According to Geoff Hutchison:
> This is a barely tested patch to fix that--I'd recommend using this
> instead of the patch on StepWise that most people have been using.
Instead of, or in addition to? The StepWise patch had other fixes,
like getting rid of things like "::read()" if I recall correctly, which
were fixed in the 3.2 CVS tree, but not in the 3.1 code. Maybe these
aren't needed with the latest Mac OS X release?
> I'll commit this to the CVS tree tomorrow so it will make it into
> 3.1.6. (At which point, I'll see if I can get a Mac OS X binary
> installer together.)
>
> The patch fixes the malloc.h error in DB2_db.cc, as well as the
> stream.h error in String.cc. The bulk of the patch is an update to
> the db/dist/config.guess that enables it to recognize a variety of
> flavors of Mac OS X out of the box.
I'm still not convinced that replacing the include of stream.h with
iostream.h is the right thing to do on all systems. I can't recall
for sure, but it seems this had come up a long time ago and there
was a reason for the distinction between the two. I don't know the
C++ library classes enough to say for sure what it should be, though.
I think part of the problem was that the classes were evolving, so we
can't count on all systems to use the same standard.
> The bug in the Configuration.cc code is pretty simple--the code
> currently attempts to "peel off" any whitespace after a config
> attribute, but never bothers to make sure the string has any length.
> After all, getting string[-1] is a Bad Idea(TM).
But because the loop peels off characters, shouldn't the test to make
sure len >= 0 be right in the loop? I.e.:
while (len >= 0 && (value[len] == ' ' || value[len] == '\t'))
value[len--] = '\0';
I guess we've already stripped off the leading white space, so we should
run into something before the start of the string, if len >= 0, but I'm
into defensive programming.
> I'll check to make sure the next 3.2 snapshot compiles cleanly on OS X as well.
Great, thanks. That'll clear up a few longstanding threads of discussion.
--
Gilles R. Detillieux E-mail: <[EMAIL PROTECTED]>
Spinal Cord Research Centre WWW: http://www.scrc.umanitoba.ca/~grdetil
Dept. Physiology, U. of Manitoba Phone: (204)789-3766
Winnipeg, MB R3E 3J7 (Canada) Fax: (204)789-3930
_______________________________________________
htdig-general mailing list <[EMAIL PROTECTED]>
To unsubscribe, send a message to <[EMAIL PROTECTED]> with a
subject of unsubscribe
FAQ: http://htdig.sourceforge.net/FAQ.html
