There's a bug with Multiple Touch object where "A new touch on object has started" gives a "double click" on a string, I didn't tested on my device but it happens on Windows runtime.
Printable View
There's a bug with Multiple Touch object where "A new touch on object has started" gives a "double click" on a string, I didn't tested on my device but it happens on Windows runtime.
thanks for the example i will start check tomorrowQuote:
Originally Posted by NaitorStudios
post a small example for me to checkQuote:
Originally Posted by NaitorStudios
2048m means ?Quote:
Originally Posted by Fernando
and from ?
Does child events has been changed? I have a lot of issues with child events not triggering correctly in this version, for example, this event trigger only one of both actions:
Code:* NOT mask_spartan-foe animation Disappearing is playing
+ Enemy life of mask_spartan-foe <= 0
+ Pick one of mask_spartan-foe
mask_spartan-foe : Set Flag 10 off
mask_spartan-foe : Change animation sequence to Disappearing
+ Killed by of mask_spartan-foe = 0
counters_all : Add 1 to medals_multikills
counters_all : Add 1 to medals_killingspree
counters_all : Add 1 to ct_frags
+ val_gamemode <> 16
OR
+ val_gamemode = 16
+ zombie round of counters_all = 0
+ set_team = 2
OR
+ val_gamemode = 16
+ zombie round of counters_all = 2
+ set_team = 2
OR
+ val_gamemode = 16
+ zombie round of counters_all = 1
+ set_team = 1
counters_all : Add 10 to ct_xp
check-xp : Subtract 10 from Counter
Seems not to be related to this build as reverting to 293.5 or 293.1 doesn't change anything. (or the MFA is broken?)
But if i let the child event with its OR it trigger only the last action of the list. If i remove the OR, it trigger both actions.
I've fpound the issue, here is a MFA that reproduce it:
Attachment 30395
Pressing SPACE BAR should add 1 to counter 2 and add 1 to Alterable Value O of Active 2.
The add 1 to Alterable Value O of Active 2 action is ignored.
My bad, just figured out it was the normal behavior of OR when adding an object in the condition. As it scope for this particular object and ignore it if is not scoped in one of the OR. (case of my example)
So i should use "OR (logical)" here to make the example to work.
PS: why is the 'edit post' button still missing?
Another bug on mobile: Set position of Layer causes touch/mouse input to be misaligned. I'm not sure if this also happens with set Scrolling.
so far you should understand that values are megabyte of memory, you can get more information in gradle documentation https://docs.gradle.org/current/user...ing_jvm_memoryQuote:
Originally Posted by pradeep
MFA please to compareQuote:
Originally Posted by NaitorStudios
Sure, you can clearly see how the actual object moves when you set the layer's position but the actual hitbox stays in the original position.Quote:
Originally Posted by Fernando
This doesn't happen on Windows, HTML5, I couldn't test others... Edit: But Shinbaxter verified on iOS and it seems like it doesn't happen there as well.
MFA
https://www.dropbox.com/s/dbn9ko3hx6...touch.mfa?dl=0
APK
https://www.dropbox.com/s/1jfckwqxc7...touch.apk?dl=0
Here's a MFA showing this other bug:Quote:
Originally Posted by NaitorStudios
Attachment 30397
Number error，当重复按下鼠标右键，计数器开始增加，当移动鼠标到窗口外的桌面悉放鼠标右键，计数器有时并不 立即停止计数，而是又继续增加了一会数字才停止，这让我很是烦脑
Number error, when the right mouse button is pressed repeatedly, the counter starts to increase. When the right mouse button is placed on the desktop outside the window, the counter sometimes does not stop counting immediately, but continues to increase the number for a while, which annoys me very muchAttachment 30399
I set 2048m but not workQuote:
Originally Posted by Fernando
delete the .gradle folder
download all needed framework, libs, and plugins
but problem not resolve.
