Re: Replacement for MS STL?
"Garry W" <ListMail3@extremelyserious.org> wrote in message
news:gbi0i2h5qksje13ijl5up3cu0737cl7f0e@4ax.com...
Ulrich Eckhardt <eckhardt@satorlaser.com> wrote:
I think it's a lack of understanding on your side. The point is that I
must
be able to do this:
#define i 7*6
#define x 23,
#include <vector>
and still have a working vector afterwards. That is why that code, except
for the interfaces required by the standard, uses names reserved to the
compiler and standardlibrary implementation like __foo and _Bar.
Well, the MS STL -almost- always remembers to put the _'s in there. :)
Nope, it *always* does. The only names you'll find that aren't
"obfuscated" (defined in the implementer's name space) are the
ones mandated by the C++ Standard.
Here's the beginning of the MS STL "de-obfuscator" that I had, by
necessity,
started working on. Old internal names are on the left, replacement ones
are
on the right.
.....
So far, you've reported:
1) that you don't like the documented way that debugging gets
turned on and off, so you experimented at length and got in
some trouble
2) that you tried to do something with strings that's documented
as not supported, and you got in some trouble
3) that you *think* the names internal to the library are
inconsistent, and you *feel* they are in bad style
But you began this thread by saying:
: And I've given up on the Microsoft implementation. When I tried to use it
I
: immediately fell over several bugs in it that took a lot of time to find.
And
: working with it has been really difficult because the code is kind of
: illegible. At least for the uninitiated. (Apologies if there are authors
: here!)
So far you are zero for three, and seem to have trouble
reading documentation. But if you have other bugs to report,
I'm still very interested in hearing about them.
P.J. Plauger
Dinkumware, Ltd.
http://www.dinkumware.com