Build 295.10 - Release Version

Quote from Linky

It's in CEvents.m a find all should reveal it

LOL, thanks for the info, global find could make this line in my Xcode. if not because of you or Yves posting where this code is I wouldn't find it.

Checking it now.

============

Will continue with some test tomorrow but try this at your side.

int nString=0;
NSUInteger strLength = [pString length];

if (strLength==0)
{
...
return;
}
while([pString characterAtIndex:nString]==32 && nString < strLength-1)
nString++;

testing with this inputs seems fine

funcVal(@"");
funcVal(@".");
funcVal(@"0.");
funcVal(@"0");
funcVal(@".0");
funcVal(@"1234,456.0");
funcVal(@"78976.0");
funcVal(@"1e-06");
funcVal(@"23673644.989e-05");
funcVal(@"1954637.989e05");
funcVal(@" ");
funcVal(@" ");
funcVal(@" 56");
funcVal(@" 56.678 ");
also tested for nil input

Quote from Fernando

LOL, thanks for the info, global find could make this line in my Xcode. if not because of you or Yves posting where this code is I wouldn't find it.

Checking it now.

============

Will continue with some test tomorrow but try this at your side.

int nString=0;
NSUInteger strLength = [pString length];

if (strLength==0)
{
...
return;
}
while([pString characterAtIndex:nString]==32 && nString < strLength-1)
nString++;

testing with this inputs seems fine

funcVal(@"");
funcVal(@".");
funcVal(@"0.");
funcVal(@"0");
funcVal(@".0");
funcVal(@"1234,456.0");
funcVal(@"78976.0");
funcVal(@"1e-06");
funcVal(@"23673644.989e-05");
funcVal(@"1954637.989e05");
funcVal(@" ");
funcVal(@" ");
funcVal(@" 56");
funcVal(@" 56.678 ");
also tested for nil input

Display More

Thanks, will try to check when I can.

Unrelated, but I think I found the problem related to saving\loading array files (NOT the embeded binary data one, it was fixed somehow in 295.5)
Basically if I don't add a forward slash here to the path, it doesn't work anymore, but it used to work in earlier versions:
Please login to see this picture.
This may be related to paths in general regardless of the object, but didn't test that so i'm not 100% sure

Quote from Linky

Minimum android version can't go older than 5.0 or you will get errors, make sure target is 13, newest API and minimum 5.0

Minimum version already 5.0, targer is 13 and API is 33

(The error mentioned API 31, but I already checked with API 33 as well. Still same error)

Quote from Linky

Thanks, will try to check when I can.

Unrelated, but I think I found the problem related to saving\loading array files (NOT the embeded binary data one, it was fixed somehow in 295.5)
Basically if I don't add a forward slash here to the path, it doesn't work anymore, but it used to work in earlier versions:
Please login to see this picture.
This may be related to paths in general regardless of the object, but didn't test that so i'm not 100% sure

The event that you mention as 1698 that used to work, it was never correct, since to point correctly to datastorage folder you will need a file separator "/" as you do in the event 1700. It is a well-known comment in this forum regarding datastorage or publicscopedstorage, used now, and in previous version externalstorage. They always need a "/"

one example of this is here Please login to see this link.

Quote from Linky

Basically if I don't add a forward slash here to the path, it doesn't work anymore, but it used to work in earlier versions.

Please login to see this picture.

The forward slash was definitly required before. There should be everything ok here and nothing has probably changed.
Are you imagining things, Linky?
Perhaps you wrote in the past "/directorypath/" + "File.ext" instead of "/directorypath" + "/File.ext"
- what you have in your image at the top ("/folder" + "File.ext") would be not correct.

Quote from Fernando

The event that you mention as 1698 that used to work, it was never correct, since to point correctly to datastorage folder you will need a file separator "/" as you do in the event 1700. It is a well-known comment in this forum regarding datastorage or publicscopedstorage, used now, and in previous version externalstorage. They always need a "/"

one example of this is here Please login to see this link.

Quote from Wodjanoi

You are probably wrong with this assumption. There should be everything ok here and the forward slash was definitly required before.
Are you imagining things, Linky? Eventually you simply you wrote "/folder/" + "Track.ogg"? This would also work I guess, but not "/folder" + "Track.ogg"

I know event 1698 is incorrect, i'm very aware of this, but i'm not imagining things and it used to work in 294.14
Thing is, one of the games I worked on that is now in production, and I made this silly mistake where I genuiniley forgot to add a slash there, sometimes i'm too used to do AppPath$ + "xxx" without slashes, that I did the same thing with DataStorage for android, and because it awlays worked I never noticed such problem until it was late and the game was already in production.

