Issues in desktop client and mobile apps

[shibacomputer]

Is your feature request related to a problem? Please describe.
I'm here because of your CEO's response to the Twitter news, and my feature request is: Don't.

Describe the solution you'd like
Finish your half-hearted Dropbox clone before you start work on your half-hearted Twitter clone.

Describe alternatives you've considered
A world where Nextcloud has a tight focus, prioritises data integrity, and commits to a cohesive strategic vision instead of branding borderline prototypes as production ready and shipping software updates that endanger user data.

Additional context
Fix these first (especially issues that describe data loss) then reconsider again whether Nextcloud should build a Twitter clone:
https://github.com/nextcloud/server/issues
https://github.com/nextcloud/desktop/issues
https://github.com/nextcloud/ios/issues
https://github.com/nextcloud/android/issues

[shibacomputer]

Examples of the Nextcloud Desktop app's potential data loss:
#4016
#4451
#4446
#4440
#4319

Examples from the Nextcloud iOS app:
nextcloud/ios#1950
nextcloud/ios#1395
nextcloud/ios#1489
nextcloud/ios#1187
nextcloud/ios#1788

Examples from the Nextcloud Android app:
nextcloud/android#7905
nextcloud/android#8285
nextcloud/android#3291
nextcloud/android#6570

Contributions welcome!

[jancborchardt]

Thanks for listing some specific issues you’d like to see prioritized. cc @tobiasKaminsky on that, also cc @mgallien @AlvaroBrey @marinofaggiana specifically for desktop, Android and iOS.

Just a note @shibacomputer that the folks working on any web part (like Social) are different people than would work on any of the clients, so these things are not mutually exclusive.

[ph00lt0]

Could not agree more, Nextcloud is super unstable, releases often crash behavior of apps like e2ee and groupfolders which are poorly implemented. The design is a very poor experience on desktop both in app and client. Sync is horrific compared to alternatives and the performance of the web client is a joke. Please focus on what your goals are.
If you want to start something ambitious implement e2ee in the web, instead of saying this is impossible (it is not), but first address the enormousness amount of problems. The current desktop app of nextcloud causes dataloss for more then a month now... personally switched to mountain duck because Nextcloud can't seem to maintain its core functionality in good fashion.

[mediaformat]

A decent round up of issues that have nothing to do with this specific app.

[mrgohin]

This is also related since e2ee is broken like ever. I tried it once and yes it is not useable at all. Implementation and protection of your data is questionable since e2ee has been famous for losing data or at least is helping you with that.

[shibacomputer]

A decent round up of issues that have nothing to do with this specific app.

Just a note @shibacomputer that the folks working on any web part (like Social) are different people than would work on any of the clients, so these things are not mutually exclusive.

I am aware that the Nextcloud social app is maintained by separate folks to the clients team, and also that the Social team is not responsible for the direction that Nextcloud takes. Regardless, I created this issue in response to Frank Karlitscheck's public announcement that Nextcloud intends to invest heavily in this very project due to current events. I am someone trying to maintain a mid-sized Nextcloud instance who wants to believe in Nextcloud's mission but is in absolute despair at the state of the project.

[jaredthirsk]

Brand risk makes this relevant to nextcloud social.

I also don't know where else to post this, since every other place gets into details rather than bottom line.

I am someone trying to maintain a mid-sized Nextcloud instance who wants to believe in Nextcloud's mission but is in absolute despair at the state of the project.

I want to believe in Nextcloud's mission too and did until this week when I finally looked into some "modification time" messages in my sync program and embarked on a journey to figure out what I need to do about it as the maintainer of the instance for our small business. That's when I discovered that modification times have been corrupted, and files have apparently stopped syncing for many of our users. Yikes.

Do my users want to try to recover modification times? I have to find out now. Did my users change files on their own computers and create sync conflicts without knowing? I don't know. If their disk dies, will we lose data? It sounds like it.

