From cowan@ccil.org Tue Sep 05 19:43:47 2000 Return-Path: Received: (qmail 15357 invoked from network); 6 Sep 2000 02:43:47 -0000 Received: from unknown (10.1.10.26) by m2.onelist.org with QMQP; 6 Sep 2000 02:43:47 -0000 Received: from unknown (HELO locke.ccil.org) (192.190.237.102) by mta1 with SMTP; 6 Sep 2000 02:43:45 -0000 Received: from localhost (cowan@localhost) by locke.ccil.org (8.8.5/8.8.5) with SMTP id XAA05897; Tue, 5 Sep 2000 23:38:33 -0400 (EDT) Date: Tue, 5 Sep 2000 23:38:33 -0400 (EDT) To: pycyn@aol.com Cc: lojban@egroups.com Subject: Re: [lojban] RE: emacs, etc. In-Reply-To: <34.a12ea97.26e6ee6b@aol.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-eGroups-From: John Cowan From: John Cowan On Tue, 5 Sep 2000 pycyn@aol.com wrote: > << Microsoft makes it hard for people to develop applications that > run well both on Windows and non-Windows. >> > > Whereas Linux, say, makes it easy to develop something that runs on that > system as well as Windows and Macs? Hooohah! Of course not: the barrier works both ways. Windows programs aren't portable off Windows, and Windows development is hardly possible except on Windows. Microsoft deliberately makes Windows development as complex as possible, so that only with liberal use of assistive tools is programming possible at all. (The raw C version of the Windows program that just displays "hello, world" is 300 lines long!) > My problem is not with Windows (etc.) per se but with the fact -- to continue > the example -- that of the seven critters I have called BASIC no program > written and running well in one will run under any of the others. (None of > them are really BASIC either, but that is another point -- though none of > them run the real stuff, either) With the exception of GM- and Q-, these > programs did not come with the machines but are produced by outside > providers, who provide similar services for at least various Apples as well > as DOS machines. The only hope is external standards, but the standard for Basic was deliberately kept weak by the vendors so they wouldn't have to do the hard work of interoperability. > But, in fact, programs written in such a BASIC for a DOS > will not run on the same company's "equivalent" system for another machine, > either. I had similar experiences with Forth (an anti-establishment > programming language, it says) Forth is so individualistic, the standard hardly helps much. > I do wonder how the designers of Java managed > to avoid the problems (aside from simple competence, that is -- but then I > don't exactly understand how Java works so competence may not be an issue) Sun defined the language *exactly* and then kicked hard, with lawyers, anybody who tried to implement it in ways that violated the definition. > Overcoming this > "little problem" would go a long way toward relieving the general problem of > which I have complained: There *are* languages that are highly portable: Perl and Python have only one implementation, which certainly helps. > too much stuff for the peripheral systems like unix > etal and nothing for the central systems, which most people (by a factor of a > 1000 remember) have. That's the way the vendors of the central systems want it. This is *not* a wacky conspiracy theory: it's the ruling of the antitrust judge. > A program in that could be compiled or interpreted on > any system (as the corresponding names falsely suggest they all can) would > take care of that more or less immediately, at no increase in price. And we > would all be happy. But then there would be no reason for us to pay two to three figures for Windows, would there? We could *all* run Linux for precisely nothing.. And Microsoft can't have that, can they? -- John Cowan cowan@ccil.org One art/there is/no less/no more/All things/to do/with sparks/galore --Douglas Hofstadter