 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Microcontroller and PC projects : Popping the Gosub stack...
Page 1 of 3   |
|||||
| Author | Message | ||||
Grogster Admin Group  Joined: 31/12/2012 Location: New ZealandPosts: 9063 |
Hi everyone. 
Is there any easy way to pop the gosub stack? Sometimes, I need to exit from a gosub WITHOUT a return, and go somewhere else in the code. You can do this(goto some other routine) without explicitly issuing a return for the gosub you are leaving, but it is bad practice, as you end up with the situation where returns from gosubs are dropping off the stack(depending on how many levels of gosub you can have in MMBASIC). Back in the days of the Atari, they had a POP command, which was specifically used to pop the gosub stack, and bring you "Up one level" in the gosub stack, so that you did not end up with stack overflow errors. You would use that POP command just before you left the gosub routine, if you were not leaving the gosub by the normal RETURN route. I don't suppose there is any way to do this in MMBASIC is there? Smoke makes things work. When the smoke gets out, it stops! |
||||
kiiid Guru  Joined: 11/05/2013 Location: United KingdomPosts: 671 |
This is something that you shouldn't be needing in a well written program. Even using the GOTO command is commonly considered as a bad programming style. The best way to deal with this is just to redo this part of the code. Even if there was such thing as popping a RETURN from the stack, it will be a really bad way to go and a potential source of many other issues in the program. http://rittle.org -------------- |
||||
| Grogster Admin Group Joined: 31/12/2012 Location: New ZealandPosts: 9063 |
REALLY? I used it with great success in the Atari years ago...... It's not that easy, really, to exit from a Gosub with a Return, if you need to be somewhere else, not back at the point where the Gosub was called from. Smoke makes things work. When the smoke gets out, it stops! |
||||
| Grogster Admin Group Joined: 31/12/2012 Location: New ZealandPosts: 9063 |
Now you're taking the piss. 
Goto another line number or label is a native command in just about ANY language you care to nominate. How do you justify that statement?  Smoke makes things work. When the smoke gets out, it stops! |
||||
BobD Guru Joined: 07/12/2011 Location: AustraliaPosts: 935 |
The answer to the exit from a sub routine is see the EXIT command. Kon is not taking the piss. I am sure he is deadly serious about GOTOs and exiting sub routines early. Use DO LOOPs etc. rather than GOTOs. The idea is called structured programming and you have a main sequence which calls sub routines and functions. Other languages call them procedures and functions. All the same thing. Try Googling structured programming. |
||||
| kiiid Guru Joined: 11/05/2013 Location: United KingdomPosts: 671 |
Hey, this is not to drag you into some pointless disputes
Just stating what is a common knowledge. GOTO is considered as a "bad" command. This is the truth. I know it is native to BASIC, but that does not change the fact. Also, you can probably exit from a subroutine the hard way, but again, doing it does not mean it is good. Better post the piece of code which needs that and I am sure there will be another way. Just trying to help, that's all... http://rittle.org -------------- |
||||
djuqa Guru Joined: 23/11/2011 Location: AustraliaPosts: 447 |
Should N ever,ever or need to jump out of a Nested gosub/goto location. Goto is BAD programming practice. Gosub's ok but used correctly. Jumping directly out of nested gosubs would get any programming assignment a "F"ail mark . Yes several crap versions of BASIC had ability to POP the stack, but still is BAD practice. VK4MU MicroController Units |
||||
Geoffg Guru Joined: 06/06/2011 Location: AustraliaPosts: 3165 |
I agree with Konstantin, excessive use of GOTOs can lead to unreadable (and therefore buggy) code. Popping the stack would only make things worse. Geoff Graham - http://geoffg.net |
||||
| Grogster Admin Group Joined: 31/12/2012 Location: New ZealandPosts: 9063 |
Well bugger me. 
I have been using Gosubs and Goto's for years - ever since I first learned programming. I was totally unaware that Goto is considered bad programming - that really has come as something of a genuine shock. So, is it fair to say, that most code written these days tries to avoid Goto's completely? I will re-evaluate the routine that needed the POP, and see if I can teach myself some new tricks.
Oh, and BTW kiiid - I was not trying to be nasty(hence the smilies) in my post about you taking the piss. I truly thought you were taking the piss, as Goto is one of the first commands you learn in BASIC. Well, in the BASIC that I learnt years ago. Having said that, it is pretty well acknowledged, that Atari BASIC was not the best BASIC even then, but habits can be hard to break. 
