Likewise, the android application is so bad the best way is to open the web UI on your phone to do stuff (and sometimes, it's the only way to achieve some goals).
There is definitely a lot of room for improvement for easier adoption, yet, this is one of the few sync app that work for me, even behind NAT, with closed ports, and so on.
The second thing I like is that it is a successful FOSS project, meaning it works and is being regularly and often improved by its users.
Your comment, as helpful as you may feel it is, is contradictory to my second - very strong - feeling.
I really think to UI works quite good, and by god, if a better solution is so obvious to you, why haven't you proposed it yet, or - god forbid - made a merge/pull request??
Because there isn't any obligation to? Why do you think you would be entitled to ask for that?
I think freedom of speech allows for it.
Can I have a million pounds, from you personally, please?
Also, that might be the small nudge which potentially has a user actually contribute in some circumstances.
I'm not sure your question is fair.
We're just here having a conversation. The expectation that a user MUST do something for an open source project because they have an issue is a tired take. It's not as though Reventlov is blowing up Syncthing's development team with demands about how to make it better. And, for all any of us know, Reventlov may have already submitted PRs or opened issues for Syncthing. Or, it could be the case that there's a different piece of software suiting their needs better, and that's fine. Or, Reventlov maybe just doesn't possess the technical skill to contribute in that way. And that's okay too.
Bludgeoning people with an open source virtue-stick for sharing an opinion isn't helpful.
But just as much as you consider an "expectation that a user MUST do something for an open source project" is a tired point, I consider the endless ranting by entitled FOSS users an obnoxious trend.
And btw. I never said something about "must", I asked why not choosing - OK, maybe strongly suggesting - another option.
Maybe you took my question the wrong way, but (maybe just as much misguided) I took the "opinion"/"discussion" not as opinion, but for the lack of any helpful suggestion as merely a rant.
I was raised not to needlessly complain about free things, without considering to take things into my own hands.
And even if you don't know how to code, suggestions and discussions are better had in an issue tracker, right? Otherwise, what's the point?
If one is not a user, and considers something else to be better, why voice such a strong opinion?
There is a massive difference between voicing your opinion about something in a public discussion forum and hounding the developers of a project because they don't fix bugs or implement new features on your say-so. One of those is making conversation, the other is entitlement.
> Maybe you took my question the wrong way, but (maybe just as much misguided) I took the "opinion"/"discussion" not as opinion, but for the lack of any helpful suggestion as merely a rant.
Sometimes, you can recognize that something is bad without knowing the best way to fix it.
> I was raised not to needlessly complain about free things, without considering to take things into my own hands.
Okay, so that was how you were raised and how you operate. I don't see the reply as needless complaining. It is a critique of some specific issues. It was constructive criticism, because it presented a specific set of things that could be improved upon. I happen to agree with those criticisms, even though I generally think Syncthing is a great piece of software.
> And even if you don't know how to code, suggestions and discussions are better had in an issue tracker, right? Otherwise, what's the point?
What's the point that any of us are here talking about anything?
> If one is not a user, and considers something else to be better, why voice such a strong opinion?
Because people make conversation and have opinions. Are you familiar with socializing? It's okay to not like something. It's okay to not like _parts_ of something.
How would you make things clearer?
NextCloud integrates into iOS's Files app pretty tightly as a 'storage provider' or whatever Apple calls it. End result is that my password database is magically up-to-date whenever I go to use it in my password manager, probably because Files sees an app trying to access the file, pings NextCloud to say "yo, is this shit up to date?", and then my password manager opens the file. I don't have to worry myself about background sync and such.
I didn't have to do jack shit to open the database the first time; the built-in iOS file-picker that came up let me select nextcloud as a source, and then my password database. Done. It is two taps to get that database open now - one to hit "passwords" in the keyboard area, and the second is me TouchID'ing to unlock the database.
Looking over the Mobius faq, it appears to be vastly inferior, with no guarantee your files are synced at any point in time, and you have to manually push files from Mobius to Files, and then access them in apps from there:
> iOS apps cannot access each others’ files. This means you will need to copy files in and out of Syncthing using the Apple Files app.
I wouldn't do that even if the software was free. And they want me to pay for that? No?
A better way should be creating a whitelist of apps that can access the storage of other apps. Gpx tracker app? Add Syncthing to its whitelist, and the file manager I use instead of the system one.
The difference is that instead of having an ever-running daemon, the developer schedules tasks with the system, and system decides when to run them based on network availability, battery charge, etc as well as the app’s behavior (badly written/inefficient background tasks and frequent high intensity task requests are penalized).
We see a NextCloud dev explaining that background uploads don't work herehttps://github.com/nextcloud/ios/issues/215
And we can see syncthing devs lamenting the many issues with an iOS client here https://github.com/syncthing/syncthing/issues/102
It may be the case that file eventing now works, but a quick check with an iOS dev friend suggests that the filesystem sandboxing is too restrictive to be meaningful anyway.
Further on this issue, consider that a functioning syncthing client is a node in a p2p network, so must be able to advertise and listen to requests in the background as well as the background jobs that NextCloud requires (as NextCloud is centralised it doesn't need node level co-ordination) - so a partially working NextCloud client is good enough, but a partially working Syncthing client is woefully broken.