I have spent hours on this, looking at a wiki page from Nextcloud, 15 or so gitlab issues (I lost count), a couple merge requests, a bunch of forum posts, a few reddit threads, and I still have no idea what to do. It sounds like the latest version up until a few days ago (3.4.4) still had the problem, and I can't tell if 3.5 is safe. I don't have time for this. We just switched email providers from a small potato to Microsoft, and hey look, we happen to now also have OneDrive for free! At this point, if someone in our company wants to migrate our data there, I am not going to push back. I hate vendor lock in, but we also have things to get done and have to have confidence that our file sharing system won't corrupt our data or stop syncing without us realizing what's going on for months.

I have seen mention of sync issues in Nextcloud over the years requiring rescans or whatever but never been bitten by them. Now I have. It's a good thing I keep backups (I also use syncthing, which seems way more reliable at syncing data -- maybe Nextcloud can learn from them -- maybe hijack their go-based sync engine if that would solve problems long term, I don't know), but it's still a nightmare in progress about how to go about fixing it.

And it's not just about the bug -- it's about communication and handling bugs when they do arise. Is there some central place to discuss this? A status tracker? One gitlab issue? I see several gitlab issues open, and one auto-closed by a bot, which is infuriating. I see one new fix-it feature for occ but the milestone is Nextcloud 25 -- I can't wait until then. I see there is an entire gitlab repo for a toolkit to fix this issue -- good but I still don't know how to best recover mtime's or the status of 3.5. And I still have 32 Nextcloud related browser tabs to go through. It cannot be overstated enough how big of a nightmare data loss / data corruption is. Nextcloud has built a ton of good will, but an unresolved data loss issue is like fire that will grow from a small flame to an inferno before you know it.

Nextcloud does a ton of things, but I think the top feature has always been file synchronization, and if Nextcloud can't get that right, why should it be trusted with anything else?

[ThatCoffeeGuy]

Here I am, at 10:30PM troubleshooting why a (hopefully) tested android application simply stopped exercising it's main purpose after an update, I found this thread. Ironic.

Honestly,

• after losing about 5% of my data because of a bug about a year ago,
• after ruining my file modification times and spamming logs full of errors generated by itself about a month ago,
• after updating my android client (genuine phone, genuine android) using google play and now auto-upload stopped working since 4th of May,

please stop even considering jumping into new projects.

"If there are specific enhancements and fixes you'd like to see, they can be brought to devs attention or directly contributed."

Well, we could start with the one where people's files are losing part of metadata and nextcloud just dies, spamming errors and manual fixes are needed to be applied.
#4378

I really appreciate the effort and work put into this project, but for real, I've lost data and can't even count how many hours wasted on troubleshooting. Not to mention that time when a featured(!!!) application wasn't working at all with the stable released Nextcloud Server.

[decentral1se]

Thank you for raising this @shibacomputer. Important.

[mrgohin]

@jancborchardt is there any statement from Nextcloud to be expected about this whole mess in near time or at all?

If we can do something as community or whatever jumps in mind we should start to talk about it as a community and begin finding solutions about these summarized issues by @shibacomputer

[jancborchardt]

If we can do something as community or whatever jumps in mind we should start to talk about it as a community and begin finding solutions about these summarized issues

Definitely thanks to everyone who is already opening issues and providing details, logs, etc. It is often difficult to debug cause there is just such a big possible matrix of setups.

Helpful is:

• Staying positive and friendly in communication (can't stress this enough :)
• Contributing to fixes if you can
• Reviewing pull requests
• Testing betas and opening issues when you find some
• Seeing that this is an open source project where everyone can join in and make a difference
• Checking out our job openings: https://nextcloud.com/jobs – we especially encourage applications of people from groups underrepresented in tech

Also keep in mind that it's all a balance – Nextcloud is not and never was just focusing on file sync & share. Calendar & contacts and the app ecosystem have been an integral part of the system from the beginning.

Thank you all!

[jancborchardt]

FYI @tobiasKaminsky @AndyScherzinger transferred to desktop repo, since while also relevant for mobile, it's not fitting in the social repo at all. To be discussed and probably best split into the issues listed.

