search

airsdk / Adobe-Runtime-Support

Report, track and discuss issues in Adobe AIR. Monitored by Adobe - and HARMAN - and maintained by the AIR community.
206 stars 11 forks source link

ITMS-90428: Invalid Swift Support #892

Open idanasher opened 3 years ago

idanasher commented 3 years ago

After completing & compiling my distribution ipa , I uploaded it via the Transporter app on my mac.

After the upload, I got this message from the app store connect:

ITMS-90428: Invalid Swift Support

AIR 33.1.1.476 fro ios Compiling with Animate 2020 on Windows 10

spielepalast commented 3 years ago

What is the status of this ticket? We got the same error from Apple when Uploading our ipa file:

ITMS-90428: Invalid Swift Support - The files libswiftFoundation.dylib, libswiftos.dylib, libswiftMetal.dylib, libswiftDispatch.dylib, libswiftUIKit.dylib, libswiftObjectiveC.dylib, libswiftCoreGraphics.dylib, libswiftDarwin.dylib, libswiftCoreFoundation.dylib, libswiftCore.dylib, libswiftCoreImage.dylib, libswiftQuartzCore.dylib don’t match /Payload/doppelkopf.app/Frameworks/libswiftFoundation.dylib, /Payload/doppelkopf.app/Frameworks/libswiftos.dylib, /Payload/doppelkopf.app/Frameworks/libswiftMetal.dylib, /Payload/doppelkopf.app/Frameworks/libswiftDispatch.dylib, /Payload/doppelkopf.app/Frameworks/libswiftUIKit.dylib, /Payload/doppelkopf.app/Frameworks/libswiftObjectiveC.dylib, /Payload/doppelkopf.app/Frameworks/libswiftCoreGraphics.dylib, /Payload/doppelkopf.app/Frameworks/libswiftDarwin.dylib, /Payload/doppelkopf.app/Frameworks/libswiftCoreFoundation.dylib, /Payload/doppelkopf.app/Frameworks/libswiftCore.dylib, /Payload/doppelkopf.app/Frameworks/libswiftCoreImage.dylib, /Payload/doppelkopf.app/Frameworks/libswiftQuartzCore.dylib. Make sure the files are correct, rebuild your app, and resubmit it. Don’t apply post-processing to /Payload/doppelkopf.app/Frameworks/libswiftFoundation.dylib, /Payload/doppelkopf.app/Frameworks/libswiftos.dylib, /Payload/doppelkopf.app/Frameworks/libswiftMetal.dylib, /Payload/doppelkopf.app/Frameworks/libswiftDispatch.dylib, /Payload/doppelkopf.app/Frameworks/libswiftUIKit.dylib, /Payload/doppelkopf.app/Frameworks/libswiftObjectiveC.dylib, /Payload/doppelkopf.app/Frameworks/libswiftCoreGraphics.dylib, /Payload/doppelkopf.app/Frameworks/libswiftDarwin.dylib, /Payload/doppelkopf.app/Frameworks/libswiftCoreFoundation.dylib, /Payload/doppelkopf.app/Frameworks/libswiftCore.dylib, /Payload/doppelkopf.app/Frameworks/libswiftCoreImage.dylib, /Payload/doppelkopf.app/Frameworks/libswiftQuartzCore.dylib.

idanasher commented 3 years ago

@spielepalast , the status is "in process"

@ajwfrost Andrew is working on the air sdk and trying to solve this issue.

You can see this tread as well for updates: https://github.com/airsdk/Adobe-Runtime-Support/issues/776

ajwfrost commented 3 years ago

Should be fixed in 33.1.1.533, please check and re-open if necessary. Thanks

urthling commented 3 years ago

Just pushed a 533 build to the App Store and I'm seeing the same:

ITMS-90428: Invalid Swift Support

ajwfrost commented 3 years ago

Can I check what build target type you used - was it "ipa-app-store"? Any chance we can look at your IPA file please? https://transfer.harman.com/requests/TQ1vfER8YcDkRh5jp7nLUr thanks

urthling commented 3 years ago

Aye, using: -target ipa-app-store -- just sent the IPA..

ajwfrost commented 3 years ago

Thanks. So that all looks fine, the signatures check out, and apart from these signatures the content of each architecture block are identical. From what we could tell - having inspected an IPA that previously worked - this is all good.

But, the Apple message suggests that there should be no post-processing done on these other dylib files. The two differences that we do have are (a) we're only including the architectures that we're using, and (b) we're re-signing them.

Let me see whether we can change either/both of those, and get it to a point where Apple accept it...

urthling commented 3 years ago

Gotcha, sounds good.. Just fyi, I did notice it's about 20MB larger than a .476 build..

ajwfrost commented 3 years ago

Yes, the size difference is likely to be because we are just moving the swift libraries into the "SwiftSupport" folder unchanged .. which means they have all 4 CPU architecture variants in them.

It would be lovely if Apple published details of what they actually want this IPA format to contain.. sadly it's a bit of trial and error, and inspecting packages that work, for us to work this out...

urthling commented 3 years ago

Understood.. power to you and the team!

ajwfrost commented 3 years ago

Hi @urthling - we have a version of ADT that's got some extra configuration settings in it, are you able to try this please? Details are in the below message: https://transfer.harman.com/message/bOpJ9PNdRCeyCRsJQ6mgQ9 Essentially, if you can use this and add the below two lines to the "lib\adt.cfg" file, it may then end up being accepted by Apple:

OnlyIncludeSwiftUsedArchsInSupport=false
OnlyIncludeSwiftUsedArchsInPayload=false

thanks

urthling commented 3 years ago

Excellent news to report -- so, with your first suggestion of:

OnlyIncludeSwiftUsedArchsInSupport=false
OnlyIncludeSwiftUsedArchsInPayload=false

I received the first email from Apple with no mention of ITMS-90428 and I received the second email from them saying that the build had completed processing -- so this issue appears to have been resolved. Well done!

ajwfrost commented 3 years ago

Great, thanks for checking that! So I think we should set this behaviour as 'default' for app store release builds..

urthling commented 3 years ago

Quick note on this -- I notice that the 554 adt.cfg does not have the two entries

OnlyIncludeSwiftUsedArchsInSupport=false
OnlyIncludeSwiftUsedArchsInPayload=false

I'm getting the ITMS-90428: Invalid Swift Support email from Apple

After adding them back in -- I'm still getting this email from Apple

ajwfrost commented 3 years ago

Yes, those changes came after the 554 release candidate was built. We're going to have to do (yet) another release to update the behaviour here so that these values are default false... In the meantime, the ADT from the above link was I think built on top of the .554 release candidate anyway so you can use that for the meantime.

thanks

urthling commented 3 years ago

Ah, my bad, gotcha, k, thank you Andrew