0 Favourites

Will the top level design be removed in C3?

  • Wonder if the top level design/problem will be removed from C3 as its really not good in my opinion. Especially because its kind of encourage, especially new users, to solve problems using wait = 0, which is a bad way to handle problems in general, even though in some cases its really difficult to avoid it. But every time you do or are forced to do it, you are almost 100% certain that you are going to have problems later on. So wondering if that design is going to be removed from C3? So the Wait functionality can be removed as well, for good?

  • i don't see why anyone would use wait(0) since it calls the wait function not to wait. but i've had a couple of good wait usages and i find it pretty usefull.

    also i don't see what is your "top level design" problem? if you are thinkin' on gui/UI i don't see how removing that would solve your problems

  • i don't see why anyone would use wait(0) since it calls the wait function not to wait. but i've had a couple of good wait usages and i find it pretty usefull.

    also i don't see what is your "top level design" problem? if you are thinkin' on gui/UI i don't see how removing that would solve your problems

    What im referring to is the problem with objects not existing before next top level when created, which you can "solve" by throwing in a wait = 0. Which is part of the top-level design problem or what to call it, there are other issues which can be solved with just throwing in waits.

    But a functionality in the form of a wait is a poor design decision in my opinion in general. Especially when its used to solve varies issues like working with newly created objects, but more, why on earth would you ever make a design that would force the program to wait? Unless you actually want the user to wait, but then there are timers way better suited for that, as they have built in triggers. I don't disagree with you that waits can be useful, but this is not because they are needed, its because C2 is designed the way it is. But if that was redesigned you wouldn't have to use Waits at all. It have nothing to do with the UI/GUI.

  • Wait( 0 ) is good to use as an Event Sheet LateUpdate. An update that occurs after everything else in EventSheet and Plugins, yet also triggers before the next tick. It's actually very helpful in a few instances as it's one of the few techniques that allows some measure of finite control. However it is a problem due to how it works. Personally the only time to use wait are for automated sequences. if at any point there is a possibility of interuption or reset, then wait will hurt. It's best to use Timer.

    I'm not sure about the top level reference either. Are we talking about UI or are we talking about how all programming is in C2 is Functional and not say Object Orientated, which would be I suppose downward up.

    -------------------------

    Replied before I sents.

    Wait doesn't hold the program. Wait encapsulates the current set of information and starts a JS Timer to execute after the given period of time. However once doing so obliviously moves that piece of code out of our control.

  • Actually wait has it's uses, you just haven't encountered them yet.

    Like pressing a sprite button. You switch the buttons animation, wait one second, and switch it back.

    Having to also set up a trigger to switch it back is as silly as wait(0).

    or like not allowing a key to work for 3 seconds after it's pushed:

    (I didn't already have an image of the first example.)

    Wait could be made clearer to the user, when you make a wait action, it immediately forms a little subevent, so that it's clear the only thing waiting is the code in that subevent. (however making that kind of subevent look right might be a challenge.

    Knowing how things work on the back end makes it obvious why you "wait" until the next tick, but I suppose it could be built into the problem events. (they all include a hidden wait(0) the user doesn't know about.)

  • It's actually very helpful in a few instances as it's one of the few techniques that allows some measure of finite control. However it is a problem due to how it works. Personally the only time to use wait are for automated sequences. if at any point there is a possibility of interuption or reset, then wait will hurt. It's best to use Timer.

    I'm not sure about the top level reference either. Are we talking about UI or are we talking about how all programming is in C2 is Functional and not say Object Orientated, which would be I suppose downward up.

    -------------------------

    Replied before I sents.

    Wait doesn't hold the program. Wait encapsulates the current set of information and starts a JS Timer to execute after the given period of time. However once doing so obliviously moves that piece of code out of our control.

    As I said its not because Wait() can't be useful, my point is that its shouldn't be needed at all, at least not when it comes to solving problems, or what to say.

    I don't mean like if you throw in a wait of 10, then nothing will happen in the program for 10 seconds. What I mean is that you create a "fake" break in your program that shouldn't be needed.

    Actually wait has it's uses, you just haven't encountered them yet.

    Like pressing a sprite button. You switch the buttons animation, wait one second, and switch it back.

