Basically the main thing is simply to make sure that things only happen for a given player or AI at a given time. Which sounds logic obviously

What you can do to control it is simply to make a Variable that keeping track of who's turn it is and make sure that its a condition all the places where its needed. When that player / AI is done you just change the variable to the next turn and so forth.

Using lots of Functions would be a good idea (in general its a good idea) But what they allow you to do, is to call the function with different variables. In which the Variable that keeps track of who's turn it is could be one of them. That way you can reuse pretty much all the code for both player and all the AIs, if they are suppose to be the same.

There ain't really a huge difference besides the above mentioned when it comes to a turn based game and a non turn based. Even in real time games things happens in turns, they are just so fast that you don't really notice it.

Meaning all units in a real time game doesn't reacted at the same time, they simply appears to.

There are several things that I would suggest, mostly to make things easier for yourself and so you might be able to troubleshoot what is going wrong.

Your picking of objects is a bit confusing, which I think make things harder for you. You shouldn't mind adding several picking sub levels. A good rule of thumb I think, is to always make sure that a Pick <something> is always the last thing you do in a picking event. If you need to pick more different objects, add a new sub level.

The first picking order should be:

(X) = line in your the picking

(1) Every 0.032 seconds <------------ This can be removed, as (2) will check every tick and doesn't really impact performance a lot.

(2) Collisionsprite is overlapping Front

(3) For each Collisionsprite <----- This will act as a Pick <Something> so will be the last condition in the first pick, so add a new sub level.

So the above two conditions will pick all the relevant collisionsprites that you are interested in.

In a new sub level pick, you add:

(4) Collisionsprite is overlapping front <----------- You don't need this, as you have already tested for that in (2), so you already know that you have the correct objects, so this should be deleted.

This is what is important for the Front object. So add those next or add them above line (2) in the first Pick event.

(5) ai_enabled = 0

(8) Front is not attacking

Something seems wrong with (6) and (7)

(6) System Pick.... Front where ....

(7) Pick instance with UID collisionspite.AI_ID

(I don't know what is stored in collisionsprite.AI_ID, But I would double check that what you are doing here is correct and working. But these two conditions seems opposite of each other and my guess is that your problem is found with these two picks.)

But assuming these are correct. you can add the last Pick here as well:

(9) Pick a random Front

The sub event (10)

I assume that if the "Front" object is alive it should cast a fireball. If thats the case i would move it to a separate Attack group that handles Attack of AI. Mixing things together like movement and Attacking etc. Is most likely going to cause problems for you at some point.

In your actions, you use a rotate behavior, which seems to complicate things, there is an action for sprites called rotate towards position, which will let you rotate the sprite at a certain speed towards a target. That way you can get rid of the timer and rotate behaviors.

Not sure why you disable collisions? It seems wrong, but thats just a guess and there might be a reason for it.

I would strongly advice that you never use Wait actions as it can cause serious problems, and should only in my opinion at least, be used if you are 100% sure it wont cause problems. Besides that adding the wait as the last action will as far as I can see have no effect unless you actually create an object that you need to change some settings for later during the tick. (meaning until the program have ran through all the code and started a new cycle (tick), if this makes little sense, it have something to do with when objects are created in Construct and when they actually exist in your program. So an object created during a tick will not be available before the next tick, so it needs to be initialize at the same level or a lower sub level...., The wait can be used to "fake" it, and that is where problems starts, as the wait doesn't actually make Construct wait, it simply fires off after the wait is over, which is when problems can occur and make it very difficult to pin point the exact cause of bugs, as these can happen with something that is completely unrelated to the place you used the wait. If it still makes no sense, just remember Wait = problems )

Last you have a check called "On collision" which will trigger once while its true, you could use that instead. However if you have a lot of objects that need to use it, it will hurt performance a lot more than using "Is overlapping".

I think you might misunderstand what the problem is or I didnt write it clear enough

I know that DOM elements are rendered on top of non DOM elements, which is fine as its not the problem, so its not important.

The issue is if you click and hold down the mouse button on a progressbar as if you wanted to drag it somewhere. (As I wrote in the initial report, this obviously wont work as it doesn't have dragdrop behaviour). But regardless of that, after you have done it, other objects like the sprites which actually have dragdrop behaviour will no longer work. The mere fact that you simply "simulate" a dragdrop with the mouse on a progressbar will make it ruin the dragdrop behaviour for other objects.

