[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[DotGNU]Does cscc need a pedantic mode ?
From: |
Gopal V |
Subject: |
[DotGNU]Does cscc need a pedantic mode ? |
Date: |
Tue, 26 Nov 2002 17:42:43 +0530 |
User-agent: |
Mutt/1.2.5i |
Hi All,
I've been pestered by marcusU every time I hit IRC with a
bug which does not affect cscc in any negative way ... IMHO it
even works the way I want to .. So here is the code
using i32=Int32;
Should this throw an error saying "'Int32' not found" ?..
ECMA spec says that it should throw an error ... I'd like to
comply .. but that would mean throwing out a some code I
wrote to support double aliasing (alias to an alias) , and
circularity checks ....
Maybe I went overboard with features ... But DotGNU always promised
to run what MS .NET does ... it never said "we will fail where MS .NET
fails" ..
Nov 26 13:35:28 <marcusU> Just cause cscc compiles the example in
the ECMA std that's supposed to illustrate an
error!
........
Nov 26 13:37:28 <t3rmin4t0r> well marcusU, I'm more concerned in
getting cscc to compile things .. rather than
other way around
Nov 26 13:37:51 <t3rmin4t0r> especially something that does not
affect bytecode output (like using)
Nov 26 13:38:01 <t3rmin4t0r> which is inherently difficult to debug
Nov 26 13:38:08 <marcusU> I guess, but the invalid programs that a
compiler rejects is just as important as the
valid program
a compiler accepts.
.........
Nov 26 13:39:44 <marcusU> Okay. But cscc isn't a C# compiler. It
accepts some other language that looks like C#
but isn't.
Ok ... having said all that .. do we need a 'pedantic' mode for complete
and total subservience to what MS CSC does ?
Gopal
PS: I wrote it so that it works like a common macro #define in C ...
--
The difference between insanity and genius is measured by success
- [DotGNU]Does cscc need a pedantic mode ?,
Gopal V <=