poser@csli.Stanford.EDU (Bill Poser) (10/03/90)
I'm curious about the problem some people apparently have with the fall-through property of switch. I've never had the slightest difficulty with this, even when I was first learning C. Modestly putting aside the attractive hypothesis that I am a genius at learning programming languages, I wonder if the people who have a problem are those who have previously learned a similar construction in a language that lacks fall-through, such as the Pascal case statement? When I learned C, I knew Fortran (Fortran IV, the real thing, not something semi-modern like F77), a bit of BASIC, MIX assembler (only on paper), and the assembler for an obscure Japanese laboratory minicomputer. None of these had anything like switch, so I had no preconceptions about how it should work. But an awful lot of people these days come to C after learning Pascal. Are they the people who are having trouble with this? Bill
ark@alice.att.com (Andrew Koenig) (10/03/90)
In article <15611@csli.Stanford.EDU>, poser@csli.Stanford.EDU (Bill Poser) writes: > I'm curious about the problem some people apparently have > with the fall-through property of switch. Some people have difficulty understanding things that aren't Politically Correct. -- --Andrew Koenig ark@europa.att.com
stelmack@screamer.csee.usf.edu (Gregory M. Stelmack) (10/03/90)
OK, I can't resist anymore :-) I thought one of the nice things about C was the programmer's ability to do anything, regardless of whether or not it makes sense. If you use the break; you can separate each case. If you don't, you allow follow-through. It seems to me that if this flexibility makes it hard for you to work with, maybe C isn't the language for you. I personally like having the compiler allow me to do whatever I want without delving into Assembler -- that's why I use C instead of other languages. Am I off-base with this opinion? -- Greg Stelmack -- Email: stelmack@sol.csee.usf.edu -- USmail: USF Box 1510, Tampa, FL 33620-1510 -- Amiga: the only way to compute!
dkeisen@Gang-of-Four.Stanford.EDU (Dave Eisen) (10/03/90)
In article <11428@alice.att.com> ark@alice.att.com (Andrew Koenig) writes: >In article <15611@csli.Stanford.EDU>, poser@csli.Stanford.EDU (Bill Poser) writes: > >> I'm curious about the problem some people apparently have >> with the fall-through property of switch. > >Some people have difficulty understanding things that aren't Politically >Correct. And some of us have problems with unjustified snipes. I doubt anyone has difficulty understanding the semantics of switch, they are simple enough. That doesn't mean we agree with them. Any decent C programmer uses switch correctly every day of eir life, but I doubt that I am alone in having omitted it by accident once or twice and having been left with annoying bugs. We don't want to take away your precious fall-through cases, I'm glad that option is available too. But the each case separate type of switch clearly is the more common type, is much easier to follow, and is more correct politically, err, I mean is more structured. Besides, it always seems silly to me to have to write break a dozen times in a long switch to do what in my mind is the canonical operation. Changing the default to breaking and allowing the word continue to be used for fall-through where would make switch statements cleaner and would save typing in both the fall through and other case --- we wouldn't have to type "break;" and we wouldn't have to type "/* YES --- I REALLY MEAN TO FALL THROUGH */." -- Dave Eisen Home: (415) 323-9757 dkeisen@Gang-of-Four.Stanford.EDU Office: (415) 967-5644 1447 N. Shoreline Blvd. Mountain View, CA 94043
ark@alice.att.com (Andrew Koenig) (10/04/90)
In article <1990Oct3.160403.517@Neon.Stanford.EDU>, dkeisen@Gang-of-Four.Stanford.EDU (Dave Eisen) writes: > In article <11428@alice.att.com> ark@alice.att.com (Andrew Koenig) writes: > >In article <15611@csli.Stanford.EDU>, poser@csli.Stanford.EDU (Bill Poser) writes: > >> I'm curious about the problem some people apparently have > >> with the fall-through property of switch. > >Some people have difficulty understanding things that aren't Politically > >Correct. > And some of us have problems with unjustified snipes. I quite agree. However, I went back and read the original posting and my response and cannot find any unjustified snipes there. To what are you referring? -- --Andrew Koenig ark@europa.att.com
enag@ifi.uio.no (Erik Naggum) (10/04/90)
In article <1990Oct3.160403.517@Neon.Stanford.EDU> dkeisen@Gang-of-Four.Stanford.EDU (Dave Eisen) writes:
Besides, it always seems silly to me to have to write break a dozen
times in a long switch to do what in my mind is the canonical
operation.