So simply, make a progressbar or a few and scale them down a bit as it doesnt seem to happen everytime and with every progressbar, after that make some sprites which have dragdrop behaviour. When you run the app. first try to dragdrop the sprites, this should work fine. Then try to dragdrop the progressbar and after that try to dragdrop the sprites again and see if they still work.

I know, but still they shouldnt break behaviours like dragdrop or did i misunderstood what you meant by what you wrote? I just find it weird why a dragdrop on a progressbar, which doesnt have it enable would affect it on other objects. At least to me it seems like a bug. But have you tried testing if you experience the same issue?

Problem Description

The problem im having is when using the dragdrog behaviour with a sprite and having a progressbar pinned to it. This will cause the dragdrop to stop working, lag or make the sprite jump around.

Attach a Capx

https://www.dropbox.com/s/xl7ina3c0olwoum/DragDrop_bug.capx?dl=0

Description of Capx

The capx contain 3 sprites with dragdrop behaviour, each of them have a progressbar pinned to them. The progressbars have been resize to different sizes as it seems to matter, however not sure.

Steps to Reproduce Bug

1. Simply dragging each sprite around seems to work fine. 2. Place the mouse cursor above the middle progressbar and try to drag it. (This will obviously not work as it have no dragdrop behaviour) 3. Now dragging a sprite will now (in most cases) cause the sprite to lag, jump and ruin the dragdrop behaviour. This doesn't seem to happen for all progressbar. Starting by dragging the first progressbar to the far left and then dragging a sprite doesnt seem to cause problems. However it seems that it will happen regardless of whether the progressbar is pinned or not, simply dragging on a progressbar will ruin the dragdrop behaviour for other objects. (If someone can test it, and confirm the issue it would be much appreciated.)

Observed Result

Dragdrop behaviour stops working when you first drag on certain progressbars

Expected Result

That it shouldnt matter whether you dragdrop on a progressbar or not.

Affected Browsers

Tested in NW.js (v0.27.2)

Operating System and Service Pack

Windows 7

Construct 2 Version ID

r250

nimos100

Hey it seem your image doesn't show up, can you please reupload ?

Thank's !!

It should be fixed now.

[quote:2ppeokc9]Was looking over the long list there it seemed like most of them were "nice-to-haves" and not really crucial. Some of them definitely nice to haves, so maybe that's why some things often are left out.

A lot of the things are totally "nice-to-haves" I would never deny that, I would not argue that each suggestion is equally important, which I also made clear above. But this is a personal list of what is important to me when I work with C2. If you notice when looking through the list, there is not a single suggestion aimed at adding further support for additional mobile phones. I don't make anything for mobiles so its of no importance to me. Others might find this very important and will complain and request certain things for these devices. But for me its not important. And in general these suggestions was never meant to be huge ground breaking requests, there are a few in there, but weren't the main idea behind the post in the first place.

There's too much here to be able to comment on everything. I'd recommend going to the new feature request system. Many of these I have never heard of from other users, so they sound more like personal requests - which isn't to say it's not important, it's just with our finite resources we try to focus on what lots of people are asking for. The list is also diluted with some things which are pretty trivial, like I don't see any purpose in adding a "random boolean" action, that's easily done with events.

There is a lot of things in the list and I saw the new feature request system, which looks very cool

But as mentioned above a lot of these are personal requests and are minor things that have annoyed me while working with C2. This is from the original post of why I made these suggestions in the first place:

This post is not aimed towards a single object type, but just in general when using different objects, it gets a bit frustrating that some things are missing which should be there in my opinion. As these are "minor" things you tend to forget them, so ill add them here as I find them and hopefully they will be added at some point.

And you could say that a lot of these things are pretty trivial things that can be solved in other ways, but by reading the above, its exactly what these suggestions are meant to improve. So its not because that the random Boolean can't be done, but it would be nice if it were there, instead of having to solve it using a integer for instant, is it absolutely needed? Of course not. But you could say that about a lot of things when it comes to workflow in C2.

For instant that you can't delete more than one variable from an object at the time, so there is no need to change that since you can still delete them one at the time, but its annoying, why these trivial things that you call them, are not added in C2 as well, I guess I could argue .

No need to make it possible to move several objects in the object list as long as you can move them one at the time.