    Having to also set up a trigger to switch it back is as silly as wait(0).

    or like not allowing a key to work for 3 seconds after it's pushed:

    Using wait for something like a button and the audio in your examples can be ok. But you might as well hook up a timer. That when its up, triggers and you act on that. So my point in general is not that Wait() is broken and can't be used at all. But again it shouldn't be necessary to use to solve problems. You using it for a button or for the audio is fine. But you are just showing how it can be useful and that is not what im talking about.

    So to show you the problem and this is of course a simple example, but if you have a huge game and you solve problem using Wait() its could/will cause problems.

    Imagine we have a game where we spawn a Unit (Red square) based on this unit we want to spawn something else (Green square), but depending on some settings we might spawn different things. In this example I have only added one which are the Green square, just to keep it simple. But imagine it could be a Blue, Yellow, Black or a Pink square, all depending on some value when the Red square is spawned.

    So the Red square have two variables A and B and we assume that based on whatever, we ended up spawning the Green Square and we store the UID of the Red Square in the Green one. So when you hit the button to create the squares, we want Red Sq. A to be 1 and B to be 2 and write the UID stored in the Green Sqr. as well.

    In the top part of the image when I click the button it fails so all it writes is 0.

    Adding a Wait(0) in the Green square create event and then trying again, will solve the problem and now the values are correctly read and written to the text objects.

    So Wait(0) can be (mis)used to solve problems like this to fake It going to top level, instead of just having the Red and Green squares exist straight away, so the wait is not needed. So the code would do the same thing but the Wait(0) could be removed.

  • O star of layout- set text to "hi"

    -wait 5 sec

    -set text to "This is an example of a timeline"

    -wait 5 sec

    -set text to "wait can do other things besides text"

    -wait 5 sec

    -set text to "like make a trigger out of a variable"

    -wait 1 sec

    -set variable "dostuff" to 1

  • O star of layout- set text to "hi"

    -wait 5 sec

    -set text to "This is an example of a timeline"

    -wait 5 sec

    -set text to "wait can do other things besides text"

    -wait 5 sec

    -set text to "like make a trigger out of a variable"

    -wait 1 sec

    -set variable "dostuff" to 1

    And again Wait in itself is not a problem and yes there are lots of examples of how it can be used And your example is also a fine way to use it, but again it could be done with a Timer just as well. But its fair enough to leave Wait and not remove it, its not really that huge of a problem.

    But If you as in my example have to create several objects which depends on each other in one go, then its not all that simple as it allow Wait() to be misused, not because there is something wrong with wait itself, but because C2 when creating objects doesn't make them available before the next top level.

  • i see what you're tryin' to do.

    you want to use the object you created at the same moment as it was created (in the same tick).

    so basically what you do is use wait ( 0 ) to go to next tick and continue where the it left off (at that function).

    by default that is the expected behaviour - suggesting that your desing is bad. and probably top level won't be removed any time soon, but if you design your game good it's not even needed. also if you're aiming for the instant spawning of more elements that need to be "stuck together" and you want to do that in one tick - use containers. that's what they're used for. also "on created" is used for defining start state of an created object. but only after one tick is done your object gets UID, therefore you never had any output without wait(0).

    also, if you aim 60+ fps, have in mind that 1 tick is 0.016s (1/60)s which is really fast, and if you pick your objects in next ticks that won't look / work wierd. i know it takes some time to design the game correctly, but it works good.

    also what newt told you - do an action and if you want to set up some basic object items do it on created, otherwise use wait for some behaviours/events in your game that need to happen after some time /at some defined moment. instantly picking item that you created won't do much since that tick you're in has to PASS for object to really appear there.

  • i see what you're tryin' to do.

    you want to use the object you created at the same moment as it was created (in the same tick).

    so basically what you do is use wait ( 0 ) to go to next tick and continue where the it left off (at that function).

    by default that is the expected behaviour - suggesting that your desing is bad. and probably top level won't be removed any time soon, but if you design your game good it's not even needed. also if you're aiming for the instant spawning of more elements that need to be "stuck together" and you want to do that in one tick - use containers. that's what they're used for. also "on created" is used for defining start state of an created object. but only after one tick is done your object gets UID, therefore you never had any output without wait(0).

    also, if you aim 60+ fps, have in mind that 1 tick is 0.016s (1/60)s which is really fast, and if you pick your objects in next ticks that won't look / work wierd. i know it takes some time to design the game correctly, but it works good.

    also what newt told you - do an action and if you want to set up some basic object items do it on created, otherwise use wait for some behaviours/events in your game that need to happen after some time /at some defined moment. instantly picking item that you created won't do much since that tick you're in has to PASS for object to really appear there.