After the frame rate fix for sub apps other problems started to appear, here is some of which I in countered so far:
- On one of my projects, sub app crashes the whole project
- Timer is out of sync when the main app is on a temp pause (eg. When resizing window or minimizing it)
- when changing the FPS on runtime the whole app freezes (even the debugger)
- when changing sub app size it crash the project randomly (very rare)
does the error still Java Heap Space?, if so what are you building there my only idea is that the .aab is bigger than 150 MB and therefore is declared invalid, which is the size for the apk?Quote:
Originally Posted by pradeep
Tested with latest version of multitouch, and it works OK, indeed there is a problem for windowsQuote:
Originally Posted by NaitorStudios
@NaitorStudios, a fix was prepared for both multitouch and android runtime, but I will like you to test and review in real projects, contact me on discord.Quote:
Originally Posted by NaitorStudios
My app is educational. In this app I use png {of questions (MCQ's) of different units and topic) , List if thus png.}add in Binary Data. In this total files 2690,89, total sizeQuote:
Originally Posted by Fernando
mfa size 1.85 MB (Total Frame 5)
1. Old build (Updated April 30, 2021) uplode on playstore : https://play.google.com/store/apps/d...PatelMECHANICS
2. Size of .aab = 106 MB (Again Create without any change 05 August 2021) (apk size of this is 112 MB)
3. In this total files 2690,89 , total size 107 MB
4. same mfa with file use to build abb using 293.5 (no any change,error still Java Heap Space).
5. using 293.5 :
But some file and folders remove,
In this total files 1707,59 folders = total size 68.8 MB.
Then build abb using 293.5, build success. size of .aab = 49.0 MB.
When add some file (total size of file greater then 68.8 MB like 75 MB) error still Java Heap Space.
Problem Resolve.
Thanks Fernando.
This probably isn't worth your while, since there probably aren't many people other than me who would particularly want this. But is there any chance you could add a List View mode to the Layers toolbar (ie. no thumbnail - just layer names and numbers)? It would save some horizontal real estate (allowing us to reduce the width by ~100px while still seeing full layer names). It would also allow those of us who use dozens of layers to see them all at once.
That isn't a bad idea at all, it will definitely make it easier for managing lots of layers at once.Quote:
Originally Posted by Volnaiskra
It seems like Easing object behaves differently on Windows, to be more specific, it's bugged... Since it's open source, could we get a fix?
I believe the issue must be something related to something rounding the math, cause depending on the movement it can get closer or further away...
To add:Quote:
Originally Posted by Linky
- keyboard and mouse: "On any key press" condition doesn't work with sub-apps, I had to use Control X object
I'm pretty sure I've encountered this too.Quote:
Originally Posted by NaitorStudios
Another bug in Android:
Sometimes minimizing the app and opening it again causes transparent parts of objects to have a "additive" or "overlay" effect, like the blending mode from Photoshop, similar to some shaders of the same name...
It's just weird, sometimes it can get stuck in this state, it's just a visual glitch, but it can really mess with games that relies a lot on alpha channel...
I'm still on 293.1, but I have a couple of strange behaviors to report & I did not see them listed as fixed in the recent changelogs. Forgive me if they have already been fixed.
1) the expression function random() doesn't seem to be able to generate numbers outside the range 0-65535.
RRandom() is similarly affected. It doesn't generate random numbers outside the range 0-32767
Also, it appears that when referencing a Global String by index, it's using a 1-based index, instead of zero-based like with Alterable Strings of Active Objects. Could anyone confirm this one?
MFA demonstrating above issues. I am on 293.1 though.
This is intentional, if you see the Values properties of the app you'll see that it shows a numbers starting with 1.Quote:
Originally Posted by YesNoYes
Alterable Values/Strings are in fact 0-based but it does show in the corner of the calculator that it works this way. Same goes for globals.
That's an old 16-bit limitation, int can go from 0-65535 if not including negatives, with negatives it just splits the value in two.Quote:
Originally Posted by YesNoYes
I'm not sure if they can change this at this point, must be tied to the old code from previous softwares like K&P.
Either way, the built-in randomization isn't good enough if you're doing something more advanced, it will repeat the pattern pretty clearly...
You can try MT Random object, Random Pool, Randomizer object or String Generator.