So its not about whether its ground breaking new features, but small things that can be frustrating when you work. Certain data you cant get in an easy way, but instead have to make some annoying workarounds and so forth.

So if people think the main idea behind a lot of these suggestions are stuff like C3 should support 3D and so forth, its nothing like that, its to improve the workflow and reduce frustrations and annoyances when working, simply to speed up work. And as mentioned, its aimed towards all these small things that you run into but tend to forget not being there or that you never get around to make a suggestion post for.

Your concerns and pessimism are absolutely fair. I do have a pretty simple explanation that may ease your mind though.

Simply put - rewriting the runtime will take a lot of time and effort. Scirra feels like they have created something that is worth money to some people, and hopefully will finance this effort. After all they have been supporting C2 for years and years now. As a company, they also need income to operate and continue working.

So the alternative is wait and release C3 after a new runtime is complete which may come way later, or release what they have now and give users the OPTION of subscribing now or later.

They have been very responsive about engaging the community and clearly documenting their plans and thought process - to an absurd standard even. So I do have high faith that the new runtime will come eventually, as long as they don't shut down.

For any developer, it really comes down to opportunity cost and prioritizing what features to develop given the limited resource of development time - for example, do you want them to work on backgrounds for text that third party addons have already addressed, or do you want them to work on the underlying runtime which would open up a whole new realm of possibilities?

Don't get me wrong, Scirra are in general very good at communication with their users and for that they really get huge thumbs up from me.

However regarding:

[quote:2awcby2u]for example, do you want them to work on backgrounds for text that third party addons have already addressed, or do you want them to work on the underlying runtime which would open up a whole new realm of possibilities?

First of all im not really sure what this new runtime offer in terms of possibilities, if it allows third party developers to add a lot of new cool plugins, that's amazing.

I do however not feel that adding a lot of these "standard" features would be very demanding compared to the extra functionality it would give. And my guess is, for the majority of these missing "Get/Sets" its not because they are especially difficult to make, but simply that they were forgotten when the feature were added, maybe due to some organisation issues of simply not working with a philosophy of if you can Set something you should also be able to get it. But maybe im wrong and these are deliberate decisions that certain data should simply not be possible to receive, I honestly don't know. But if that is case, I would very much disagree to such decisions as its very frustrating for a user to not being able to fully use all features of such functionalities. But instead having to make workarounds that can be extremely time consuming.

While I didn't read through each item in detail, my understanding is that if the runtime is updated, a lot of not possible becomes possible.<br abp="701"><br abp="702">With the caveats that if and when are in the air, and possible doesn't mean probable. Basically I wouldn't hold my breath for anything. Work with what you have or get what you need with alternative software.

[quote:gndn71ny]As stated in a former post im not a fan of working in a browser, however I think C3 does a decent job there, its not as bad as I thought, everything seems a bit more cramped, as you now have the browser navigation panel in the top as well, which causes you to loose a bit of screen space.

This was my biggest concern with C3, but you can solve it beautifully:

Go to the editor page in Chrome > click options button on top right of browser (under close x button) > More Tools > Add to Desktop... > *important!* check "Open as Window".

Now behold: C3 is just like any other app . It's great (no browser UI).

[quote:gndn71ny]Path finding that haven't been fixed from what i can see.

Sounds good and needed with a new runtime if It allows for a lot of new things and I will to some degree accept that, especially when it comes to the path finding and maybe the unload image feature (Which might never been there, new runtime or not ).

However I still think and feel that my concerns are not completely unfounded and I wont bash C3 based on that, I do acknowledge that it is in beta, so obviously its not the final release. But when that is said, I do still feel when going through the program and testing all these issues listed above, that the majority of these things are none runtime related, yet they are no where to be found in the program. That you for instant can't set the background color of a text object, which can be done in C2 with 3 party plugins, cant be done in the current beta of C3, it doesn't require a new runtime.

Also all these missing "Set/Get" functionalities that are missing in C2, are for whatever reason also missing in the current C3 beta and will suddenly appear at some point. And without sounding to pessimistic, Im not really sure I buy that. I find it much more plausible that its because these features are simply not there currently and might never be there. It would make no sense not to have added them when launching the free beta to show off how many more possibilities there is with C3 compared to C2 and to be honest that concerns me a little.

Hi all

Have been very excited about trying out C3 as a lot of others might have been. Since im a long time user of C2 and through the years have experienced my share of problems, bugs and general things I would have liked to see improved in C2, these will make the case for which I will judge C3 against.

