Board index » kylix » A suggestion regarding lack of support for Kylix

A suggestion regarding lack of support for Kylix


2003-09-04 05:21:18 PM
kylix1
Considering these thoughts:
- .NET IL (i.e. bytecode) might get an (ISO) standard (maybe even
independent of Microsoft)
- IL code can be _compiled_ at the target system. The native code
result is said to be not slower (sometimes even faster) than native code
compiled from Delphi language or C++. The compilation from IL to native
code is said to not significantly increase the load time of an
application.
- The IL compiler always needs to stay active on the target machine,
but RAM is chap
- Upcoming 64 Bit processors (Itanium or AMD64) would need new
compilers if Delphi/Kylix want to stay excellent tools (much work for
Borland)
- Everybody in Linux Land complains about Kylix not creating code for
other CPUs but IA32, which would need new compilers (much work for
Borland), too
- IL interpreters / compilers soon will be available on all OSes and
CPU architectures (Windows anyway, Linux: MONO, support for 64 Bit
processors will come automatically, PALMs will get it, too...)
- Deploying IL programs will be much easier than native (only one
version for all targets, very small files)
So maybe Borland decided to drop anything but IL generation (Delphi for
.NET). Here it's not very important if or if not the IDE runs on
Windows, Linux or both.
-Michael
 
 

Re:A suggestion regarding lack of support for Kylix

"Michael Schnell" < XXXX@XXXXX.COM >wrote in message
Quote
Considering these thoughts:

Good thoughts. Make it so.
 

Re:A suggestion regarding lack of support for Kylix

"Michael Schnell" < XXXX@XXXXX.COM >wrote in message
Quote
- .NET IL (i.e. bytecode) might get an (ISO) standard (maybe even
independent of Microsoft)
The whole idea of a standard is it's independence. The CLI is already a
ECMA standard, the ISO equivalent is just a formality.
As far as Kylix is concerned, the real problem is not the lack of support
but the fact that so much support is required --- mainly due to external,
systemic factors beyond Borland's control. In a nutshell, Kylix is merely a
visible manifestation of problems inherent to Linux. The best current hope
for a solution is Mono IMO.
Kylix was a risk and not one with a particularly high probability of
success. A more careful and in-depth analysis upfront should have revealed
some of the reasons why.
 

{smallsort}

Re:A suggestion regarding lack of support for Kylix

Michael Schnell < XXXX@XXXXX.COM >wrote in message news:< XXXX@XXXXX.COM >...
Quote
Considering these thoughts:

- .NET IL (i.e. bytecode) might get an (ISO) standard (maybe even
independent of Microsoft)
I doubt you can make MS's IL a standard independent of them.
Isn't it their intellecual property? I suppose you could make an
open source equivalent though. I can't imagine MS invented the idea.
Otherwise the "One IL language to rule them all" idea is kind of
cool.
It's propably what Java had in mind. Hopefully you'd have the option
to byte-compile into an .exe as well.
 

Re:A suggestion regarding lack of support for Kylix

Quote
As far as Kylix is concerned, the real problem is not the lack of support
but the fact that so much support is required --- mainly due to external,
systemic factors beyond Borland's control. In a nutshell, Kylix is merely
a
visible manifestation of problems inherent to Linux. The best current
hope
for a solution is Mono IMO.

Maybe you are too much too indulgent with Borland.
Kylix is full of problems that are of the sole responsability of Borland,
then cames the problems that appeared due the lack of Linux expertise inside
Borland and because a Windows centric point of view, then comes the lack of
support and continued development to make the program succesful and the
comes the inherent problems of Linux.
There are many successful development systems that do not suffer the
problems that Kylix has, didn't you think about that? Even if they do not
have all the bells and whistles that Kylix *pretended* to have, they work
and are pretty solid.
It seems you suffered the same kind of problems that Borland had, didn't
you? Perhaps you should try to look for the origin of some of those problems
inside you and try to understand the platform a little more, instead of
simply stating grosse assertions...
Regards,
K.M.
 

Re:A suggestion regarding lack of support for Kylix

Quote

The whole idea of a standard is it's independence. The CLI is already a
ECMA standard, the ISO equivalent is just a formality.

Does that mean that we are safe assuming that "Delphi for .Net" created
projects will run on MONO just fine ?
-Michael
 

Re:A suggestion regarding lack of support for Kylix

"Michael Schnell" < XXXX@XXXXX.COM >wrote in message
Quote
Does that mean that we are safe assuming that "Delphi for .Net" created
projects will run on MONO just fine ?
Personally, I'd feel safe assuming that "Delphi for .NET" will conform to
the standard. As for the rest, it's up to the Mono project.
 

