Describe the bug Multiple crashes due to runtime.raise_trampoline, causing thread to crash with SIGABRT ABORT.

To Reproduce Unknown. Collected via Crashlytics.

Smartphone (please complete the following information): Seems to occur more often on iOS 15 (currently in beta), but few instances of it occuring on iOS 14. App version ranges from 1.1.2 to 2.1.4.

Additional context runtime.raise_trampoline appears to be from Go runtime, indicating an issue with GopenPGP library. Some crash reports indicate that the crash occurs when using the non-streaming API, at the point where finish() is called on the new low memory attachment processor.

Describe the bug A clear and concise description of what the bug is.

App crashes when i click something

To Reproduce Steps to reproduce the behavior:

1. Go to '...'
2. Click on '....'
3. Scroll down to '....'
4. See error

Click anything like photos, setting options

Expected behavior A clear and concise description of what you expected to happen. Show contents

Screenshots If applicable, add screenshots to help explain your problem.

Smartphone (please complete the following information):

• Device: iPhone 7
• OS: iOS 14.6
• Browser: [e.g. stock browser, safari, firefox]
• App Version: 2.1.1

Additional context Add any other context about the problem here.

https://www.icloud.com/iclouddrive/0Ac-QPcqV0YP7gncvbKnBPIzg#swiftybeaver

• Issue for release 2.1.9(.1) only.
• Occurs due to a http timeout which is reached when transferring a large file. The new Soto framework has a default timeout set to 20 seconds.
• Solution is to increase the http timeout for the S3 service client. 7 days is the default according to https://developer.apple.com/documentation/foundation/nsurlsessionconfiguration/1408153-timeoutintervalforresource.

Describe the bug Title. From Crashlytics:

Crashed: NSOperationQueue 0x107e09960 (QOS: UNSPECIFIED)
EXC_BAD_ACCESS KERN_INVALID_ADDRESS 0x0000000000000030

To Reproduce Unknown. Collected via Crashlytics.

Smartphone (please complete the following information):

• Device: iPhone 12, iPhone SE 2
• OS: 15.0.0, 15.0.1
• App Version: 2.1.7.1, 2.1.7.1

Additional context Believe this is happening due to workaround implemented in AWSAdapter to stop completion callback being called twice on multi-part upload failure.

Photos are currently loaded into memory and encrypted in full, using a custom Protobuf format for storing the key and data packets.

Instead, photos should use the streaming API and be encrypted as standard PGPMessages, which is currently used for videos only.

1. Encrypt new photos with streaming API and store as standard PGPMessages.
2. Decrypt newer photos with streaming API.
3. Provide fallback decryption for photos encrypted the old way (until migration upgrade is implemented).

Note: this should hopefully resolve #24 too.

1. Video length labels on cells.
2. Date labels on header views of collection views.
3. Selection counter badge number (usually once reached a triple digit count).

Describe the bug Occurs when attempting to cast the html returned after evaluating the javascript – unable to cast the returned html as String.

To Reproduce Unknown. Collected via Crashlytics. Presumably due to an invalid cast, which must occur when the javascript has failed to load for whatever reason (e.g. no internet connection perhaps?).

Smartphone (please complete the following information):

• Device: iPhone 6S
• OS: iOS 14.7.1
• App Version: 2.1.3.1

Describe the bug Heap corruption according to Crashlytics.

Issue 1:

__destroy_helper_block_e8_32s40s48s56s

Crashed: NSOperationQueue 0x105b26b10 (QOS: UNSPECIFIED)
EXC_BAD_ACCESS KERN_INVALID_ADDRESS 0x000000042e707080

Issue 2:

-[AWSURLSessionManager taskWithDelegate:]

Crashed: NSOperationQueue 0x10701aef0 (QOS: UNSPECIFIED)
EXC_BAD_ACCESS KERN_INVALID_ADDRESS 0x00000008f56bbfa0

To Reproduce Unknown. Collected via Crashlytics. Occurs while uploading data to S3.

Smartphone (please complete the following information): Crash 1:

• Device: iPhone 12, iPhone XS
• OS: iOS 15 (beta)
• App Version: 1.1.2.17, 2.1.3.1

Crash 2:

• Device: iPhone 12
• OS: iOS 15 (beta)
• App Version: 2.1.3.1

Additional context Possibly due to #24 as other threads are at the same point as mentioned in that issue, but uncertain.

• Import operation should save original filename of asset.
• When sharing and saving, use original filename.
• Write an upgrade operation to import original filenames from photo library.
• Handle assets without original filenames or that haven't been upgraded yet.

"Backup and Sync" setting might give the impression that any deletions in the Photos app will be reflected in TripUp Cloud backup. Change to something more clear, like "Backup Library".

App should notify user to keep app in the foreground when running a manual import – unlike auto import, manual import and follow up actions (such as being added to a group album) is at risk of being lost if app state is lost.

• [x] Refactor code to support custom S3 endpoint.
• [ ] Refactor code to support custom server endpoint.
• [ ] Refactor API caller to support 2 endpoints – one for self-hosted server and one for central server calls (for calls related to Firebase service and notifications).
• [ ] Design user interface to support custom endpoint entry on user sign in.
• [ ] Call API to set token claim and refresh token once custom endpoint is entered.
• [ ] Force refresh token when creating a new user.
• [ ] Force refresh token when authorisation details are updated/changed.
• [ ] Update README.

Describe the bug Number of references to "photo" in the user interface (e.g. when saving an item to the photo library from fullscreen view). As TripUp now supports videos, should be changed to a more generic term like "item" or "asset".

Describe the bug Title. Crashed main queue. EXC_BAD_ACCESS KERN_INVALID_ADDRESS 0x0000000116a7eb54

To Reproduce Unknown. Collected via Crashlytics.

Smartphone (please complete the following information):

• Device: iPhone SE 2
• OS: iOS 15.0.0, build 19A5307g (beta 4)
• App Version: 2.1.1, build 1.1.1.1
• Orientation: Portrait
• RAM free: 306.44 MB
• Disk free: 75.52 GB
• Jailbroken: No
• Date: 28 Jul 2021, 16:09:59

Additional context Possible bug due to beta iOS. Monitoring for future issues. Included incomplete crash log from Crashlytics (cut out minor threads). crashlytics-relevant-logs.txt

AssetManager operations should run until an error is encountered, instead of preemptively and continuously checking system state and external variables to determine if operation should execute. Instead, fire operation and handle errors accordingly.

Depends on tripupapp/tripup-server#8, as operation must fail when user doesn't have the required entitlements.

Describe the bug Import operations started in the foreground remain suspended in background state. This makes them unable to be set as dependencies (as required by #15) as well as unresponsive to cancellation events until the app returns to the foreground.

To Reproduce Steps to reproduce the behavior:

1. Enable Debug logs and Backup & Sync.
2. Leave phone on standby and on charge to begin background process.
3. Later, pick up phone and open app.
4. Observe in the logs that the foreground import operations resumed only after app returned to foreground, regardless of other events including cancellation.

Expected behavior Foreground import operation should resume and complete in background mode (without spawning another operation for the next batch).

Dependent on #17