Its important to make clear that these are my own personal things that I find crucial in regards to C3 and therefore might not be equally shared amongst others as being relevant or of any importance. But these will be heavily deciding factors for why or why not I would consider buying C3. Now some tests weren't possible as the Demo didn't allow it and I will make that clear.

For some time ago, I started a thread in the general section with ideas and suggestions of how to improve C2, since I have kept adding stuff to this thread as well as other people, I will use these as my check list, I haven't checked anyone else suggestions in that thread, so those I test against are those added by me.

All the original suggestions will be written in black and red will be used for things not working or that ain't possible, and green for those that have been added, orange will be used for those that I weren't able to test as the demo didn't allow it. (This will be a rather long post)

In the end you will find my first impression of C3.

------

Video

Frame rate - Would be great if you could get the frame rate of the video.

Not possible

Number of frames - Self explanatory, but the number of frames in the video.

Not possible

Format - there are other ways to get this, but would be handy if you could just get the format through a variable (Less important).

Not possible

Check vs Frame rate / Playbacktime - For some reason there are no way to check vs this.

Not possible

Play at FPS - Play a video at a certain fps, so you can make slow motion and fast forward.

Not possible

File chooser

Path - There are no way to get the path to a file on local disk. Which is strange since you can get the filename.

Not possible

NWjs

Filename - Opposite the File chooser object you can get the path including the filename, but there are no way as far as I can see that you can get just the filename.

Uncertain as I can't get any of the NWjs functionality to work.

Sprite

Add frame - Add a blank frame to a sprite

Not possible

Delete frame - Delete frame at index

Not possible

Insert frame - Insert frame at index

Not possible

Replace frame - Replace frame at index

Not possible

Fill frame - Fill frame with a certain color at index

Not possible

Erase frame - Erase frame at index

Not possible

Video

Missing expression checks - Even though you can check whether a video object is in a current state such as muted, paused etc as an event. These options seems to not be available as expressions in the video section. If you use a "Compare two variables" it would make sense that these were also there, since its actually possible to check vs them as events. At the moment only 3 are available (Duration, PlaybackTime, Volume) so if (Has ended, Is muted, Is paused, Is playing) were there as well it would be nice.

Not possible

Source - Would be nice if you could get the source. Both filename and path to the file. And if there are no source it could just return -1

Not possible

NWjs & Video & File chooser

Syntax consistency - Depending on what functionality you are using, there seem to be different syntax for how you work with files, which is really annoying to say the least. For instant using the NWjs.Userfolder will return the following: "\User\<Name>\"

However setting the source for a video require the exact path "C:\Movie.webm". Which you can get using the open dialog in NWjs, but that require user interaction. Since there are no functionality that allow you to get just the drive letter, filename, or just the path to a location makes working with files a living hell to say the least, and this changes completely depending on what functionality you use.

It is suggested in the C2 manual that you should use the User folder due to permissions, so it would make sense if all the file interactions also worked in a consisting manner, so you don't have to create a different workaround for each function depending on which one you use. Being able to get the drive letter of where a folder or file exist would be very useful. Basically just divided, into the separate parts that follows what you would expect when working with files on a computer. (Drive, folder, File).

Not possible

NWjs & Debugger

Incompatible - While running your project in debug mode its not possible to open any NWjs dialog boxes. Might be considered a bug, but regardless, it would be nice if the debugger and NWjs were compatible.

Uncertain as Nwjs doesnt seem to work

NWjs

Expanding triggers - Would be very useful I think if the following triggers could be added to NWjs.

On file being copied - Trigger when a file is currently being copied. Would be useful for making copy file progress bars.

Not possible

On file copied - Trigger when a file have completed being copied.

Not possible

On file deleted - Trigger after a file have been erased.

Not possible

On file moved - Trigger after a file have been moved.

Not possible

On file being moved - Same as "On file being copied" would be handy for progress bars.

Not possible

C2 General

On exit program - Would trigger when your program is closing. Could be used for saving progress etc.

Not possible

NWjs

Delete folder - For some reason its possible to delete a file, but not a folder. So would be nice If that was possible as well.