    Think you are missing the point. Using containers for something like this is not very good I think. In my example I said imagine that you have several options and you have to choose one at random, lets say you have 10 options. Then when you spawn your object you spawn 10 other objects as well that is part of the container, to only use one of them, how is that a good design?

    And you are correct that the problem you run into is that it will be a bad design.

    "if you aim 60+ fps, have in mind that 1 tick is 0.016s (1/60)s which is really fast, and if you pick your objects in next ticks that won't look / work wierd. i know it takes some time to design the game correctly, but it works good."

    I don't think this is a valid reason. Computers doesn't work as we do, where you can say that if we don't see it, then the computer wont either. It notice every tick, and if something in your code does something in that 1 tick it gives you problem whether you see it or not.

    Newt just gave an example of how wait works, but I can come up with lots of example of how to use it where it works fine. But im not interested in that.

    So try this and you can use my example if you want, but when you make it and spawn the Red square, make it so, depending on a random roll of 3, which we assume is some important setting for the Red square, you make it spawn another object depending on the roll all in one go and print out the values from both of them. Its not unrealistic that your game when creating a unit might spawn something else that rely on some setting from the first unit.

  • i was using containers as an explanation for how simultaneous objects should be created in general, not for your example. trust me, containers are GREAT, and they lower the amount of coding. also, when it comes to your example, i don't see the problem. random 1-3, have a var which defines what is 1,2,3, spawn depending on the number. also i was talking about design in general.

    that's not true. during my time of development with c# / c++ i've noticed a lot of things that don't work as expected and need workaround because computers calculate them differently. the same is with this wait thing. you either accept it, or leave it. you will see much more trouble with unity / ue4 / any engine. but it all depends how masochistic you are

    to work your ass off on something. and in the end it all comes down to learning how computers work, how interpreters work, how compilers work, to just fix that ONE bug.

    of course it's not impossible that something relies on something else. but you used "on created" that was connected to another "on created" that was calling a function that does something. you can't do that in 1 tick since you've got a chaining of events. that's why wait (0) stops the chain before calling function for 1 sec so that the objects are completed and then continues from there to get the right values and print them out.

    i'm pretty sure you don't know what happens, and how c2 engine works.

  • Construct 3

    Buy Construct 3

    Develop games in your browser. Powerful, performant & highly capable.

    Buy Now Construct 3 users don't see these ads
  • i was using containers as an explanation for how simultaneous objects should be created in general, not for your example. trust me, containers are GREAT, and they lower the amount of coding. also, when it comes to your example, i don't see the problem. random 1-3, have a var which defines what is 1,2,3, spawn depending on the number. also i was talking about design in general.

    that's not true. during my time of development with c# / c++ i've noticed a lot of things that don't work as expected and need workaround because computers calculate them differently. the same is with this wait thing. you either accept it, or leave it. you will see much more trouble with unity / ue4 / any engine. but it all depends how masochistic you are

    to work your ass off on something. and in the end it all comes down to learning how computers work, how interpreters work, how compilers work, to just fix that ONE bug.

    of course it's not impossible that something relies on something else. but you used "on created" that was connected to another "on created" that was calling a function that does something. you can't do that in 1 tick since you've got a chaining of events. that's why wait (0) stops the chain before calling function for 1 sec so that the objects are completed and then continues from there to get the right values and print them out.

    i'm pretty sure you don't know what happens, and how c2 engine works.

    Sorry Saiyadjin, but you speak to me like you assume that I have no clue how C2 works, which of course is fair enough if that's how you see it. But I did make you aware of this issue in the first place, which is known by Scirra, that it is designed that way. Why I asked about it in the first place. I fully respect if you do not see a problem with this, but that doesn't make it go away. And I made a simple example of how and when it happens, not because I was trying to show good design, but to explain the problem to those people, including yourself that related this issue to being maybe about the UI. If you have never used a wait() in your code to solve varies issues even though you knew it was bad, then I fully understand why you haven't noticed this.

