[AC21.5] C++ fixes
Stephen J. Turnbull
stephen at xemacs.org
Fri Apr 9 07:22:20 EDT 2004
>>>>> "Jerry" == Jerry James <james at xemacs.org> writes:
Jerry> "Stephen J. Turnbull" <stephen at xemacs.org> wrote:
>> C++ build is busted. This needs to be fixed.
Jerry> Mats is right.
In theory somebody could sign up for the SourceForge build farm.
Jerry> There are too many combinations for me to
Jerry> keep up with them all.
Well, of course. That's why I ran the C++ build when I was reminded
of it by the configure.in fiddling I did, especially in view of
Mr. Royer's experience with union type. I actually didn't expect it
to break :-(.
I don't mean to pick on you (tease yes, I did, but imply your
practices are substandard, no), and I'm hardly more than an armchair
standards lawyer at best. I'm just repeating what people who are much
better programmers than me (Ben) and much better maintainers than me
(Martin) have insisted on over the years. If Ben and Martin agree,
then I'd bet it's true, you don't have to tell me the third time. ;-)
FWIW, I would say that using GCC C w/ dis-union (I'm pretty sure there
have been syntax bugs that didn't appear with strict type-checking but
showed up in the C/disunion build), C++ w/ union pretty much covers
the bases for this kind of breakage. If you want to do other
compilers, that's very fine, but IMHO this pair is a pretty good
baseline for individuals, especially in the stage you're at.
The bignums v. not, modules v. not, assorted platforms stuff is
important on a beta-to-beta basis, but if at all possible the code
should be kept as clean as possible vis-a-vis the C++/union build on a
day-to-day basis (again IMHO).
Institute of Policy and Planning Sciences http://turnbull.sk.tsukuba.ac.jp
University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Ask not how you can "do" free software business;
ask what your business can "do for" free software.
More information about the XEmacs-Beta