Not possible

NWjs

On folder dialog Cancel - Inverse version of "On folder dialog OK"

Working, but was already fixed in C2

On dialog Cancel - Inverse version of "On dialog OK"

Working, but was already fixed in C2

On save dialog Cancel - Inverse version of "On save dialog OK"

Working, but was already fixed in C2

General

Get layer background color - Possible to get the RGB values of a layer"

In general it would be nice if all things in C2 was made in a way, that if you can set a value then you can also get it.

Not possible and it doesnt seem that C3 improves any when it comes to setting or getting variables, its pretty much what you have in C2 some you can and some for whatever reason is unavailable.

Path finding

Change nodes - Possible to change individual node locations when a path have been found, so its possible to manipulate the path before an object moves along it.

Not possible

Audio

Tag property - Possibility to get the tag of an audio file when its playing, ended etc.

Not possible

Any tag ended - Trigger when any audio tag is ended.

Not possible

Form controls

Tab index - Using form controls (Textbox, Buttons, etc) there seem to be no way to set/change the tab index to navigate them. Would be nice if you could specify it in the properties panel.

Not possible

Text objects

Auto size - Would be useful if you could turn on auto size for text objects. So they would automatically adjust to the size of the written text.

Not possible

Change text align during runtime - For some reason its not possible to change this, even though it can be done in design mode.

Not possible

Use background - Make it possible to add a background to an text object and change it color and opacity. There is a plugin capable of doing that, so thought it would be a good idea to just add to the default one.

Not possible

General

Enable/Disable Objects - Should apply to all objects.

An option to enable or disable any object in the properties window and through an action. Object that are disabled would appear as if they were destroyed, without actually being so. The benefit of having an option like this would be to quickly disable and enable objects that you would otherwise have to move off screen. Would be useful for stuff like HUDs or sprites where visibility for instant have been turned off, but are still reacting to mouse inputs like "Mouse over effects" and so forth. And so you would avoid having to reposition objects all the time etc.

Not possible

DragDrop behaviour

Axes angle constrain

The option to choose to constrain the axes to the angle of the object using the dragdrop behaviour. At the moment if you have a sprite and you apply the dragdrop behaviour and want it to only allow dragging along the length of the sprite you can change it to horizontal. But the moment you change the angle of the sprite the constrain doesn't follow along. So being able to change the axes constrain so it always follow the angle of the parent would be very useful.

Not possible

Form objects

On focus - When a form control gets focus this will trigger true.

Not possible

On focus lost - When a form control looses focus this will trigger true.

Not possible

Has focus - When a form control has focus this will trigger true.

Not possible

This is inspired by a 3rd party plugin capable of doing this, so thought it would be good to add so its part of the standard form controls, since its possible.

Collision object

This is not a functionality in C2, so its more of a suggestion.

The idea is to be able to add an empty invisible collision object instead of having to use sprites, which is the normal way of doing it now as far as I know. But doing this uses a lot of memory for no good reason. So what this object would do, is to make it possible to specify an area kind of the same way you would set up a collision mesh on a sprite, but this would be done in the layout, instead of in the graphic editor. The object would work in the same way as a sprite would when it comes to collision detection, but wouldn't be able to hold any form of graphical information, thereby making it possible to use more complicated collision solutions that wouldn't eat up memory for storing empty sprites.

So basically the object simply keep track of a collision mesh, maybe with the ability to manipulate it at runtime and whatever other functionality could be found useful.

Not possible

Rotate objects

The general idea is to extend the rotation of objects. As it is now they rotate around there origin. So adding two new options could be very useful, I think.

Rotate around imagepoint - Instead of rotating around origin, you can choose whatever imagepoint on the object you like and it will rotate around that point. (Could maybe just replace the normal one, as it does exactly the same. And the origin could just be set by default.)

Not possible

Rotate around position - Same as above except you can specify a X,Y position and the rotation will happen around this point.

Not possible

Screen resolution

A way to get more control over the resolution when the "Fullscreen in browser" settings doesn't cut it.

Get screen resolution - The ability to get the screen resolution set in windows/OSX.

Not possible

Set screen resolution - The functionality to change screen resolution to suit what you need. Like you can in pretty much any modern computer game today.

Not possible

On screen resolution change fail - Will trigger if the change fails.

Not possible

On screen resolution change done - Opposite of when it fails, will trigger when the resolution have changed successful.

Not possible

Form controls

Disable tab - Will exclude a form control from appearing in the tab index list so it can't get focus.

