Re: [Q21.5] Incremental Garbage Collector (revised)

Marcus> What about this patch: [...] Marcus> Note that the XE_DIE case is commented out: Currently I'd Marcus> prefer to test if the platform supports POSIX if it is not Marcus> listed explicitly. I understand your desire for testing, but we really don't want people using a configuration where we expect one thing and configure finds something else. Since this is known to cause crashes, and I assume you know why so more experiments are not terribly useful, we should only actually build the configuration when everything checks out, or the user told us what he wants and we didn't have to guess anything. I'm pretty sure we can arrange that configure will not die until it finishes its detection. Then we can have it die with a request to mail XEmacs Beta with the information about the platform. Alternatively we could have it simply force the vdb off (as if the user had specified --with-vdb=fake), but I think people are more likely to notice and write the bug report if configure fails than if it succeeds with a warning. Which do you prefer? SJT> However, I would like to ask about the extreme repetitiveness SJT> of the patch; Marcus> Currently I cannot come up with anything nicer or less Marcus> repetitive. Someone else? I've merged the Xft code, and in the process I had to study what you're actually doing there. In fact most of them have no #else, so it's new code, and probably not redundant at all. This can wait, I think. (And it's the kind of thing that Ben can figure out very quickly.) BTW, as it worked out the only conflicts were in configure.ac in the argument section, and in lrecord.h. Both were trivial to resolve. -- School of Systems and Information Engineering 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.