[ros-kernel] about headers

[KJK::Hyperion]

At 21.52 19/10/2003, Danny Smith wrote:
>>Also, our headers should not contain anything not found in the MS headers.
>As far as the mingw runtime headers are concerned, you will have a hard 
>time convincing me to remove all the ISO C99 extras.

isn't it possible to write universal headers, that could be used on all 
known compilers? (I'm lying, I know it is. See STLPort)

>You will also have a hard time convincing me that removal of all the 
>POSIX-isms is a good thing, since they really assist building of things 
>like binutils, gcc, libiconv, gettext, make, etc.

I'd object that a build environment doesn't exist primarily to build 
itself, but to build stuff people write. If you mean the fake POSIX headers 
like <unistd.h>, though, I can only say: bring more! I'd really like them 
in pre-packaged format and compatible with major Windows compilers: then I 
could just install them, instead of rewriting them by hand every time :-)

>Adherence to the MS "gold standard" has meant that mingw is forced to use 
>sjlj exceptions rather than DWARF2.

but you know MinGW isn't forced to use setjmp-longjmp exceptions: it's 
forced to use SEH-based exceptions :-) no, really, you can run but you 
can't hide: C++ on Windows means support for SEH. You may find synchronous 
signals with exception semantics a questionable design, but it's so handy 
in practice that everyone exploits it. MinGW failing to implement SEH-based 
exceptions isn't very dissimilar from Borland C++ insisting on the OMF 
binary format while the rest of the world speaks COFF: it means hassle, 
hassle and hassle