Board index » delphi » Re: My Suggestion : A New C# compiler for native code

Re: My Suggestion : A New C# compiler for native code


2007-01-09 01:29:45 AM
delphi103
"Renato" <XXXX@XXXXX.COM>writes:
Quote
about a new C# compiler for native Win32/64.
What you guys think about this ?

Regards.

Renato
Technically, it is possible, but been practically,
It won't be much of a bussiness for CodeGear/Borland.
Maybe a good idea for someone's personal open source project...
There are several "native Java" projects, doing trying to
do the same stuff, but with Java.
The C# programming language by itself, adds stuff that wasn't
available to C++. A few years ago, Borland proposed to expand
the C++ language to support properties, events, and other
stuff that is currently SUPPORTED BY JAVA AND C#.
Search in Goggle's newsgroup search:
borland.public.cppbuilder.non-technical
"Borland proposes PME additions to the C++ standard"
Many C/C++ rejected those additions, altough necessary.
There is another important feature about C#/Java versus
C++/Delphi: Module support.
In C#/Java all units/modules are "dynamic libraries",
and are called on "the fly". Developers doesn't need to
"prepare" functions from a "dynamic library", as it does
in Delphi or C++. The NET/Java compiler adds the implicit
necessary code.
A native C#/Java compiler with static libraries may be
an interesting idea, but doesn't provide enough benefit/cost
for a commercial compiler, not at the moment, I think.
Just my 2 cents...
maramirezc
 
 

Re: My Suggestion : A New C# compiler for native code

"Marco van de Voort" <XXXX@XXXXX.COM>writes
Quote
On 2007-01-08, Renato <XXXX@XXXXX.COM>writes:
>
Note that the fact that he used webbroker was not known when the
alternative
was suggested.
Then, maybe the question should have been asked before suggesting, yet
again, Lazarus as a possible solution?
Alan.
 

Re: My Suggestion : A New C# compiler for native code

Nils Boedeker writes:
Quote
you can run 32bit applications in Win64 IIS...but this is only an
workaround and not a solution:
There's a big caveat in that text, though:
"On 64-bit Windows, the World Wide Web Publishing service does
not support running 32-bit and 64-bit worker processes concurrently
on the same server."
--
Arthur Hoornweg
(In order to reply per e-mail, please just remove the ".net"
from my e-mail address. Leave the rest of the address intact
including the "antispam" part. I had to take this measure to
counteract unsollicited mail.)
 

Re: My Suggestion : A New C# compiler for native code

Alan Garny writes:
Quote
>>I like to see a native 64bit complier + VCL for Delphi... we have a big
>>ISAPI WCMS Application and want to port them to Win64 bit.
>Lazarus Win64 bit is here!
>michael-ep3.physik.uni-halle.de/Lazarus/

Don't you think that people know about Lazarus by now?! Rather than
suggesting Lazarus here and there, you should rather ask yourself why some
people (including myself) who want to port their application to Win64 (for
instance) have decided against Lazarus.

Actually, I was wondering when Lazarus would have a Win64 compiler.
So this is good news for me.
I use Delphi but I like knowing there's an alternative to compile my sources.
This is especially important for me now since there isn't a Delphi Win64 compiler yet.
Good job Lazarus team.
Michael
 

Re: My Suggestion : A New C# compiler for native code

Rod writes:
Quote
Nils Boedeker writes:
>I like to see a native 64bit complier + VCL for Delphi... we have a
>big ISAPI WCMS Application and want to port them to Win64 bit.

Lazarus Win64 bit is here!
michael-ep3.physik.uni-halle.de/Lazarus/
This is great news. Thanks for sharing this.
Michael
 

Re: My Suggestion : A New C# compiler for native code

Quote
Lazarus Win64 bit is here!
michael-ep3.physik.uni-halle.de/Lazarus/
Is there a working version somewhere?
I've tried it several times (since it has been first announced) and
never got a version that would work, recurring issues being:
- garbled icons in the IDE (have to use the hints to find which
component in the palette you're selecting)
- compiled output that isn't recognized as an EXE (and thus won't run at
all)
I know these are auto snapshots or something that aren't guaranteed to
work, but having at least one version that a human has minimally tested
and that offers minimal functionality would be nice :)
Eric
 

Re: My Suggestion : A New C# compiler for native code

On 2007-01-08, Alan Garny <XXXX@XXXXX.COM>writes:
Quote
>Note that the fact that he used webbroker was not known when the
>alternative
>was suggested.

Then, maybe the question should have been asked before suggesting, yet
again, Lazarus as a possible solution?
How do you imagine this? A list of 200 questions to fill out? Webbroker was
the question to ask in retrospect, but nobody could know this up front.
 

Re: My Suggestion : A New C# compiler for native code

Hi,
Quote
There's a big caveat in that text, though:

"On 64-bit Windows, the World Wide Web Publishing service does
not support running 32-bit and 64-bit worker processes concurrently
on the same server."
Yep
Nils
 

Re: My Suggestion : A New C# compiler for native code

"Lars Fosdal" <Lars(q)Fosdal.com>writes:
Quote
Will obfuscated code run (somewhat) slower than un-obfuscated,
Probably.
Quote
Wouldn't a protection against easily reversing the CIL code, be to
compile the CIL to native code at installation instead of JIT?
The source CIL assemblies have to be installed and available even after
NGENing.