ELKS and .NET
- Hi everybody!
Before I start to call for discussion, let me say some words about
.NET. Personally I do not share the general excitement. I see a
strong possibility that it will suffer the same fate as other plans
"to revolutionize computing" out of Redmond have suffered, namely a
quiet death. Furthermore I don't trust MS one bit to get it right. As
somebody that does research into security, I have no trust at all into
I also do not see many people using main-stream languages to create
high quality components where knowing the interface is really enough
to use them. This takes a lot of discipline even in Eiffel. Mixed
language projects are also not more attractive because of .NET.
The ability to link modules from different languages is not
new (in fact I did it on an ATARI ST a decade or so ago), and
was never widely adopted. The reason may be an other problem than
the binary format.
Don't get me wrong, ignoring .NET might not be an option for
people doing work on MS platforms. But I do not believe it
makes our work here obsolete or redundant. Being able to switch
from one compiler for a language to another one with minimal
problems is a major advantage for a language.
Furthermore we are not only specifying basic classes here. We are
also setting an example on how it should be done. An example that
everybody doing more complex things on top of ELKS can benefit
from. And on a more personal level, I gained a lot of insights here in
the past about the nature of the things the are actually modelled by
O.K., enough of that. I will now open the discussion on
COMPARABLE in a separate message.
Please feel free to discuss changes in the overall agenda as long as
we have not gotten to that part of the agenda yet. I feel that we are
competent enough to update the agenda from time to time without
loosing our overall direction.
Arno Wagner, Communication Systems Group, ETH Zuerich, wagner@...
GnuPG: ID: 1E25338F FP: 0C30 5782 9D93 F785 E79C 0296 797F 6B50 1E25 338F
For every complex problem there is an answer that is clear, simple,
and wrong. -- H L Mencken
> The ability to link modules from different languages is notIt's largely true, but there has been some instances of mixed
> new (...), and was never widely adopted.
languages in widespread use: on Unix, scripting languages associated
with C libraries (python etc), and with Windows UIs with VB as glue
for C/C++ UI components.
> I do not believe it makes our work here obsolete or redundant.Agreed.
> Furthermore we are not only specifying basic classes here. We areWe're constrained by the existing though, although there's
> also setting an example on how it should be done.
still scope for doing good things within that constraint.