[antlr-interest] More, Status of C++ backend?

[Maurizio de Pascale]

You're right Jim, that's why I've said "ideally" ;D

I've often listened to people complaining about poor compilers and 
linkers (especially for embedded devices), simply I used to believe that 
at least the meaning of extern c was standard across compilers and that 
non templated C++ code should compile correctly even on toy compilers.

Maybe some ANTLR user has a working C++ to C compiler ;)

cheers,
Maurizio
mdepascale at dii.unisi.it

Jim Idle wrote:
>   
>> -----Original Message-----
>> From: Gavin Lambert [mailto:antlr at mirality.co.nz]
>> Sent: Thursday, January 03, 2008 11:53 AM
>> To: Maurizio de Pascale
>> Cc: antlr-interest at antlr.org
>> Subject: Re: [antlr-interest] More, Status of C++ backend?
>>
>> At 23:59 3/01/2008, Maurizio de Pascale wrote:
>>  >My suggestion is that the right road (ideally of course, I know
>> the
>>  >C runtime is already in place) would be to have the runtime
>> written
>>  >in C++ (as close as possible to the java counterpart) and
>> provide
>>  >(also) a C API for C users.
>>
>> No, I disagree with that.  It's important to keep a pure C
>> runtime, for use with embedded hardware.  Most of the embedded C++
>> compilers I've seen are fairly flaky.  (And that includes gcc.)
>>
>>     
>
> Yes. Maurizio - I guess you have never had to maintain C++ libraries on 
> 12 different operating systems all with different compilers with 
> different ideas about C++ linkage and interaction with C ;-). I don't 
> wish to put that burden on everyone who wants a C interface as I have 
> had to do this with ANTLR 2 and it was a complete nightmare, having 
> nothing to do with the quality of the code, but the C++ compilers and 
> the linkers. 
>
> It is going to be bad enough looking after the C++ library for C++ 
> programmers who are linking with C++ object :-)
>
> Jim
>
>
>