Not possible

Delete object warning

This is an improvement or extension to the warning that you get when you want to remove an object from your project. Normally when you do this you get a warning telling you that it will remove all actions, conditions and expressions related to the object. Which is really nice, however when you are working on big projects and you might not recall all the places where this object might have been used etc. The removal can cause some serious problems if you are not careful and normally, at least for me I keep an eye on the number of event in the status bar to see if it changes when I remove an object and if it does I can undo and then check through the project to make sure that it wont cause problems before I do it.

What would be nice and since C2 already have to go through the project deleting the actions, conditions etc. when you press yes. Would be if it in the warning message wrote how many actions, conditions, expressions it would delete, so you would know what impact it would have.

I can't imagine that being a difficult thing as it already "scan" the program when removing it. So it could just "scan" and count the elements without deleting them before pressing yes.

Just a small extension that I think would be handy.

Not possible

Support for multiple aspect ratios

The fullscreen in browser settings are really nice when you want your game to support several resolutions. However a huge limitation occur when you want to support multiple aspect ratios.

For instant the most common used ratios are 16:9 (1680x1050) and 16:10 (1920x1080), however there are no way, as far as I can see that you can make your program support both of these correctly when using "Request fullscreen in browser". And regardless of which method you use (Scale outer or letterbox scale) you are left with black bars either in the top and bottom using letterbox scale or only bottom using Scale outer. Since C2 can already support a single aspect ratio, extending this to be able to support multiple ratios would be a huge improvement I think.

As an extension to the former suggestions regarding "Screen resolution", it would be nice with the following.

Get system aspect ratio - Would return the aspect ratio of the system.

Not possible

Extend the fullscreen browser methods - The ability to tell C2 to target specific aspect ratios, so instead of just "letterbox scale" it would be "letterbox scale - (16:10)", "letterbox scale - (16:9)" and so forth. These methods should also be available in the browser object when requesting fullscreen mode.

Not possible

Bound to layout behaviour update

The bound to layout is a quick way to keep objects within boundary limits. However this behaviour runs into problems if objects that fits the layout are suddenly rotated. In this case the object will start flickering as it tries to stay within the layout. So a minor update to this behaviour could fix it.

Not possible

Fit to layout - When this is turned on, an object will automatically scale to fit the layout in case it becomes to big.

Not possible

Timer behaviour

Adding a few extensions to this behaviour would make it more complete and easier to work with.

Stop all - Would allow you to stop all timers running for an object, regardless of the tag.

Not possible

Pause timer - Would allow you to pause a timer using its tag.

Not possible

Pause all - Same as above but would just pause all timers regardless of tag.

Not possible

Resume timer - Would allow you to resume a paused timer using its tag.

Not possible

Resume all - Would resume all paused timers for the object.

Not possible

Textbox form control

Some of the 3rd party textbox plugins have a select all functionality, so thought it would be handy to integrate that as a standard functionality to the textbox object together with a deselect all.

Select all - Select all text in a textbox

Not possible

Deselect all - Deselect all text in a textbox

Not possible

NW.js dialog tags

When using NW.js "Show folder dialog", "Show open dialog", "Show save dialog" and afterwards the triggers "On folder dialog OK", "On open dialog OK", "On save dialog OK" it would be very useful I think if it were possible to set a tag for these as you can with audio.

As it is now there are no build in functionality to specify which "On ... Ok" will trigger if you use several of them. You can and have to do this using manual variables, but would be handy to simply make it a standard option in the NW.js functionality, I think.

Not possible

Quick navigating functions

This was an old suggestion made, that I still believe would be one of the most useful features in C2. So ill add it here for that reason.

The ability to Ctrl + Click (or whatever hotkey is useful) a function call to quickly jump to the where the function is declared.

Not possible

Reset object variables

The ability to reset an object variables to their default values. Would be very helpful when working with for instant sprites that contain a lot of temporary variables for tracking things, state machines and so forth. Rather than having to do this per variable.

Reset variables - Will reset all variables added to an object to the value they were by default.

Not possible

Smart family

This is not a feature in C2 so should be considered a feature suggestion maybe for C3.

This is basically a normal family, except that it will allow any combination of object to be part of it. An issue with the normal families is if you have features in your program that need to be the same for varies types of object then there are no other way than to duplicate code for each type.

How the smart family would work is to check for similarities between the objects added to the family and only those which are shared by family members would be available. Since for instant UID is available for all objects, it could be used, whereas something like "Angle" might not be, and therefore it won't be available.