[tobiasKaminsky]

Hello,
first of all thanks for taking your time to analyze and collect the topics where you see that our product has problems.
We already try to do as much bug fixing as possible, besides new features.
We try to fix most urgent/annoying ones first. This may either a rare data loss which affects only specific settings, or a "minor" papercut, which affects many users, but has no real consequences.
For this we monitor GitHub bugs, but also check how many reactions/answers a specific bug has.
Thus means that such meta topics collecting different bugs are not helping in terms of helping to prioritize.
Therefore please next time answer in the individual bugs or use reactions on initial post within a bug.
I will close this issue, but will check with my team on those tickets.

Best regards
Tobias

[decentral1se]

Classic Stuff 🙄

[mrgohin]

I guess I might look around for another solution. Nextcloud proofs really nothing. Move it around and then close it in silence. Congrats guys. May your social stuff make you rich!

[jaredthirsk]

Helpful is:

• Staying positive and friendly in communication (can't stress this enough :)

We want to love Nextcloud. Please have some empathy for our situation. I want you to succeed. You live in a bubble of being inside the development world, and users can be annoying and wrong in how they complain -- I get it, but I hope people in the user world who are having problems can be constructive in suggesting improvements in how to handle a situation like this. I'm not sure you guys fully get it yet, which is why we are asking for more empathy here, and doing random desperate things like creating a rollup issue on the social project, and asking you to not make a social product -- these are cries for help, even if not articulated precisely. At least be glad that we are still trying to talk to you, rather than giving up and telling the world instead.

Situation:

• Files are being corrupted, by desktop client 3.4.0, with reports of there still being issues in 3.4.4. We don't know if 3.5.0 fixes it. Sure there are changelogs, but they are a bit obtuse and we are not the changelog whisperers. (I am a programmer and I couldn't tell. I could read the code in the commits, but that's not reasonable.)
• There are a kajillion little gitlab issues and merge commits that are very difficult and time consuming to piece together.
• What we need is a sort of task force and status page for this kind of issue.
• There is a wiki page, but it hasn't been updated to acknowledge when or if the issue was resolved, or the status of the six (!) new versions since the wiki page was created. (3.4.1 3.4.2 3.4.3 3.4.4 3.5.0 3.5.1) A user reported it was still broken in 3.4.4. I asked in a gitlab issue if it was fixed in 3.5.0 but nobody answered me.
• There is no single place to discuss the wiki page, or this issue, just a bunch of gitlab issues that are sometimes overly specific to a problem. Gitlab comments seem too easy to miss.
• There is one place in the wiki that rubs me the wrong way (honestly, it infuriates me): "...the original, correct date and time are lost. In most situations, this is sufficient." Now this may be entirely true, but what about the rest of us who don't want data loss? I wish there was a status page and blog for major issues like this. (https://github.com/nextcloud/desktop/wiki/How-to-fix-the-error-invalid-or-negative-modification-date)

Also keep in mind that it's all a balance – Nextcloud is not and never was just focusing on file sync & share. Calendar & contacts and the app ecosystem have been an integral part of the system from the beginning.

Ok, fine, but this doesn't really help your cause on this issue right now to say this. How I interpret it: "we care even less about file sync than you thought, and care less about your data loss problems than you thought. You may care about your ongoing data loss issue, but please be mindful that we are busy at the moment adding more calendar features."

Even for a 100% non-commercial open source project, I would expect some sort of task force response team style communication if a critical emergency sort of issue such as data loss and stoppage of synchronization (potentially leading to more data loss) arose. But you guys have commercial revenue, so I would expect this even more.

For a 100% non-commercial open source project, there should still be some leadership and diplomatic response shown for critical issues, otherwise, the project should be regarded as unmaintained (and forked by someone who can provide sufficient leadership.) But you guys have lots of people -- surely someone can show some leadership?

