sb> Karl M Hegbloom <karlheg(a)inetarena.com&gt; writes in sb> xemacs-beta(a)xemacs.org: Hrvoje> karlheg(a)inetarena.com (Karl M. Hegbloom) writes: Hrvoje> This is unfortunately not true. More often than not, Hrvoje> these people don't *care* to support anything other than Hrvoje> Linux. sb> I don't know, perhaps, but ... It's not a case of absolutes. Hrvoje unfortunately omits the modifier "enough". People don't care enough to be portable. Unfortunately, Linux has made it hard to care enough. It's getting better, the documents are starting to say (accurately) what's Posix and what isn't, but there's a lot of legacy hackery out there. And it gets borrowed every day. And Linux is a GNU system. GNU isn't bad, as documents go, but one could wish (as you do) that something as central as glibc was better documented (or Emacs 20.3 Elisp for Mule...). Hrvoje> Didn't you yourself say you didn't care only a few Hrvoje> messages ago? sb> That's certainly a good reason. The only general way I can sb> write portable code is to have multiple systems to test on. I thought when you wanted portability you wrote in ELisp? :-) Only half a joke; Linux hackers, like Dose/Windose hackers, often go direct to the OS when there are perfectly good library calls. And they are often amateurs (as I am myself), and aren't aware of the benefits of encapsulating system dependencies in separate modules and so on. They could learn, as I have done a little, but coding is more fun.... I don't think blaming the Linux _programmer_ is useful, but we should be aware that the Linux environment is probably going to be a source of portability issues for a while. -- University of Tsukuba Tennodai 1-1-1 Tsukuba 305-8573 JAPAN Institute of Policy and Planning Sciences Tel/fax: +1 (298) 53-5091