Uppernate also did a great custom random number generator, you can find on the community forum I believe.
It's intentional? I must say it seems a bit bizarre that 2 different base-indexes would be used for 2 built-in functions that are virtually identical. I did notice that Global Values started at 1 in the properties, but I didn't think it related to the actual index. I'm sure there's valid reasons why it's been implemented in this way, but it kind of adds an extra layer of confusion & potentially adds more code to routines which address global variables just trying to make the math work. Not a big issue for me though, I'll just use Alterable Strings from Active Objects instead.Quote:
Originally Posted by NaitorStudios
As for the random number issue, that's a bit of a gotcha imo. I don't personally need random numbers to be either that precise or even that random for any particular project I'm working on, but I was trying to generate a series of random numbers across the full range of word sizes for a test I was doing regarding masking out select bits using "0bxxxx xxxx" formatted strings stored in globals as a bitmask (which is how I discovered the indexing idiosyncrasy), and I wondered why it was returning 0 when masking into the 3rd & 4th byte. Again, not a massive problem, but something I had expected to be able to do in a 32-bit application running on a 64-bit computer XD
There's a bug with the new "Play Sample (all parameters)" action, that overrides a channel volume instead of just changing the volume of the sample that it's playing.
Not really, if you're using loopindex or random(), you just need to add 1.Quote:
Originally Posted by YesNoYes
I guess it is this way cause Global Values was introduced later on, we always have to assume it's like this because of a old limitation, either way, if they change it now it will break everybody's projects, so this ain't gonna happen, perhaps on Fusion 3 ;)
I see so much potential on the Tools menu + toolbar customization... Definitely a underrated (almost hidden) feature!
But it has some major problems:
- You can only have 10 tools in the tools menu (you can't even choose which ones if you have more than 10 EXE on the tools folder!)
- The icon editor for the toolbar customization is ancient, it has only 16 colors... Would be awesome if it was updated to the new icon editor!
- Command line arguments can only be set by tool and not by icon on the toolbar... Would be awesome if we could also have arguments for each icon placed on the toolbar, so a single tool could do multiple stuff.
The last suggestion alone could bypass the 10 tools limit and make it so much more useful!
Since you can actually have multiple different icons pointing to the same tool, being able to change the command line arguments individually would be amazing.
I know it's not a much talked feature, but I think if it gets a nice update (even if it was 2.5+ exclusive) and get a nice video showing what you can do with it, it would become a lot popular!
A nice little addition to fusion would be the ability to resize objects in the frame editor while snapping the resize handles to the grid. If I remember correctly Klik & Play could do that so why not fusion, it can be a new optional check box in grid properties :)
I've found an issue, i don't remember if it has been already reported, the black color (0,0,0) of quick backdrop object is slightly red (0,0,8).
Not a bug, the black 0,0,0 is used for transparency by default, so it's normal that anything visible will not be it.Quote:
Originally Posted by Dobermann
You can change it by double clicking the transparency box on the image editor, but it can only be changed per object (and per animation I believe), changing this to something like 255,0,255 you are able to use 0,0,0.
I've getting an error when exporting to Android with one extension that seems to be related to the changes on the beta... (the extension is PushOneSignal from ClickStore)
> Task :app:processDebugMainManifest FAILED
C:\Users\NaitorStudios\Documents\Clickteam Fusion Temp\And8C11.tmp\app\src\main\AndroidManifest.xml: 17:3-51 Error:
Attribute application@requestLegacyExternalStorage at AndroidManifest.xml:17:3-51 requires a placeholder substitution but no value for <Legacy> is provided.
C:\Users\NaitorStudios\Documents\Clickteam Fusion Temp\And8C11.tmp\app\src\main\AndroidManifest.xml Error:
Validation failed, exiting
Seems like it either changes something on the AndroidManifest or it request something that isn't available on Fusion anymore... Can this be fixed?
This extension is currently the only notification extension available that still works, I wish Fernando had finished the promised notification extension (found about it on an really old article)
Edit: Seems like the extension did added the placeholders to PrepareAndroidBuild, so something something seems wrong with Android runtime...