S3Drive - Scrolling weirdness in the sync section

Hi, I am facing an unusual issue on the sync section. The scrolling doesn't seem to work properly and it gets stuck occasionally. I haven't seen this issue in any other app or section within this app on my Pixel 8 Pro. I am attaching a video. Let me know your thoughts. Also, are there plans to do something about these errors? (edited)

Thanks for your feedback. Scrolling issue will be addressed in a next release. Speaking of sync issues, I can see you have at least 3 sync entries, on the video I can only see 2 errors related to Proton back-end. One error points looks like Proton API gave up or/and there were connection issues from your end. Can you let me know what were the circumstances, did you start these sync yourself and then waited for the results or perhaps it was started when app was in the background? Was it on WiFi or mobile connection? Speaking of other error "wrong credentials" it may well be that Proton integration: https://github.com/henrybear327/Proton-API-Bridge/ isn't entirely thread safe, so if there is more than one parallel execution weird things may happen. I only speculate, but it looks like security token in one of the execution was possibly overwritten by the other execution. Quick question did you set up Proton using username/password or perhaps used just the access tokens / refresh token as mentioned here? https://www.reddit.com/r/ProtonMail/comments/18s211d/comment/kzfqub7

devutils's comment on "Proton Drive for Linux"

Explore this conversation and more from the ProtonMail community

Thanks for getting back to me. That's great to hear! Regarding proton, things seem to crumble when you start sending a few gigabytes of data. Eventually you get a message like "Your client has been blocked due to suspected abuse. Go to https://proton.me/support/appeal-abuse to appeal" (which I have done and the support haven't been terribly helpful). The solution I have found has been to recreate the "proton" profile. So I keep doing this and rush to get the 6 digit code into JSON (see below) and add it into S3 Drive before it expires... I don't know if it's concurrency related. This feels more like locking out API users based on client ID. I don't blame this on S3 Drive. What is problematic is the handling and displaying of errors across most of the sync jobs I have set up (not just proton ones) and the fact that they seem to not vanish. I have other errors like: hash mismatch (corrupted on transfer?), exception on transfer (no useful info) etc. Ideally this would retry automatically and hide the logs & errors from the user. Some of them I can't do anything about so there is little point in notifying me about it.

Quick question did you set up Proton using username/password or perhaps used just the access tokens / refresh token as mentioned here?

Sort of. What I paste into the new profile input box is something like this: {"proton-vN": {"username": "myusername", "password": "mypassword", "2fa": "123123"}} This seems to generate the other bits: { "2fa": "1231234", "client_access_token": "xxxxxxxxxxxxxxxxxxxxx", "client_refresh_token": "xxxxxxxxxxxxxxxxxxxxx", "client_salted_key_pass": "xxxxxxxxxxxxxxxxxxxxx==", "client_uid": "xxxxxxxxxxxxxxxxxxxxxx", "password": "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx", "type": "protondrive", "username": "myusername" } (edited)

Also it'd be amazing to see the progress of sync tasks. I have seen the spinning icon showing that it's syncing but % complete or X/Y GiB transferred or something would be amazing as well because right now it's not clear if a job is stuck, it's just begun, it is finishing etc...

3yb

Thanks for getting back to me. That's great to hear! Regarding proton, things seem to crumble when you start sending a few gigabytes of data. Eventually you get a message like "Your client has been blocked due to suspected abuse. Go to https://proton.me/support/appeal-abuse to appeal" (which I have done and the support haven't been terribly helpful). The solution I have found has been to recreate the "proton" profile. So I keep doing this and rush to get the 6 digit code into JSON (see below) and add it into S3 Drive before it expires... I don't know if it's concurrency related. This feels more like locking out API users based on client ID. I don't blame this on S3 Drive. What is problematic is the handling and displaying of errors across most of the sync jobs I have set up (not just proton ones) and the fact that they seem to not vanish. I have other errors like: hash mismatch (corrupted on transfer?), exception on transfer (no useful info) etc. Ideally this would retry automatically and hide the logs & errors from the user. Some of them I can't do anything about so there is little point in notifying me about it.

Quick question did you set up Proton using username/password or perhaps used just the access tokens / refresh token as mentioned here?

Sort of. What I paste into the new profile input box is something like this: {"proton-vN": {"username": "myusername", "password": "mypassword", "2fa": "123123"}} This seems to generate the other bits: { "2fa": "1231234", "client_access_token": "xxxxxxxxxxxxxxxxxxxxx", "client_refresh_token": "xxxxxxxxxxxxxxxxxxxxx", "client_salted_key_pass": "xxxxxxxxxxxxxxxxxxxxx==", "client_uid": "xxxxxxxxxxxxxxxxxxxxxx", "password": "xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx", "type": "protondrive", "username": "myusername" } (edited)

Speaking of Proton blocking rclone (and S3Drive as a result), there was some backlash ~12 days ago: https://www.reddit.com/r/rclone/comments/1d2pm6p/rclone_blocked_by_proton_drive_it_seems/ We've asked them if they have any advice for Rclone/S3Drive users that we could pass on: https://www.reddit.com/r/rclone/comments/1d2pm6p/comment/l6albgs Proton reply:

Hi! Yes, The particular issue seems to be that many users of rclone run a sync every 10 minutes. We have now modified some settings on our end which should make it easier to use rclone, but we advise not to run sync too frequently.

I hope that eventually they will either provide an API or find out ways to improve their API performance and tweak the abuse auto detection. ==

What is problematic is the handling and displaying of errors across most of the sync jobs I have set up (not just proton ones) and the fact that they seem to not vanish. I have other errors like: hash mismatch (corrupted on transfer?), exception on transfer (no useful info) etc. Ideally this would retry automatically and hide the logs & errors from the user. Some of them I can't do anything about so there is little point in notifying me about it.

Current verbose setup is somewhat deliberate action to clearly display the cause of failure, so users can report. We weren't able to reliably test it on all back-ends and platforms, so we've released early version in: [1.7.0] - 2023-12-29 release and then in February [1.7.9] - 2024-02-09 released Folder sync on Android along with first major redesign of Sync functionality on all platforms. History: https://s3drive.app/changelog

1

7:30 PM

Yet another redesign, including better error handling, transfer/progress indicator, notifications will come next (possibly later this year) as we keep discovering some gotchas around Rclone and build knowledge about ways to integrate with its core. Before that redesign happens we'll still need to address other higher priority items, e.g. improved cipher: https://github.com/rclone/rclone/issues/7192 to allow secure sharing and write mode for Android mounts: https://s3drive.app/roadmap2024 Background processing for Rclone will likely be executed in 2025, as our 2024 roadmap is already tight enough. Please bear with us and keep sending your feedback, it's not lost... and it helps us to better allocate our development resources. Thank you. (edited)

1