    Besides that im not comparing to other software solutions and whether they have problems or not, because its of no concern to me. But if you have experience with C++/C# you know just as well as me, that you would never solve problems with creating objects by throwing in waits(). So why you would even compare those with C2, im not sure of.

    Besides that you have moved from my initial question, which I have no clue why you would even argue against in the first place, that if objects could be created so they exist instantly, that it wouldn't be better than having to wait? to how I made the example, which shows the problem and then explain to me, what happens in it? Since I made the example, im well aware of what it does, again it weren't made to show good design, but to show where the problem occurs and in what situations. And if you can solve it some other way under the same conditions that I put up, then I would be glad to see it and learn from it. But still why defend how it works now, if it can be changed in C3 so this wouldn't even be an issue, since this is a part of the core of how C2 works, it would make sense to change it for C3 if possible, which is why I asked the question and not to be shown 10 examples of how to use waits for varies things, that have nothing to do with what this is about.

  • I'm guessing you're referring to when you can pick new objects.

    New object's do exist instantly, but here's info on what's happens with regard to picking new objects:

    It's basically do to how the object filtering works in the event system. The question is how to implement it better.

    Maybe having the ability to make lists of objects. It would be a bit tricky when dealing with destroyed objects though.

    I'd like to think the whole event system could be overhauled to be simpler and more powerful. I've been trying to come up with a design that addresses things like picking two separate instances and this toplevel thing in a simple way. Imo a good design is key before I make argument how I think things could be better. Maybe eventually I'll find that holy grail of perfect simple design.

  • I'm guessing you're referring to when you can pick new objects.

    New object's do exist instantly, but here's info on what's happens with regard to picking new objects:

    It's basically do to how the object filtering works in the event system. The question is how to implement it better.

    Maybe having the ability to make lists of objects. It would be a bit tricky when dealing with destroyed objects though.

    I'd like to think the whole event system could be overhauled to be simpler and more powerful. I've been trying to come up with a design that addresses things like picking two separate instances and this toplevel thing in a simple way. Imo a good design is key before I make argument how I think things could be better. Maybe eventually I'll find that holy grail of perfect simple design.

    In the picking selection window, just add txt input where you place a name of "set pick group" , or even just a dropdown under same name with groups "A" "B" "C" and "D". Sorted

  • I'm guessing you're referring to when you can pick new objects.

    New object's do exist instantly, but here's info on what's happens with regard to picking new objects:

    It's basically do to how the object filtering works in the event system. The question is how to implement it better.

    Maybe having the ability to make lists of objects. It would be a bit tricky when dealing with destroyed objects though.

    I'd like to think the whole event system could be overhauled to be simpler and more powerful. I've been trying to come up with a design that addresses things like picking two separate instances and this toplevel thing in a simple way. Imo a good design is key before I make argument how I think things could be better. Maybe eventually I'll find that holy grail of perfect simple design.

    Yeah think you make a good point and recall that thread you are linking too as well. How to solve it is of course the whole issue, also why I was curious to whether it was on the table when designing C3.

    One way I thought it might be possible or at least worth a consideration, could be something in the lines of creating "temporary" objects/data. Meaning that when a create takes place, all the information is stored temporary until the actually creation takes place, said in another way. Because you of course is correct that the object is created. But in lack of better terms I just refer to it as "not being created". But what I mean is that during the period until a top level is reached, maybe the information could be stored and all interactions etc. would then be done on this temporary data/object and copied to the actually object when its ready. If that makes sense? But whether something like that is possible, a good solution or even worth it, I don't know.

    But that way you would be able to just destroy the temporary data, instead of the actual object, or said in another way never create it. Anyway just some thoughts about it. Hopefully Ashley have already figured something out, if its of course being considered

Jump to:
Active Users
There are 1 visitors browsing this topic (0 users and 1 guests)
Similar Topics Posts Views Last Post
Unread hot topic
0 Favourites
Spriter/C2 - Spriter 2 Teaser!
1,800 322,332
tarek2's avatar
tarek2
Unread hot topic
217 18,570
humanescape's avatar
humanescape
Unread hot topic
203 18,531
Lumicreative's avatar
Lumicreative