I looked at your job posting board. It looks like you have a lot of positions to fill. Here's a random suggestion:

• for critical issues, (such as data loss), create an incident in a "Customer Relations Incidents" spreadsheet.
• Schedule meetings for every two weeks until the issue is resolved. Discuss customer response, and updates on fixes.
• People required at the meeting:
  • your future Developer Relations Manager (whose role should perhaps be expanded, if not implied, to monitor and provide initiative on critical gitlab issues such as data loss)
  • Documentation writer, to assist in communicating status to end-user admins (if above person can't handle this)
  • Marketing Manager or delegate (better they're involved now than when people get so frustrated they resort to twitter-shaming / reddit)
  • Lead Developer or delegate
  • Primary developer on problem area
  • Secondary developer to give support/review to primary developer
  • (One individual may cover multiple roles in some cases, but shouldn't in other cases -- as a developer myself, I suggest letting the developer focus on fixes and nothing else, and taking off burdens of responsibility and stress as much as possible. Also don't blame the developer for the issue, since testing should be the purview of QA.)
• Create a status page (a forum post would probably be good enough, if you edit the top post in the thread to keep it updated) to keep people updated, and let people ask for clarification.
• Link to the status page (or foum post) in the wiki, so that there is a nexus for the problem, rather than a kajillion overly specific gitlab issues that get closed and ignored.
• It sounds like another member who commented above doesn't think enough users are impacted by data loss, and therefore it's not a big deal: "rare data loss which affects only specific settings". This is why I recommend involving Marketing/Sales, because they may have a better appreciation than you (developers who want to work on new features and invent a social app) about how well "98% of our users don't experience data loss!" works as an advertisement.

Final status report from my own situation:

• I manually tracked down modification timestamps for all affected files from client machines and server "versions" folders, and manually set the date from the command line, and didn't lose any data that I know of. Fortunately I only had around 100 files to fix.
• I got all my users on 3.5.0.
• I let my team know I wasn't happy with how this was handled, and if anyone comes to me with a recommendation for an alternative, I am going to find it difficult to say no.
• But if Nextcloud figures out how to respond to a situation like this, it is not too late to convince me.

So I am done with this issue, as far as I can tell, and thought I'd share my final thoughts in case it can somehow ease future problems or improve the issue handling response to the extent that it saves me the hassle of migrating off Nextcloud someday, because I don't want to have to do that, because it is extra work for me, and because I love the mission of Nextcloud and want it to succeed.

Cheers

[ThatCoffeeGuy]

Hello, first of all thanks for taking your time to analyze and collect the topics where you see that our product has problems. We already try to do as much bug fixing as possible, besides new features. We try to fix most urgent/annoying ones first. This may either a rare data loss which affects only specific settings, or a "minor" papercut, which affects many users, but has no real consequences. For this we monitor GitHub bugs, but also check how many reactions/answers a specific bug has. Thus means that such meta topics collecting different bugs are not helping in terms of helping to prioritize. Therefore please next time answer in the individual bugs or use reactions on initial post within a bug. I will close this issue, but will check with my team on those tickets.

Best regards Tobias

Well, my open issue about data-loss is at 86 reactions now. Is that considered low? Or since it's "only metadata" it doesn't matter as much?:

#4378

[ph00lt0]

@ThatCoffeeGuy not even just loss of meta data. The solution provided now in the wiki actually doesn't work. At least I can't control what version of the client my users are running, while they share data with each other, so no matter what I can't get this to work in a correct way. No user can update to latest client when that one doesn't sync, but in order to fix the issue everyone needs to be upgraded. Chicken and egg problem.

[shibacomputer]

Can we please re-open this issue? it appears as though Nextcloud leadership is doubling down on another half-finished feature.
It's hard not to see this as vision-less and reactionary. Nextcloud inserting itself into the complexities of social media while suffering from multiple, serious data-loss bugs is extraordinarily concerning. I appeal to Nextcloud's leadership to divert the resources put into the semi-abandoned Social app, and instead solve some of the issues listed in this issue.

[shibacomputer]

Update: now it's Nextcloud leadership's fixation with so-called Ethical AI. My protest achieved nothing and the worst of the issues still exist. I despair.