dotgnu-general
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [DotGNU]RFC : [Proposals]ILAutoStubber Proposal


From: James Michael DuPont
Subject: Re: [DotGNU]RFC : [Proposals]ILAutoStubber Proposal
Date: Wed, 5 Mar 2003 23:48:20 -0800 (PST)

--- Rhys Weatherley <address@hidden> wrote:
> On Thursday 06 March 2003 09:16 am, James Michael DuPont wrote:
> 
> > > A IL Processor that can create stubs out of the IL processed.
> 
> Seems useful, although it does set my "derived work" light blinking a
> little.  
> What you have specified is essentially a decompiler, which stops
> before 
> decompiling the method bodies.

Yes. 
Good Point, and I am glad you brought that up.

That is true. And it it a real Issue in general,
the creating of compatible interfaces might even be patent violation in
some cases. The webservices for example.

The Destructive ILStubDestruct can be seen that way,
it just guts the il, and the perl script I attached
 already does some of it. That is not the real value of the
AutoStubber.

I think the ILStubBridge Program is the best example 
of a valuable usage of the AutoStubber, Create a compatible interface,
Log all the calls, and somehow translate to our Compatible interface.
We can use that for example to create a bridge from WinForms to QT, for
example, at least on the level of the simple classes.

On the other side, it is a linker. 
This is basically creating a header file to link against.

And after all my research, it is not a problem to 
copy the header file of windows (or linux) and implement a brand new
implementation of the API.

The AutoStubConstruct only will stub out
the methods called! this is a form of an intellegent linker
that will help users free themselves from non free software.
It can be argued that it is a derived work of the caller, and
all the calls are written by the user.

> I usually prefer that people write the entire class themselves,
> including the 
> boring bits.  Then there is no question that we wrote it ourselves,
> in our 
> own words.  It may take a little longer, but it is easier to defend
> in court 
> should it ever come to that.

Yes, that is a very good point. This is exactly what
I suggested to Frank the newbie yesterday. About the securty model,
I even went as far as to say that he create an entire new API that is
inspired by multiple implementations (OpenSSl, among others)

> 
> In any case, we already have an auto-stubber of sorts that Gopal
> wrote: 
> "csdoc2stub".  This takes an XML description of the module as input. 
> What is 
> currently missing is a program to convert an IL binary into XML
> first, which 
> is one of the things on my list for the next few days.

OK, That would be good. 
This is similar to what we talked about, and I had put it on my task
list : 
** TaskIlRdfDumping
      DumpIL into RDF/XML for
      status : on hold 

I would like to work on that, because it would be a good opportunity
for me to learn more about IL and to combine my interests in XML with
DotGNU. 

If you permit, then I would start by creating a Fleshed out
specification of the problem, like this :
http://wiki.dotgnu.org/TaskFoafDotGnu


Please Advise,
mike

=====
James Michael DuPont
http://introspector.sourceforge.net/

__________________________________________________
Do you Yahoo!?
Yahoo! Tax Center - forms, calculators, tips, more
http://taxes.yahoo.com/


reply via email to

[Prev in Thread] Current Thread [Next in Thread]