Why isn't this a lawsuit yet?
It's not like Apple specifically advertised background services, then lied about it.
The device is specifically built like that, and you can buy a different phone if it's that big of an issue.
It feels like complaining your toaster doesn't have a dark enough toast setting that you want. Is there a lawsuit in that?
- I can't change my car's satnav. You can buy another car if it's that big a deal.
I think we can agree, the above statement reads silly.
Sometimes, no matter the deal size - buying a new thing isn't always a reasonable option.
If you really care about replacing parts in your car like the satnav, should you need to do research before buying instead of assuming it works that way?
Or should we have rules that all sat navs need to be replaceable so consumers don't need to do that?
I honestly can't draw the line myself, I just try my best to identify what I want in a product before purchasing. Especially huge purchases like a car.
What does it do if you're not connected? Don't mind it checking whether the local copy is up to date, but I'd be concerned if the sync trigger is you opening the app.
Works flawlessly, is rather cheap, and the developers encourage donations to the main SyncThing project.
I usually sync PDF and epub files to my iPad and open them directly from the sync folder in Apple Books.
Here's that fork: https://github.com/Catfriend1/syncthing-android
You can also search up Syncthing-Fork on Google Play.
Now I use Minio with FolderSync (Android App, I use the paid one, but the free is perfectly capable) to backup my camera roll and I wrote a very simply WebDAV server in Go to backup my (Android) password manager DB which only supports WebDAV, I sit NGINX in front of them both to terminate TLS and to handle basic auth for the WebDAV server, though I could easily implement basic auth in that too (again, my password manager only supports basic auth).
If someone can suggest a better password manager for Android with a good UI, good integration (e.g. auto-fill), has a Linux desktop app and can backup to WebDAV or Minio I'd jump in a heartbeat.
I considered Bitwarden, but I don't want to run the Mono/Windows server container, and I don't want to rely on the Rust port which is behind in features and is susceptible to the upstream breaking APIs.
Minio is zero-maintenance, if it needs an upgrade I just pull a newer Docker image. My WebDAV server is fewer than 100 lines of Go and extremely easy to maintain.
I use Nextcloud on my phone (Android) to backup my camera too, the automatic sync is wonderful and works well. It does stuff I don't need too, but it's pretty easy to use and setup and it integrates well with the OS (Windows and Mac at least).
"The server project is written in C# using .NET Core with ASP.NET Core." https://github.com/bitwarden/server
Seems like alternative DB providers is in alpha stage right now.
Mono and netcore is not the same, netcore is one of the more wonderful things I've worked with if I'm to be completely honest. Then again, I also kinda like PowerShell
My fix it to run `touch .stfolder` in the build directory which fixes it but seriously Syncthing? Just put in some reasonable defaults and call it a day.
Otherwise it's perfectly fine.
then you _really_ want syncthing to stand back and hold its hands up, instead of syncing deletion events to all connected devices or sending all your friends your entire home directory or something like this.
imo, it's a good safety tradeoff, and like so many of those (e.g. short vs long passwords) they might seem annoying most of the time, but prevent incredible damage that _one_ other time.
For whatever reason open source software always has these weird edge case engineering "solutions" that really aren't that great. If someone was actually a paying customer and asked for this the engineers would just figure something out instead of making excuses for why it is the way it is.
Maybe it's possible to .. not do that?
o_O weird take but k
Seems kind of obvious to my eyes, asking, talking, discussing about some change before doing the actual work. But as it seems, "obviousness" is in the eye of the beholder...
I assume you don't care about missing out on updates since that's the free work you're complaining about
That's really bizarre behavior. If the folder has no files or zero byte files then it might make sense, but if gradle's deleting non-empty files that don't concern it, that seems to be more of a problem with gradle.
I get the point though; if there's no versioning, why bother having an empty folder?
The web UI is perhaps the only pain point of syncthing. It's very confusing around folder paths VS IDs, especially around the "Default Folder".
Functions like Actions > Show ID are presented poorly.
Importantly, the UI is not intuitive at all. You need to read the docs to understand what you are supposed to do.
I used to use it on Android to keep my 'camera roll' synced to my desktop. I preferred setting the ignore file to exclude android's thumbnails directory, but otherwise it worked great. I'd take a photo with my phone and almost before I could open the shortcut on the desktop, boom, the photo was there, since the android client detects changes in the filesystem.
I also used it to sync my password database, and a cross-platform notebook app's database. Worked flawlessly for all of them.
Ditto for a folder shared between my MacOS laptop and Windows desktop.
Lots of controls, especially in the forked android client. Don't want the sync client to run on any network other than your home network? Done. Don't want your syncthing clients to try and do NAT transversal / discovery outside whatever network it's on? No problem.
The biggest bummer is that there's no iOS client. Had to switch to Nextcloud, which so far has been working OK, but the number of people that have problems with the iOS client is quite high (problems such as syncs being so slow it takes days to sync a camera roll), and it's generally sluggish and doesn't work in the background (yes, I know, Apple's fault.) However, Nextcloud does allow me to sync contacts and calendars (I think. I haven't tried to set it up yet.)
One warning: Sycnthing really does not like it when you delete the dotfile folder inside the shared folder(s). Don't do that :)