Re:A suggestion regarding lack of support for Kylix

<K.M.>wrote in message news: XXXX@XXXXX.COM ...
Quote
There are many successful development systems that do not suffer the
problems that Kylix has, didn't you think about that?
Yes, I did think about that.
The Open Source answer to long standing *nix development problems is to
release the source code. Most "successful" Linux desktop apps comply; Kylix
does not and therein lies the root of the problem.
Promoting and supporting OS development is all part of the Open Source plan.
In the process, strictly proprietary apps (like Kylix) are subtlely but
effectively discouraged ---- by simple indifference if nothing else.
Indifference at the operating system level represents a formidable barrier
to any proprietary desktop software; as Kylix demonstrates.
Mono is the best hope to transform Linux into an equally viable platform for
both proprietary and Open Source software. And no surprise (at least to
me<g>) that many in the OS community don't like it very much.
Did you think about that?
 

Re:A suggestion regarding lack of support for Kylix

Quote
Does that mean that we are safe assuming that "Delphi for .Net" created
projects will run on MONO just fine ?
I doubt if this could work.
1) There's a lot of work required by Borland to port the Borland Data
Provider to Linux. I don't think that will happen anytime soon.
There's no SQL Server or MSDE for linux, so they'd have to target
MySQL or something like that. I know MySQL is pretty good and it does
have an ADO.Net driver for the Windows platform, but I don't think
that will work on linux.
2) Mono itself isn't ready for primetime.
Maybe a minimum-function Delphi for .Net program that doesn't use the
BDP could work on Mono, but we're nowhere close to getting full
functionality.
Eric
 

Re:A suggestion regarding lack of support for Kylix

"JQP" < XXXX@XXXXX.COM >wrote:
Quote
As far as Kylix is concerned, the real problem is
Troll !
 

Re:A suggestion regarding lack of support for Kylix

<K.M.>wrote:
Quote
It seems you suffered the same kind of problems that Borland>had, didn't you? Perhaps you should try to look for the origin>of some of those problems inside you and try to understand the>platform a little more, instead of simply stating grosse>assertions...
Reasoning with a Troll? Don't waste your time.
 

Re:A suggestion regarding lack of support for Kylix

Mauro wrote:
Quote
"JQP" < XXXX@XXXXX.COM >wrote:
>As far as Kylix is concerned, the real problem is

Troll !
Idiot!
 

Re:A suggestion regarding lack of support for Kylix

K.M. wrote:
Quote
Kylix is full of problems that are of the sole responsability of
Borland, then cames the problems that appeared due the lack of Linux
expertise inside Borland and because a Windows centric point of view..
Borland produces tools for a number of platforms natively, including
Windows, Linux, Symbian and of course, .NET.
Quote
It seems you suffered the same kind of problems that Borland had,
didn't you? Perhaps you should try to look for the origin of some of
those problems inside you and try to understand the platform a little
more, instead of simply stating grosse assertions...
Perhaps you should find out more about what expertise Borland has before
making gross assertions?
--
Dave Nottage (TeamB)
 

Re:A suggestion regarding lack of support for Kylix

Michael Schnell < XXXX@XXXXX.COM >wrote in message news:< XXXX@XXXXX.COM >...
Quote
So maybe Borland decided to drop anything but IL generation (Delphi for
.NET). Here it's not very important if or if not the IDE runs on
Windows, Linux or both.
I'm socially dense, but I think I've noticed occasional hints of
hostility from some Linux people towards microsoft (even though they
try hard to hide it, it leaks out sometimes), so part of this
strategy might be a concurrent alternative path - an IL based on the
java vm.
but didn't borland try to do that a long time ago and SUN wouldn't
cooperate?
 

Re:A suggestion regarding lack of support for Kylix

Quote
I'm socially dense, but I think I've noticed occasional hints of
hostility from some Linux people towards microsoft (even though they try
hard to hide it, it leaks out sometimes), so part of this strategy might
be a concurrent alternative path - an IL based on the java vm.

but didn't borland try to do that a long time ago and SUN wouldn't
cooperate?
Hey, wot are you saying? That us Linux folks are lacking in social
graces? That we are Anti-Winders? Hey, man, I use nothing but Linux but
I fully support Windoze. In fact, I think that the death penalty should
be totally repealed and bad dudes should be sentenced to use nothing but
Redmond products for the rest of their life. Of course, it would be
expensive with the requirement for all those suicide watches on the entire
cell block:)
(Dont take the above personally. Some of my best friends use Windows, but
I still allow them into my house.)
GW