EDIT: @ BobD - thanks for the link - forgot to mention that - will read all about it. ANOTHER EDIT: If Goto is such a bad command, and it would seem, is actively discouraged, why is it included in MMBASIC at all? I mean, if there are other alternatives to using Goto, it confuses me why it was allowed into the language at all when Geoff was developing it. Force the use of the more standard methods, rather then even providing a Goto command in the first place, if you see what I am getting at. Smoke makes things work. When the smoke gets out, it stops! |
||||
| kiiid Guru Joined: 11/05/2013 Location: United KingdomPosts: 671 |
Nothing nasty. We are all in the same team here
I remember back in my early days GOTO was one of the first things to learn indeed. Then Pascal got more popular and I had to switch not by own choice. Only then I learned the structural way BobD is talking about. I remember it was written in some fundamental book (can't recall its name), that a well written program should not include even a single GOTO. Couldn't believe back then... Now thinking in retrospect... I haven't used GOTO (even in BASIC) for many years. http://rittle.org -------------- |
||||
| Grogster Admin Group Joined: 31/12/2012 Location: New ZealandPosts: 9063 |
Wow. This is INTENSLY interesting. Smoke makes things work. When the smoke gets out, it stops! |
||||
| BobD Guru Joined: 07/12/2011 Location: AustraliaPosts: 935 |
It is a good move to minimise the number of GOTOs in a program. A level somewhere very close to or below zero is very good. MMBasic when it started out at V2.1 was nowhere near the same beast it is now. You could only use line numbers and many other commands and functions that we have now were not there. There was also the goal of compatibility with one of the other BASICs. While structured is good there is no need to force people to do one or the other. The gates for the discussion were opened and you can see the flood of well meaning advice that was unleashed. I would put a good smiley in here but I was too old when they came out and I never got the hang of them. I don't want to put the wrong one. |
||||
| Grogster Admin Group Joined: 31/12/2012 Location: New ZealandPosts: 9063 |
Ahhhh - I see. Good point.
HELL yes. Very surprising indeed - I feel I have opened a can of worms. GOTO seeking worms 
LOL! - Laugh Out Loud - Smoke makes things work. When the smoke gets out, it stops! |
||||
Juri74 Senior Member  Joined: 06/02/2012 Location: ItalyPosts: 162 |
I agree that using gosub & goto commands it's very easy to insert bugs, after Geoff added the Function/Sub commands to MMbasic i stopped definitely to use gosub command and i mantain as low as possible the numbers of goto commands, but actually i can't reach 0 :D Grogster however if you use a Sub you can exit from it with a Goto command, but i will never do such thing, the problem is of course that the code tends to become unreadable. and if a bug come out start to be very difficult to fix it... i don't know how long is your program.. but think at a 1000 lines long prog... |
||||
| Grogster Admin Group Joined: 31/12/2012 Location: New ZealandPosts: 9063 |
Thanks, Juri74. Currently, my code length is 776 lines. Smoke makes things work. When the smoke gets out, it stops! |
||||
| Juri74 Senior Member Joined: 06/02/2012 Location: ItalyPosts: 162 |
ok i'm working at a game, my code length is a bit smaller than yours, 767 lines. have you noticed the general slowdown of the full screen editor after the 450th-500th line? especially when scrolling the program using keyup and keydown keys... |
||||
| Dylan Regular Member  Joined: 17/06/2013 Location: NetherlandsPosts: 81 |
http://en.wikipedia.org/wiki/Considered_Harmful in particular footnote 3 (pick a link). |
||||
MicroBlocks Guru Joined: 12/05/2012 Location: ThailandPosts: 2209 |
I use(d) pop and gotos (jumps) all the time. But that is in assembly. :) Higher level languages, as basic is, try to avoid that to minimize programming errors. Counting how many return addresses are on a stack is such a tricky business it should be left to assembly language. In assembly it is great to sometimes just pop 2 from the stack and then return if it saves codespace and processing time. A Goto statement is almost never used anymore. It is good that it is still there because sometimes there is no substitute and it can actually make a program more readable. Even in new languages like C# there still is a goto statement. And goto is still everywhere but under disguise using names like 'continue' and 'break' which can also cause spaghetti code. Goto at least has the decency to tell you right away where you are going. Break and continue send you on a search for the right opening or closing bracket. Microblocks. Build with logic. |
||||
| Dylan Regular Member Joined: 17/06/2013 Location: NetherlandsPosts: 81 |
The real danger of GOTO once you have structured constructs is that you should almost never want to jump into the middle of a subroutine, even in assembly. I believe I last used on in anger in Turbo Pascal back in the early 90's for the "standard" application, namely to get out from an inner loop to outside the outer. Nowadays the "break" version of that is to raise an exception, and of course if the innermost loop has found the answer to make it into a function (and "continue" at the end of the outermost loop). As a Pascal programmer used to case...of, it always seemed odd that the C family had a break inside switch'es, but once I learned assembler it made sense: after all, falling through saves a JMP. break and continue should, generally speaking, be very local. I've never had to search for brackets in my own code (however little I can initially understand of the functioning on revisiting a year later, it is at least readable.) It is nice to see that MMBasic has an ON...GOTO/GOSUB, which allows for menus very easily (ON ASC(INKEY$) GOTO 100, 100, 100 ...) but I imagine I would choose for the GOSUB over GOTO every time. |
||||
| JohnS Guru Joined: 18/11/2011 Location: United KingdomPosts: 3659 |
To come back to what it was wanted for, maybe you could change the SUB into a function and use its returned value to signal that you want to do something other than the current inline next statement. Some of the problems with GOTO: makes it hard for other people reading your code makes it hard to prove the code is correct - this is one of the really big issues makes it hard to maintain There are still, arguably, a few cases where GOTO makes sense and some of them are in the Linux kernel code. Basically, local jumps in smallish files where you can clearly see what's going on. But it's your code so do as you like, of course. John |
||||
| Page 1 of 3 |
|||||
 Print this page |
