Joseph Galbraith:

> I did try Platform.h... it isn't included soon enough.
> Platform would need to be included before any of the
> deprecated functions are included.

   OK.

> The only other thing I can think of is to introduce
> a private header that is included first by every
> cxx file.

   Just leave it as a -D for now.

> Eventually I'd hope there could be a compatibility wrapper
> that implemented these in the case were the system c runtime
> didn't provide them.

   That would be unfortunate. It will be much better if these
functions are standardised, which is I think Microsoft's intention,
and supplied by all compilers.

> All right, I'll fix those.  I'll also fix the NMHDR thing since
> you figured it out :-)
>
> In fact, here it is:
> ...
> + Fix NotifyHeader::idFrom to be a pointer type as defined
>    in the PSDK.  In order not to break already existing
>    64-bit clients, only do so for _WIN32.
> ...
>         void *hwndFrom;
> +#ifdef _WIN32
> +       uptr_t idFrom;
> +#else
>         unsigned int idFrom;
> +#endif
>         unsigned int code;
>  };

   I'm reasonably happy with always defining "uptr_t idFrom" despite
potential breakage. 64-bit Linux applications may fail, but that
should be rare and they will be warned by both this mailing list and
in the change notes.

   Neil

_______________________________________________
Scintilla-interest mailing list
[email protected]
http://mailman.lyra.org/mailman/listinfo/scintilla-interest

Reply via email to