Path: utzoo!utgpu!jarvis.csri.toronto.edu!clyde.concordia.ca!uunet!datapg!com50!pai!erc From: erc@pai.UUCP (Eric Johnson) Newsgroups: comp.sys.next Subject: Re: Port of g++ Summary: Because it may be useful Message-ID: <814@pai.UUCP> Date: 2 Jan 90 16:33:44 GMT References: <3897@orion.cf.uci.edu> <246300081@uxa.cso.uiuc.edu> <213@toaster.SFSU.EDU> Organization: Prime Automation, Inc., Burnsville, MN Lines: 134 In article <213@toaster.SFSU.EDU>, eps@toaster.SFSU.EDU (Eric P. Scott) writes: > In article <246300081@uxa.cso.uiuc.edu> jpd00964@uxa.cso.uiuc.edu writes: > >As I have seen this question several times in the past, I was > >curious, why would anyone want to port g++ to the NeXT? > > For the same reason they're clamoring for X Windows, MS-DOS > emulation and floppy disk drives. > > They just don't understand. Are you willing to concede that maybe--just maybe-- they do understand something? If not, save your time and go to the next message. > (Although there is merit in being > able to say one has the complete GNU suite running on any > particular platform, that's not the point. The problem is when > Marketing departments perceive requests for the hideous, wretched > things that crippled other products as "consumer demand" and > destroy what otherwise would have been an outstanding > achievement. What do you get when you "follow" everyone else > because it's "safe" and punish originality and creativity? > 90% of what's on the market! America once had a reputation for > ingenuity and excellence. Japan, et al. didn't "do us in." We > did it to ourselves. It's not too late to break the cycle.) (I won't respond to the Japan-bashing as I don't find that fruitful.) > As I understand it, NeXT is going to return their gcc with the > Objective-C support to FSF, in which case I concur--why bother > with g++? (In the absence of a LARGE amount of preexisting > code written with g++ in mind, a/k/a "the FORTRAN argument.") > > -=EPS=- > standard disclaimer Sorry, I disagree. Why would you want LISP on the NeXT? Come on, this language dates back to the 50s. Why? Because it is useful to many people for solving problems using a computer. LISP is also standardized, with Common LISP and CLOS (Common LISP Object System), so that people who have learned LISP on other systems (be they Symbolics or PCs) can transfer that knowledge to the NeXT and use that knowledge in solving problems. Why do people use these computers? To solve problems. (I know few people who can afford a NeXT merely to play games. There is usually some other reason involved.) The NeXT, with many software features built in, has many tools to help solve problems using the computer. There is no reason to decry those who want other tools as well, in addition to the fine tools provided by NeXT. In particular, G++ is a freely-distributable implementation of C++ (an object-oriented superset of the C language) from the GNU project. While Objective-C is also an object-oriented superset of C, the two langauges are not the same. Regardless of whatever position you take in the C++/Objective-C wars, many people have invested a large amout of time learning C++ and they may want to use those skills on the NeXT. C++ does seem to have an edge in the object-oriented wars (a much larger playfield than just Objective-C vs. C++). By the way, since you don't seem to like C++/G++, are you also against the following languages being ported to the NeXT: Smalltalk, Simula or Eiffel? And how about Modula-2, Ada, Oberon or Pascal? (You already have C, why would anyone ever want to use any other procedural language? :-) The NeXT offers LISP, a language very popular in Artificial Intelligence circles (especially in the USA). What about Prolog? You already have LISP, so who needs Prolog? :-) (Prolog seems to be very popular in Europe, much more popular than in the USA.) The X Window System is a graphical windowing system that runs on virtually every type of computer--from PCs (IBM PCs, Amigas, Macintoshes) to Cray supercomputers. X provides a standard protocol to distribute graphical applications over a network (I'm writing this message right now on a Sun SPARCStation as my X display server, and remotely logged in into a Hewlett-Packard 320--which runs our Usenet software.) X provides a portability not found with any other graphical windowing system. NeXTStep is now available only for the NeXT and IBM's yet-to-be-released RIOS workstations (correct me if I am wrong). NeWS is available on Suns and a few other systems (Silicon Graphics for example, I believe). X is available. X is portable. X is free (via FTP) or rather inexpensive (when ordering a tape from MIT or ISC, et al.). Many vendors have released their own implementations of X (for example, Hewlett-Packard and DEC), so if I don't like MIT's version, I can go somewhere else. Sun has even merged their proprietary windowing system (NeWS) with X11 (called OpenWindows), so you can get the best--and the worst--of both systems. Don't get me wrong, X has many faults. But so far, the advantages far outweigh the faults. In particular, I don't like a lot of the complexity nor the size of X (and X applications), but I respect software that can provide solutions on many, many different platforms. And, when I need a solution to be able to run on an HP, a Sun, a Data General, et al., I see X as the only choice available today. Others will have similiar reasons for wanting to run MS-DOS software (remember there are millions and millions of those buggers, however braindead they may be), or an inexpensive means of software distribution (hence the cry for floppies on the NeXT). If you use the NeXT, a cheap software distribution mechanism is good. You should want a lot of people to buy the NeXT (lowering costs with volume) and a lot of companies to produce software for the NeXT. In summary then, many people need (or perceive a need for) many different tools. Just because the NeXT offers some tools, it does not mean that those tools are the only appropriate tools for all users of computers (or all users of the NeXT). Remember, if the NeXT is to be successful, you want these other tools to be ported to the NeXT (if people are using these tools on the NeXT, remmember they are using--and buying--the NeXT). Have fun, -Eric -- Eric F. Johnson, Boulware Technologies, Inc. 415 W. Travelers Trail, Burnsville, MN 55337 USA. Phone: +1 612-894-0313. erc@pai.mn.org - or - bungia!pai!erc (We have a very dumb mailer, so please send a bang-!-style return address.)