#Symantec C++ 6.0 Bugs
2 messages in this thread
Andy,
You might want to post your code in the Symantec Development Tools Forum (GO
SYMDEV). I don't know if the guys over there still check this forum or not
(especially right now, since one of the sysops is on his honeymoon).
1. Indeed, constants ending in F should be treated as floats. Sounds like a
real bug to me.
2. I believe this is mentioned in the THINK C manual. Indeed, segmentation
directives in a library are not honored. They'd need to rewrite the linker to
deal with this. This has been true of the THINK environment from day 1.
3. I haven't tried a precompiled header that uses a precompiled header.
Sorry.
4. Again, check the THINK C manuals. (I refer to the THINK C manuals because
there's a lot of common ground involved here; most of what you're running into
is really in the THINK Project Manager.) I know FAR DATA doesn't mean you can
still have more than 32K of global data in a single file; I believe all it does
is let you have 32K *per file* rather than per project. Again, I'm in
unfamiliar waters here, as I've never used far code or far data.
5. See above. This does sound like a bug, although it could just be a natural
result of the code overflow.
Joe
Joe,
Thanks for responding. I did post in Symantec's forum, and called, and posted,
and called. They've got nothing to say to me?!?! It looks like they're not
going to fix SC++6.0 in the near future. It really looks like they don't have
the resources to maintain their compiler's anymore. The tech guy I talked to
didn't have a clue when I started spouting C++. This really supprises me since
the article about Apple turning over all development tools to them?!?!?
Oh well, looks like MPW is here to stay for a while. I would return the
SC++6.0 except that I got it with a product swap.
Thanks again,
Andy.