May I suggest
#define Case break; case
--
[Erik Naggum] Naggum Software; Gaustadalleen 21; 0371 OSLO; NORWAY
I disclaim, <erik@naggum.uu.no>, <enag@ifi.uio.no>
therefore I post. +47-295-8622, +47-256-7822, (fax) +47-260-4427browns@iccgcc.decnet.ab.com (Stan Brown, Oak Road Systems) (10/05/90)
About the fact that a case in a switch statement falls through to the next case unless there's an explicit break: In article <1990Oct3.160403.517@Neon.Stanford.EDU>, dkeisen@Gang-of-Four.Stanford.EDU (Dave Eisen) writes: > I doubt anyone has difficulty understanding the semantics of switch, > they are simple enough. That doesn't mean we agree with them. Any decent > C programmer uses switch correctly every day of eir life, but I doubt > that I am alone in having omitted it by accident once or twice and > having been left with annoying bugs. Bingo. The above is my own opinion and not attributable to any other person or organization. email: browns@iccgcc.decnet.ab.com Stan Brown, Oak Road Systems, Cleveland, Ohio, U.S.A. (216) 371-0043
roy%cybrspc@cs.umn.edu (Roy M. Silvernail) (10/05/90)
dkeisen@Gang-of-Four.Stanford.EDU (Dave Eisen) writes: > Changing the default to breaking and allowing the word > continue to be used for fall-through where would make switch statements > cleaner and would save typing in both the fall through and other case --- Continue, though, is already a keyword for loop control. Giving it two seperate and context-dependant tasks would, IMHO, _really_ screw people up. I'll keep the breaks, thank you, and the inherent lack of constriction C allows. -- Roy M. Silvernail |+| roy%cybrspc@cs.umn.edu |+| #define opinions ALL_MINE; main(){float x=1;x=x/50;printf("It's only $%.2f, but it's my $%.2f!\n",x,x);} "This is cyberspace." -- Peter da Silva :--: "...and I like it here!" -- me
kpicott@alias.UUCP (Socrates) (10/05/90)
In <1990Oct3.160403.517@Neon.Stanford.EDU> dkeisen@Gang-of-Four.Stanford.EDU (Dave Eisen) writes: >Changing the default to breaking and allowing the word >continue to be used for fall-through where would make switch statements >cleaner and would save typing in both the fall through and other case --- >we wouldn't have to type "break;" and we wouldn't have to type >"/* YES --- I REALLY MEAN TO FALL THROUGH */." >Dave Eisen Home: (415) 323-9757 As long as we're changing the definition of C why not just add another keyword "fallthrough" to be used instead of "continue"? That way we have our structuring, we have our ability to fallthrough, and the loop functions "break" and "continue" can be consistent in their usage. -- Kevin Picott aka Socrates aka kpicott%alias@csri.toronto.edu Alias Research Inc. 110 Richmond St. E. Toronto, Ontario... M5C 1P1 (416) 362-9181 x336
dkeisen@Gang-of-Four.Stanford.EDU (Dave Eisen) (10/05/90)
In article <RVJJq3w163w@cybrspc> roy%cybrspc@cs.umn.edu (Roy M. Silvernail) writes: >dkeisen@Gang-of-Four.Stanford.EDU (Dave Eisen) writes: > >> Changing the default to breaking and allowing the word >> continue to be used for fall-through where would make switch statements >> cleaner and would save typing in both the fall through and other case --- > >Continue, though, is already a keyword for loop control. Giving it two >seperate and context-dependant tasks would, IMHO, _really_ screw people >up. Break, though, is already a keyword for loop control. Giving it two separate and context-dependent tasks would, IMHO, _really_ screw people up. Oh hell, I guess they'll muddle through somehow. -- Dave Eisen Home: (415) 323-9757 dkeisen@Gang-of-Four.Stanford.EDU Office: (415) 967-5644 1447 N. Shoreline Blvd. Mountain View, CA 94043