Uncertain as families are not available in the free edition

Delete several local object variables at the same time

Currently you are only able to delete one variable added to an object at the time, would be handy if you could select several of them and delete them in one go.

Not possible

Move several objects

I know this have been suggested a lot of times before, but ill add here as well.

The ability to select and move several objects at once in the projects window.

Uncertain as you cant create folders in the free edition. But you can select several objects in the list so assume it works.

Automatic update of function calls

Currently when you have defined a function and later on might realize that you should have called it something else, you have to manually go through the code to find all the places where you call this function from. Would be useful if C2 could automatically do this as it does with any other object when you change their name.

Not possible

Auto/manual sorting of the list object

Would be nice if there were a property for making a list auto sort or to manually do it. This could be either a property you could turn on for the list it self in the properties panel or as an action that you could call in your code.

Action:

Sort list - Would sort the list using a basic sorting method. So "a" would be before "b" and numbers would be before letters etc. Could be expanded so you would be able to use tokenat to customize which parameter of an entry text should be used for the sorting method.

Not possible

Is between value picking

This is just a minor thing, but it would make sense if the "system -> Is between values" were also available as a picking condition for objects.

Since you can pick objects based on variable values, it would be nice if this were also a possibility without having to use other picking conditions.

Not possible

Random setting of Boolean

Yet a minor addition that would be handy I think. At the moment you have the ability to Set a Boolean variable to either True, False. But in cases where it should be random, its easier to use an integer. So adding a new action could solve this.

Action:

Random Boolean - Would set the Boolean to either true or false.

Not possible

Tilemap Set/Get rotation state

You can set the desired rotation state of tiles in a tilemap. However there is from what I can see no way to receive this rotation.

This causes problems when you switch one tile with another and want to keep the rotation of the original tile.

So a nice addition would be to get maybe a numeric value indicating which rotational state a tile is in. Could be a 0 based index, that way it would make it easier to set a new state for a tile during runtime.

Action:

Get tile state - Returns an index telling which state is currently active.

Not possible

Set tile state - The ability to pass an index to a tile at a given tile position and tell it to change state.

Not possible

On variable changed trigger

This is just a suggestion of a feature that I think might be pretty handy.

At the moment you can check a variable against varies values by manually adding these. However think it would be very useful if you had the possibility to create an event condition that would trigger whenever a specific variable changes.

It could work with both global and object variables so when you add the trigger, you get a drop down list where you can select which variable you want the event to monitor for changes.

Conditions:

On variable changed - Trigger whenever the chosen variable changes value.

Not possible

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

Last and not least.

Path finding

This have not been working correctly for a long time in C2, so I was very interested in testing that out to see if it have been fixed. And to my big surprise it seems to have been made exactly the same way as in C2 and still only uses one collision map, which means its no better than before.

Is this really the case, you redesign C2 to make it better, well aware that this feature is not working correctly, but you implement it the same way regardless, would it be possible to get an explanation why that is the case or whether you plan on fixing it or what exactly is going on, because it makes very little sense, at least to me?

My first impression of C3

Having tried C3 and testing all these things out and I know its just a Beta still, I have to say that at least for me personally, I find no reason to why I would like to buy it over C2 in its current state.

As stated in a former post im not a fan of working in a browser, however I think C3 does a decent job there, its not as bad as I thought, everything seems a bit more cramped, as you now have the browser navigation panel in the top as well, which causes you to loose a bit of screen space. However that is not crucial in itself, if that were the only problem. However quickly having scrolled through my list above, none of the issues seems to have been improved and again I want to underline that this is my personal view!! and even to me there is a difference in how important I find the above suggestions.

So to me some of the crucial ones:

Path finding that haven't been fixed from what i can see

This alone is enough for me to completely write off C3 as a whole, there are no excuse that i can come up with that would justify why I would get C3 anyway. This is a known problem by Scirra and have been a long time, even before C3 was announced as far as I remember.

If you can Set something you should be able to get it

Again absolutely crucial, its so annoying that you can manipulate and set some variables and state of things, but no way to test or receive them again. This as well is enough for me to not wanting to invest in C3.

Delete image from memory

I haven't actually added this above, but the lack of functionality to manually delete an image from memory is the last crucial thing for me. That this possibility shouldn't be there to save on bandwidth is just not a good enough excuse. Its a fair point, but this should be decided by the developer of the game and not Scirra in my opinion.

