Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.2 9/18/84; site cci-bdc.UUCP Path: utzoo!watmath!clyde!burl!ulysses!allegra!mit-eddie!cybvax0!cci-bdc!larry From: larry@cci-bdc.UUCP (Larry DeLuca) Newsgroups: net.lang Subject: Re: basic for UNIX? Message-ID: <150@cci-bdc.UUCP> Date: Tue, 19-Feb-85 10:13:59 EST Article-I.D.: cci-bdc.150 Posted: Tue Feb 19 10:13:59 1985 Date-Received: Fri, 22-Feb-85 07:12:22 EST References: <7873@brl-tgr.ARPA>, <706@ucbtopaz.CC.Berkeley.ARPA> <467@houxj.UUCP> Organization: Computer Consoles, Inc., Cambridge, MA Lines: 75 > > Why are people so intolerant of BASIC ? Every language > has some type of usefullness for some community of users. > BASIC is useful on Unix in at least two ways that I can > think of : it allows one to migrate software from BASIC-only > systems to the Unix-based system, and it allows a new Unix > user who doesn't know C to use a familiar language while > learning Unix. > > I never use BASIC myself for anything. But arguments that > it is worthless, inferior, outdated, etc are incorrect. > It may be totally useless to you if you know C and don't > have any software written in BASIC that you want to use. > But you are only one of the many types of computer user. > A version of BASIC that supports matrix operations may > be the best possible language in existence for a user > who needs to do lots of quick, throwaway programs that > do simple matrix operations. > There are several problems with this. First of which is that the power of most BASICs varies inversely as the size of the CPU (micro hackers use it a lot, mainframe hackers don't). Second, there is NO standard BASIC. You would probably want to cross DEC BASIC-Plus and Microsoft BASIC to come up with something reasonable (DEC has the matrix capabilities and a pseudo-structured approach that while inconsistent with many BASICs (though it supports standard con- structs as well) is certainly a win over the spaghetti method of control structure. Most versions of BASIC are also couched in an OS so deeply that they don't come out -- the OS goes with them. SMC BASIC comes not only with the interpreter but all the wonderful little utilites to go with it. Try to get it to speak to UNIX -- I defy you. The worst problem is that porting something like BASIC to the machine without thinking out a good UNIX interface (which would be particularly difficult in BASIC) is that it encourages people to KEEP the DAMN PACKAGE IN BASIC!!! When you're paying a bunch of people to program in C, you don't want to have to call in junior-high students to fix your packages. The UNIX interface to most BASICs is LOUSY to nonexistant. The two were just never designed with any thought as to the existance of the other. I had a HELL of a time writing some C code that was supposed to update transactions on a remote machine that were generated by BASIC code -- the worst problem was that neither environment was really aware of the other enough that it cared. This resulted in less elegant and efficient solutions to problems that would have been trivial had the entire system been written in C. > The same argument holds for any other language. Lots of > people use COBOL (which I can't stand) for good reasons : > they can't afford to rewrite all of the COBOL-based > software that they depend on, or their managers only > understand COBOL, or the government has adopted COBOL > as a standard, or some such reason. > > Bill Dietrich > houxj!wapd Yes, the same disadvantages hold true. The version of COBOL i dealt with recently had a fixed table that held the pointers to the C sub- routines that you linked in. The fixed limit was 16, and there was NO argument passing. *** SELECT THIS-SPACE ASSIGN TO YOUR-MESSAGE. *** larry... -- uucp: ..mit-eddie!cybvax0!cci-bdc!larry arpa: henrik@mit-mc.ARPA This mind intentionally left blank.