Masking does not work with AIR 29 on Desktop

[chaky]

Hi,

Masking in Starling does not work properly with AIR 29.

This was working fine with AIR 28.

This only happens on desktop. When compiled for mobile everything works fine.

Take a look the image:

https://ibb.co/na5rvc

I have enabled depthAndStencil in the xml file.

I am using latest Starling 2.3 and Feathers UI 3.4

[ASWC]

AIR 29 came out yesterday, I doubt very much starling is updated for that version .... Wait for starling to update.

[chaky]

This has nothing to do with Starling since this is working on mobile perfectly. Only desktop is affected. Starling does not write different code for mobile and different for desktop.

[ASWC]

You could be right but AIR 29 introduces a few new profiles that starling still needs to implement and support, you might be running in a situation where a new profile is used on desktop and not supported by starling while on mobile any of the new profile are not used and starling works just fine. The clues are toward the new profiles not being handled by starling and while starling does not write different code depending on platform it does write different code depending on available profiles it KNOWS about. There's an easy way to find out though, FORCE starling to use only profiles known to AIR 28 and compile with AIR 29, if you get starling to display correctly then please set my answer as correct and wait for the next starling update!

[chaky]

I already checked that. The same profile is used "Standard Extended" on both desktop and mobile.

[ASWC]

If that profile is forced then you might have found an AIR 29 bug, last thing to check is if starling does use it and doesn't fallback to any other profile.

[zwetan_uk]

I can pretty much guarantee that whatever you're describing here "as is"
will never be considered as a "serious bug to investigate"

first, you are poorly describing the bug eg. "boohooh it does not work anymore"
sure you're pointing to an image, but the bug is not that obvious nor proven

you gonna have to describe the bug clearly

you should definitively explain: the expected behaviour, the actual behaviour
and the steps to reproduce the behaviour

second, you are mentioning that the bug occurs while using 2 frameworks

nothing guarantee that those frameworks are not the ones introducing this specific bug

you gonna have to isolate the bug so it got no other places to hide

that's why you should submit a code sample (without frameworks or other dependencies)
that demonstrate clearly the bug.

third, you are not using the proper channel to report that kind of bug
you gonna have to put the bug in places where dev look for bugs

I'm not sure how much "AIR Development" is monitored for bugs,
I'm pretty sure less than the "AIR BETA CHANNEL", and certainly

much less than the official issues Tracker

because of those 3 reasons I'm pretty sure the bug you're "reporting"

is either not seen, not proven or not considered valid.

[chaky]

zwetan,

Your comment does not reflect reality at all.

First of all, if you are going to assume that I am retarded than there is no point in writing any response or commenting anything. In what universe and by what logic if the exact same code is working fine with AIR 28 and that exact same code is not working fine with AIR 29 is the code’s fault? Do you think I am retarded and have tested with different versions of frameworks or code? At best you could argue that AIR 29 introduced a backwards compatibility break but not certainly that code or framework introduced a bug.

Second thing you are commenting about importance of the communication channels. Hahahahahaah, are you serious? Have you fell from mars? You are acting like Adobe is some serious company that cares about communication channels. The same company that is known for ignoring serious and critical issues for several years whether they are in issue tracker, forum, mars, venus, or another galaxy Adobe just does not care and I do not see a point using different communication channel. I am just posting this here so that this may be a reminder one day when this bug was discovered in a far far future.

[zwetan_uk]

First of all, if you are going to assume that I am retarded than there is no point in writing any response or commenting anything. In what universe and by what logic if the exact same code is working fine with AIR 28 and that exact same code is not working fine with AIR 29 is the code’s fault? Do you think I am retarded and have tested with different versions of frameworks or code? At best you could argue that AIR 29 introduced a backwards compatibility break but not certainly that code or framework introduced a bug.

Honestly I don't know which code is to blame but I do know it is easy "to prove a bug"

by providing a simple "code sample" without any frameworks or dependencies, not sure

why you are so reluctant about that, it is very common practice.

I never assumed you were retarded but I'll keep the idea in mind next time I comment.

Second thing you are commenting about importance of the communication channels. Hahahahahaah, are you serious? Have you fell from mars? You are acting like Adobe is some serious company that cares about communication channels. The same company that is known for ignoring serious and critical issues for several years whether they are in issue tracker, forum, mars, venus, or another galaxy Adobe just does not care and I do not see a point using different communication channel. I am just posting this here so that this may be a reminder one day when this bug was discovered in a far far future.

All I can say is that wether a big company like Adobe or small dev teams on github,

every developers have a place where they ask you to submit bug reports,

again, a widely common practice.

Going against that is not very productive, some dev maybe nice and point out the obvious,

some other dev would just tag that as NAB and comment "OK flipping the bozo bit".

It seems you already got all the answers you need, good luck with that :D.

[ASWC]

I totally agree, that's why I strongly suspect that "bug", if existing, will be fixed on next Starling update ....