So adding everything together, my first impression of C3 is that its C2 in a browser, very few features that I personally would consider important seem to have been added and when i say important, I refer to workflow, C3 doesn't seem to offer anything in regards of reducing the amount of workarounds that you have to do in C2 to make certain things work. The lack of very basic functionality, like changing alignment of text during runtime are still missing which are features that have been available through 3 party plugins for a very long time, why isn't stuff like that added by default in C3, all these things put together it just make me wonder, what is the purpose of C3, what does it actually offer? Maybe these things will be added in time. Who knows, but at this very moment C3 have nothing to offer me over C2.

Now is this a rant, because they seem to have added none of my suggestions in that post?

I still love C2 and working with it and if C3 had been C2 I would have loved that, except that it ran in a browser. But I can still make stuff in C2 and good stuff, so the above might as well be seen as me having a go at C2, my problem is that the difference between these two in terms of features that are important to me, are close to none existing, which will cause me to rant a bit. Not because they didn't add any of my suggestions, but because as any long time user of this forum knows there have been a trillion suggestions to C2, some more advanced than others, but yet even the most basic suggestions from above, which might not even have been suggested by me in the first place, since I just decided to add all mine in a single post as I noticed them, seems to have made it into C3. I would understand huge and really demanding features, that it would be a lot of work adding them all, but a lot of those above isn't, they are basic functionalities that you would expect in any such program in my opinion. When you work with text and you can change alignment during development and 3 party plugin developers have already done it so you can change it during runtime, then this is what you would expect to be the new standard for a new version of construct, but it just isn't. Everything under the hood seem to be C2 in a browser and that is what disappoint me.

I think it's a bit too soon to post a topic like this, while Scirra is still revealing new and new C3 features. But just to mention some of the features that makes it worth to consider C3:

- Scirra's own cloud APK exporter.

- New features in the editor to make the development process faster and more comfortable.

- Completely redesignable and translatable editor.

- New plugin manager with the future promise of editor plugins.

- Built in web-font support.

- Etc..

The browser based editor can have several advantages, and I don't think that Scirra would go for it if they weren't absolutely sure about that it will work great.

The subscription is one of the main topics in the C3 forum section, but to cut it short: No one-off payments. And I think it's better that way (but that's just my opinion).

Don't get me wrong, they have already presented new features in C3 from what I can see and im sure more will follow. What im interested in is why a browser based editor is better than one that is not. What benefits does it give, I don't think its to early to ask such question, because it must have been something they decided or saw some benefits/advances in doing quite early, when they started to develop C3. Keep in mind as I stated in my OP post im not a fan of browser based apps in general, so im just curious about potential benefits of such solution compared to a "normal" program.

Was reading the C3 announcement and congratulation with it almost being done

Anyway as I was reading it you write the following in the beginning:

[quote:dwgza59q]We aim to redefine what is possible with the web. So we built the Construct 3 editor itself for the browser.

Construct 3 works in Chrome on Windows, Mac, Linux, and Chrome OS. There's nothing to install — it's as simple as visiting a website. There are too many new features to list here,

Personally Im not a fan of working in browsers and only use NWjs in C2 for previewing, so im probably not as excited as you are or as many other users might be. I like working in programs especially designed for a specific purpose, which might very well be the case for C3, but for me, it just feels right with a program you start and do your work in etc. Anyway what I would like to know, if you can elaborate, what are the benefits of what you write above in the announcement for me(us) as user? You give the impression that its something new and exciting, but you don't really share what we as users should be so excited about, for now having to use a browser rather than an installed application?

I read some other users concerned with doing this as 3rd party problems could affect things. In C2 we have and still experience problems with our creations due to browser issues that was outside your domain to fix. But having it affect our creations is one thing. But wont you run the risk of it suddenly adding problems or limitations which could affect the editor it self?

As many others (I think) Im not fan of the subscription payment method, but do understand why you would prefer doing it that way, but again for me I like programs and once you have bought them they are yours, subscriptions always feels like im leasing something without owning it and I don't like that in general. So was wondering if you plan on making an one time buy offline version? (I know C3 works offline) but a version that doesn't work online in the browser but one that works simply like C2 does now and where you pay a one time price?

Anyway looking forward to trying the beta.