So if this can at least behave like how it was so I can at least be able to update the app to convert old saves from the incorrect path to the correct one would be very appreciated, or users will lose there progress...

Quote from Linky

I know event 1698 is incorrect, i'm very aware of this, but i'm not imagining things and it used to work in 294.14
Thing is, one of the games I worked on that is now in production, and I made this silly mistake where I genuiniley forgot to add a slash there, sometimes i'm too used to do AppPath$ + "xxx" without slashes, that I did the same thing with DataStorage for android,
and because it awlays worked I never noticed such problem until it was late and the game was already in production.

So if this can at least behave like how it was so I can at least be able to update the app to convert old saves from the incorrect path to the correct one would be very appreciated, or users will lose there progress...

I see. Sorry, I was editing my post one more time while you was posting your last answer here:
So, you are sure that you didn't wrote "/directorypath/" + "File.ext" before instead of "/directorypath" + "File.ext"?

Fernando, can you give me a hint please? What does the "MAX_SIMUL_SOUND" property do in Android runtime?

Quote from Wodjanoi

I see. Sorry, I was editing my post one more time while you was posting your last answer here:
Are you sure that you didn't wrote "/directorypath/" + "File.ext" instead of "/directorypath" + "File.ext"?

Yeah I unfortunately did the later as can be seen here in line 667 (same thing with loading):
Please login to see this picture.

Quote from GKProduction

What does the "MAX_SIMUL_SOUND" property do in Android runtime?

Unknown property is a common error. I assume you have already checked if any outdated extensions are possibly causing this?
Wait for Fernando's return in any case, but perhaps you could track down by yourself what's causing this exactly and eventually prepare an example file.

Well, I was able to build the app in apk and aab! BUT, when I try to build into aab with expansion, there is an error that I wrote about in Please login to see this link..

Also I made mfa for test this build error. It's impossible to compile this mfa to aab with expansion:
Please login to see this attachment.

Quote from GKProduction

Well, I was able to build the app in apk and aab! BUT, when I try to build into aab with expansion, there is an error that I wrote about in Please login to see this link..

Also I made mfa for test this build error. It's impossible to compile this mfa to aab with expansion:
Please login to see this attachment.

Thanks i will check ASAP
==================

Fixed for next beta, GKProduction i sent a PM for you to try.

Quote from Linky

I know event 1698 is incorrect, i'm very aware of this, but i'm not imagining things and it used to work in 294.14
Thing is, one of the games I worked on that is now in production, and I made this silly mistake where I genuiniley forgot to add a slash there, sometimes i'm too used to do AppPath$ + "xxx" without slashes, that I did the same thing with DataStorage for android, and because it awlays worked I never noticed such problem until it was late and the game was already in production.

So if this can at least behave like how it was so I can at least be able to update the app to convert old saves from the incorrect path to the correct one would be very appreciated, or users will lose there progress...

Hey, just checking, no rush of course but are there plans to return it to work like how it was in 294.14? this "problem" have been worrying me a lot recently XD

Quote from Linky

Hey, just checking, no rush of course but are there plans to return it to work like how it was in 294.14? this "problem" have been worrying me a lot recently XD

this won't be changed as before due to all scoped storage changes now used by android 10+, proper way is as mentioned in previous post.

Hello, will effects for layers be available? My game uses the "Layer object", thanks to it I hide all objects on the layer through the "alpha-blending coefficient" effect, but due to the fact that this is not available on Android, I have to redo these events into separate objects and classifiers.

Quote from Fernando

this won't be changed as before due to all scoped storage changes now used by android 10+, proper way is as mentioned in previous post.

Sure, I may have found a workaround to convert the file, it seems to actually still read fine, but not save, I will share here in case someone falls into the same problem:
Basically do a one time conversion thing where it would load the file from the incorrect location, then save it to the correct location this time, but only once, after that you let the game load from the correct location, this will avoid users data being lost

On other notes.. it may be useful to mention more of the smaller changes in the changelog, for example billing library is now using version 5.1 and some Admob behavior seems to have been corrected, may be useful to add to the changelog, especially admob changes because I was actually using some workarounds to get past buggy behavior in 294.14, some of them were actually very useful, but as admob is now corrected, those workarounds don't work anymore and actually start breaking the whole ad system (if any of this makes sense)

What I mean by "some of them were actually very useful" is because we can't change the ads on runtime I was using this very hacky workaround of having multiple admob objects all created with each other (like admob.game, admob.shop, admob.reward..etc) and putting them all in one qualifier, then selecting the right one via name and storing the fixed value of the admob object in a value, all the conditions are reported from the qualifiers and not specific objects, here is a snippet of it (because I know the explanation is quit very confusing):
Please login to see this picture.
Please login to see this picture.