[Nitanwar]

Hi Chaky,

Thank you for reporting the issue. We were unable to reproduce the masking issue at our end.It would be really helpful if you could provide a sample project to reproduce the issue. You can either mail it to nitanwar[@]adobe.com or you can log a bug in Tracker .

Thanks.

[chaky]

Hi Nitanwar,

Thanks for your reply. It turns out that masking is working just fine. This bug seams to only affects Feathers UI List.

Why Feathers List masking is broken in AIR 29 and only on desktop (even with AIR 29 everything is working fine on mobile) while everything works fine with AIR 28 (both mobile and desktop). Could this be some backward compatibility issue?

Also one thing that is not working in AIR 29 in the same exact way as masking is not working for Feathers List is setting Starling display object scale to  negative value to achieve flipping object horizontally or vertically. So, in Starling if I set object.scaleX = -1 the objects disappears. And this bug is also only showing on desktop and on mobile everything works just fine.

Could this be happening because in AIR 29 Adobe has made some changes to the maximum texture memory  limit and maximum texture resolution limit for desktop only. Those limits were not changed for mobile and maybe that is why on mobile everything is working fine.

Did you change something in AIR 29 that could affect masking in some situations?

[ASWC]

I dunno ... 3D context profiles? As we've been saying from the beginning .... Btw I do love the "Could this be some backward compatibility issue?" When feather is not updated to AIR 29, that sentence is a beauty ... I could be a medium with superpowers when I say everything will be fixed once starling and feather run their next update ... But what do I know right? Let's just pretend it's AIR 29 fault and starling and feather can run without problem on any AIR version including the ones they were never developed for ...

[chaky]

Well again as I said before context profiles are the same.

First of all your posts are useless. They have no value whatsoever. You do not contribute anything with your stupid teasing comments.

Second of all I could not care less for desktop build of the app since I do not use it and do not need a desktop application. I am only using mobile and for mobile everything is working fine. It is just my good will that I want to point out to an existing bug that's all.

[chaky]

Ok, I just got reply from Josh. He says that masking Feathers List is working fine on Mac and scaling object to negative value is also working fine on Mac and since I am working on Windows it seams that this issue is only affecting Windows desktop machines.

[Josh Tynjala]

When feather is not updated to AIR 29, that sentence is a beauty ... I could be a medium with superpowers when I say everything will be fixed once starling and feather run their next update ... But what do I know right?

This is not helpful or professional. How about we take things down a notch?

I don't see anything obvious in the AIR 29 beta release notes to indicate that Feathers (or Starling) might need to be updated for masking to continue working correctly. While Stage 3D profiles were suggested, and the release notes say that the ENHANCED profile has been added to desktop, chaky\ says that's not the profile being used. With all that in mind, as the developer of Feathers, I would not expect to have to update anything in my library to remain compatible with the new version of AIR. This certainly seems like a bug in AIR 29 beta to me. With a beta build, it should be unsurprising that new bugs might be introduced.

[zwetan_uk]

Josh Tynjala

When feather is not updated to AIR 29, that sentence is a beauty ... I could be a medium with superpowers when I say everything will be fixed once starling and feather run their next update ... But what do I know right?

This is not helpful or professional. How about we take things down a notch?

About being helpful and professional, maybe teach chakyl how to log a bug in the Adobe bug tracker

[chaky]

zwetan,

Please, never go full retard.

Is your life so pathetic you got nothing better to do than to post stupid comments on Adobe forum?

Please, go outside, play basketball or football or go

see a movie. Make use of your time, dont be a no lifer.

[Josh Tynjala]

@chaky\ Speaking of keeping things professional, try to stop throwing around the R word here. I understand that you're dealing with some strangely antagonistic replies here, but I think it's better to just try to stay on topic and ignore anyone who isn't actually helping. Not everyone deserves a response.

[zwetan_uk]

@josh

humm yeah, I do think chakyl created himself this antagonistic situation
at least 3 persons on 2 different forums told him to log the bug in the Adobe bug tracker

I guess you have different rules, in good conscience, on what is acceptable or professional
I get it it's a guy from your forum you feel like defending him even if his attitude compare to a 5yo

@chakyl

you're not a special snowflake, your bug  (if valid) should be posted to Adobe bug tracker
like anyone else would do, it takes 5 minutes and I'm pretty sure you have the motor functions

to be able to achieve this enormous difficult task

[ASWC]

Agreed, my first reply to him was normal and his reply was so affirmative and rude that all my following reply were ... what they were. From me he got what he asked for, I got no apologies to make.

[ASWC]

Strangely? So being systematically rude and/or insulting to all persons that reply and generating in return rudeness from those people is strange?

[chaky · Accepted solution]

This bug has been fixed in the latest AIR 29.0.0.103 Beta. As expected this was bug in AIR SDK. Adobe thank you for fixing this bug.