[Top][All Lists]

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

Re: [DotGNU]Starting of the swigging of pnet

From: James Michael DuPont
Subject: Re: [DotGNU]Starting of the swigging of pnet
Date: Sun, 16 Mar 2003 06:55:52 -0800 (PST)

--- Rhys Weatherley <address@hidden> wrote:
> On Sunday 16 March 2003 08:19 pm, James Michael DuPont wrote:
> > The reason why I started on this was to help implement
> reflection.emit.
> > When the ILAsm code is wrapped in internal calls, and made
> available
> > via static linking to the C#, it can be used to implement the
> > Reflection.Emit code.
> Which is the wrong approach.  Reflection.Emit should be implemented
> using 
> internalcalls direct to the underlying engine, in the same style as
> the 
> existing System.Reflection interface.  All of the necessary
> infrastructure is 
> already in place - it just needs to be hooked through.

That is the next step to emulate that. The code that is used by the
current System.Reflection is the same as the Ilasm uses, right? Even if
not, when the swig gets going right, then it will possible to get into
all the internalcalls as well.

> Swig will *not* help you implement Reflection.Emit.  It will instead
> introduce 
> huge amounts of unnecessary complexity and bloat.  Complexity that I
> will 
> have no choice but to reject if it turns up in the patch queue.

The complexity is in the current code generation which needs lots of
work. By modifying swig to use/emulate the idioms of pnet, we will make
the code generation more efficient. That will make swigsharp a
generally useful tool.

> *sigh*  Solve the actual problem Michael, not what you want the
> problem to be.

We will get there. One step at a time. 
This exercize is getting swig running, and I am learning pnet,
everything takes its time.


James Michael DuPont

Do you Yahoo!?
Yahoo! Web Hosting - establish your business online

reply via email to

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