[Mono-list] Microsoft C# and CLI for freeBSD && Windows/Runtime metadata stuff

James Perry jeperry@uwaterloo.ca
Thu, 9 Aug 2001 03:15:24 +0000 (GMT)


On 8 Aug 2001, Reginald M Poyau wrote:

>
> Microsoft and Corel are going develop shared source implementation of
> ECMA standards C# and CLI for FreeBSD and windows.  This is very very
> sneaky move if you ask me ( there are feigns within feigns here ).
>

They announced this a while ago. What's sneakiest about this is the
copyright issue involved, though. Look at the shared source implementation
of .NET and you're 'tainted'. If you've looked at their source code and
then work on a project which develops a compatible system but under
entirely different licensing (i.e., GPL) suddenly the threat of lawsuit
becomes real and the other project is faced with the prospect of either
scrapping large sections of its code, going to the expense of
demonstrating that the code is in no way based on the shared source
implementation and the possibility of a large legal bill that could
result.

Certainly an effective way of dealing with the competition should they
become too much of a threat - and nicely explains why Miguel says not to
look at any shared source implementations or disassemblies on the Mono
website.

(I tried to explain the dangers of this to the ReactOS project w.r.t. the
shared source release of WinCE 3.0, but it fell on deaf ears. I
think reactos.com even links to the shared source download page somewhere
now. For those who don't know, ReactOS is aimed at producing a GPL'd
implementation of WinNT - not pretty graphics yet but the system executive
seems to be coming along very well.)

-
On a completely different note... Miguel, does the runtime currently
include metadata validation? I couldn't find anything in libmetadata
or any other part of the runtime that looked like it could be it.

Similarly, does the runtime include in the metadata table info information
about what the column's values are (i.e., 1,2,4 byte constants or indexes
in heaps?). It's the kind of thing that would make testing for row
uniqueness not quite the PITA it could be, but again, I couldn't find it
anywhere if it's there.

 -james