Now this doesn't work, conditions start being reported from the wrong objects, and generally the whole thing starts conflicting with eah other breaking the whole system, especially on iOS because the behaviors are actually different there, this did work fine in 294.14 and was actually a very useful system for us, but now it's all broken in 295
I don't say it has to return to the old behavior, but can there be at least actions for changing the ads on runtime? this would be very useful

Speaking of billing library, purchase failed condition is broken, both in 294 and 295, it doesn't trigger for any reason, it seems like it was working fine pre 294, for example I tried a game built before 294 (around 293.1 I think) the condition was triggering just fine, but same game built on 294 and 295 the condition doesn't trigger for any reason, I did actually report this before, but reported it again in case it was missed, if this can be fixed in a future update it would be highly appreciated as it's kind a of a major problem with the object, thanks in advance

Quote from Linky

Sure, I may have found a workaround to convert the file, it seems to actually still read fine, but not save, I will share here in case someone falls into the same problem:
Basically do a one time conversion thing where it would load the file from the incorrect location, then save it to the correct location this time, but only once, after that you let the game load from the correct location, this will avoid users data being lost

On other notes.. it may be useful to mention more of the smaller changes in the changelog, for example billing library is now using version 5.1 and some Admob behavior seems to have been corrected, may be useful to add to the changelog, especially admob changes because I was actually using some workarounds to get past buggy behavior in 294.14, some of them were actually very useful, but as admob is now corrected, those workarounds don't work anymore and actually start breaking the whole ad system (if any of this makes sense)

What I mean by "some of them were actually very useful" is because we can't change the ads on runtime I was using this very hacky workaround of having multiple admob objects all created with each other (like admob.game, admob.shop, admob.reward..etc) and putting them all in one qualifier, then selecting the right one via name and storing the fixed value of the admob object in a value, all the conditions are reported from the qualifiers and not specific objects, here is a snippet of it (because I know the explanation is quit very confusing):
Please login to see this picture.
Please login to see this picture.

Now this doesn't work, conditions start being reported from the wrong objects, and generally the whole thing starts conflicting with eah other breaking the whole system, especially on iOS because the behaviors are actually different there, this did work fine in 294.14 and was actually a very useful system for us, but now it's all broken in 295
I don't say it has to return to the old behavior, but can there be at least actions for changing the ads on runtime? this would be very useful

Speaking of billing library, purchase failed condition is broken, both in 294 and 295, it doesn't trigger for any reason, it seems like it was working fine pre 294, for example I tried a game built before 294 (around 293.1 I think) the condition was triggering just fine, but same game built on 294 and 295 the condition doesn't trigger for any reason, I did actually report this before, but reported it again in case it was missed, if this can be fixed in a future update it would be highly appreciated as it's kind a of a major problem with the object, thanks in advance

need to see a MFA for AdMob, also need to see you inapp implementation seems fine here

getting crash every time I run my game using the latest beta 295.5 due to the debugger when jumping between frames...

- add objects to debugger on the frame being jumped to causes crash (with last executed event line message)

- jumping to frame with objects already added to debugger from previous frame causes crash (without last executed event line message)

removing "add to debugger" events on the frame jumped to AND clearing all objects from debugger before jump to frame = crash does not happen

EDIT: closing the debugger prevents the crash, so only happens with debugger open

Quote from Fernando

need to see a MFA for AdMob, also need to see you inapp implementation seems fine here

Very busy with a lot of stuff currently but will try to make a separate example when I can
For inapp, it's really simple I just have an On Purchase failed condition with an ID (SKU id exists and is the one failed) with some actions to run, the event doesn't excuete for any reason
This is not just with the project I coded, but other ones I didn't work on, they seem to have been working fine in an older update, but something broke\changed in some Fusion version that this doesn't work anymore after built with 294 or 295

I hope it's not being confused with the purchase failed condition of iOS Store, i'm talking about In-app object for Android
Probably won't say much, but here is how I handle it, event 881 doesn't work while 883 works (I also tested with a harcoded ID like 2, which exists and is the one failed, but still doesn't run):
Please login to see this picture.

I think this is a very old problem, Text on DX11 for RTL languages like Arabic appears messed up, any chances of this getting fixed or it's too hard to solve?