.I.C.'s are used in the same way that anon.
methods will be used in C# Windows.Forms programming.

I have seen some huge event handlers in-line in
Java Swing that defeats the point of declaring
the event handler in-line.

The point in C# for anon. methods is to declare
small event handlers in-line, but like most of
the Swing projects I've seen, I'm afraid that
people won't write "small" anon. methods, they'll
write large, complex, and ugly ones, thus
negating the positive usage of these types of
devices.

-c

[Chad Myers]

"Heinz Getzler" <ge*******@yahoo.com> wrote in message
news:a5**************************@posting.google.c om...

Gregory Gadow <te******@serv.net> wrote in message news:<3F**************@serv.net>...

The new version of Office, recently unveiled by MS, is apparently written in .NET.

okay so how would a office .NET version be functionally different??
Would a person be able to compose word documents in a browser??? Or
would Word become a distributed Word processing application??

NOTE: I don't think you understand what .NET is, exactly, please
read up on the .NET Framework and managed code. .NET not just
web services.

It allows MS to be more agile with the product and quickly
add new features or fix bugs.

It removes the possibility for a large number of types of
bugs that occur in unmanaged code, thus making it a more
reliable product

It will integrate easier with other products, including
3rd party ISV products

It can more easily use web services and XML

The possibility of porting Office to other platforms
is greater and easier

etc, etc, etc :)

-c