jh4o+@andrew.cmu.edu (Jeffrey T. Hutzelman) (10/06/90)
dkeisen@Gang-of-Four.Stanford.EDU (Dave Eisen) writes: > In article <RVJJq3w163w@cybrspc> roy%cybrspc@cs.umn.edu (Roy > M. Silvernail) writes: >> Continue, though, is already a keyword for loop control. Giving it two >> seperate and context-dependant tasks would, IMHO, _really_ screw >> people up. > > Break, though, is already a keyword for loop control. Giving it two > separate and context-dependent tasks would, IMHO, _really_ screw > people up. Perhaps, but break already _has_ been given that meaning and plenty of C programmers deal with it all the time with no problem, and, IMHO, it makes sense. ----------------- Jeffrey Hutzelman America Online: JeffreyH11 Internet/BITNET:jh4o+@andrew.cmu.edu, jhutz@drycas.club.cc.cmu.edu >> Apple // Forever!!! <<
kdq@demott.COM (Kevin D. Quitt) (10/06/90)
In article <ENAG.90Oct4022413@hild.ifi.uio.no> enag@ifi.uio.no (Erik Naggum) writes: >In article <1990Oct3.160403.517@Neon.Stanford.EDU> dkeisen@Gang-of-Four.Stanford.EDU (Dave Eisen) writes: > > Besides, it always seems silly to me to have to write break a dozen > times in a long switch to do what in my mind is the canonical > operation. > >May I suggest > > #define Case break; case Perhaps even better (because it won't annoy the maintenance programmer) is to define yourself an editor macro called something like "make-switch-case" which inserts the "case :<NL> and "break;" (properly tabified, of course), and leaves your cursor in front of the colon. This does assume you are using a reasonable editor. -- _ Kevin D. Quitt demott!kdq kdq@demott.com DeMott Electronics Co. 14707 Keswick St. Van Nuys, CA 91405-1266 VOICE (818) 988-4975 FAX (818) 997-1190 MODEM (818) 997-4496 PEP last 96.37% of all statistics are made up.
roy%cybrspc@cs.umn.edu (Roy M. Silvernail) (10/07/90)
dkeisen@Gang-of-Four.Stanford.EDU (Dave Eisen) writes: > In article <RVJJq3w163w@cybrspc> roy%cybrspc@cs.umn.edu (Roy M. Silvernail) > writes: > >Continue, though, is already a keyword for loop control. Giving it two > >seperate and context-dependant tasks would, IMHO, _really_ screw people > >up. > > Break, though, is already a keyword for loop control. Giving it two > separate and context-dependent tasks would, IMHO, _really_ screw people > up. > > Oh hell, I guess they'll muddle through somehow. Actually, Dave, I didn't really present my point clearly. I should have said "seperate, context-dependant and _differing_ tasks". The break keyword, of course, is used now to: 1) terminate a case in the switch statement. 2) force immediate termination of a for, while or do loop. Thinking about this, I find both uses consistant with each other, as they both stop the current iteration of a control structure. The suggested addition to continue, though, isn't consistant. Currently, continue causes the current control structure to reiterate. The suggested use within a switch would do exactly the opposite, and cause the control structure to progress. A switch can be viewed as a loop that only executes once. Redefining continue within a switch? Just a disaster waiting to compile. (ask the nameless AT&T programmer about break vs. if...) But you're right... we'll all muddle through. ;-) -- Roy M. Silvernail |+| roy%cybrspc@cs.umn.edu |+| #define opinions ALL_MINE; main(){float x=1;x=x/50;printf("It's only $%.2f, but it's my $%.2f!\n",x,x);} "This is cyberspace." -- Peter da Silva :--: "...and I like it here!" -- me