I do not want to pollute the debate but: -) I remember a software QA managanager responsible for "C" coding rules also not allowing us to use 'break', 'continue', or 'return' (in the middle of a function).
Although I find them 'cleaner' than goto, would not use goto, and certainly do use 'return' in the middle of functions, I also agree that some people might think the former do reduce code readibility - ex: I , somehow, do not feel good using 'continue' and 'break'. -) Having worked on a few lex/yacc projects, I remmember being troubled ,as I ported the resulting c code into an embedded environement, that the latter extensively used gotos - Was that done only because people are not supposed to look at the generated code - or is it also simpler to generate non-structured C code ? Regards, Philippe Peter Hansen wrote: > Maxim Kasimov wrote: >> uuuuf..., i don't requesting that "goto" was available in next versions >> of python, but i'm saying if it will be so, it will be easy and quickly > > _debug_ some skripts, _not only_ for commenting > > Then we go right back to Simon Brunning's question for > you: "How does goto help you to remove bugs?" > > Maxim, nobody is really saying that *you* cannot have > found goto useful in debugging (though we're curious > for a real example, rather than just rhetoric). What > most people are saying (roughly) is that after writing > tens (or hundreds) of thousands of lines of code, we > cannot think of examples where the value of "goto", > for debugging or otherwise, outweighs the incredible > damage it does to the structure and (thus) readability > of code. And keep in mind that many of us were raised > on languages like BASIC and yet we've learned, improved, > and moved on to the point where we fully understand the > perceived value of GOTO in the mind of a newbie, but > remain unconvinced that even (or especially!) for a newbie > it is a good idea to have it available. > > So far, nobody on the "use goto!" side of the fence has > presented arguments that haven't already been shot down > dozens of times in this newsgroup and elsewhere. Of > course, we should all take this hint as a reminder that > this is a "religious" issue and that this particular > thread is not going to settle it once and for all. > > -Peter -- http://mail.python.org/mailman/listinfo/python-list