What is Nextcloud trying to be?

Hi all. I’m new to Nextcloud, just put together a server with NextcloudPi. Had a few rocky starts but now it’s working swimmingly.

I came to Nextcloud, I presume like many, because Dropbox has become a worse product with every iteration. The latest version is almost — well, no almost, absolutely — user-hostile. “All I want is a folder that syncs.”

So after getting into Nextcloud, I’m rather surprised to find so much of the development focused on apps. There are already countless great word processors, image browsers, IM apps… Why is Nextcloud chasing the same strategy that ruined Dropbox?

Also, unless I’ve been looking in the wrong places, it appears that Nextcloud isn’t even done with the “simple folder that syncs” part. Is there truly no client for simple headless servers? I’ve found some articles about running nextcloudcmd on a cron script, that solution seems hacky as hell. Dropbox, for all its faults, had a great daemon.

I don’t get what you want in a headless server.

The sync that the nextcloud client does, doesn’t satisfy you ?

I want to spin up a bunch of Ubuntu web servers that mirror a bunch of folders of static HTML content on my Nextcloud.

I’ve been doing this with Dropbox for close to a year now and it’s mostly magical. The Dropbox daemon is really lightweight and “just works.”

It sounds like I need to install the entire Nextcloud client on a server that doesn’t even have a GUI, and then install a special script that runs NextcloudCmd but doesn’t collide with runs that are already executing… I don’t quite get why the Nextcloud is making RSS readers when it doesn’t even have this basic stuff nailed down.

(that is, unless I’m a dunce and there’s a daemon I can’t find somewhere. There was some chat about it in 2016 but it didn’t seem to go anywhere.)

((here’s that script I found that seems to accomplish the collision detection: https://github.com/qcif/nextcloud-sync-cron ))

Yes the nextcloud client is for a desktop aim for the moment.
If you don’t have a GUI environment on your server it’s like you get screwed because it’s not yet optimal.
If i were you and i still want nextcloud, i would open a secure webdav connection to map onto your server. And then use a rsync daemon for syncing your folders.

It’s not a nextcloud anymore but you can use the nextcloud server for your calendars, contacts, other desktop where you have to sync folders

When comparing Nextcloud apps with Dropbox features it is fair to mention that those are apps that can be disabled (most not even pre-installed) and removed. But many users use many of those apps with makes Nextcloud great extendable for a wider use case beyond of what Dropbox is capable of. For simple sync only, a network share (NFS/Samba) or File Browser or Syncthing with web UI are probably better fitting.

Any headless webdav client can otherwise be used as sync client. The Nextcloud client is a desktop client, thus is expected to be used on desktop systems.

To answer the initial question: Nextcloud indeed tries not only to be a Dropbox replacement but allows to satisfy a wider use case, including calendar and contact storage and sharing, event and project management, collaborative work and video chats in a modular way via optional apps.


What strategy would that be? Dropbox doesn’t have apps at all.

“Simple folder that syncs” and “client for simple headless servers” are far from the same thing… I would say headless servers are atypical usage for Dropbox.

Dunno what to tell you. Glad to hear that you’ve been outside Dropbox long enough to see what a cock-up they’ve made out of it.

Atypical or not, Dropbox has had it for years, and it works splendidly.

I’d like to have a better desktop client development, there are a few issues and some new features don’t even work properly and new features were already announced.

On a server, I’d perhaps use directly webdav to mount a network drive and then use rsync or something to update. Or use directly davfs with caching (not sure if that is really stable):

1 Like

It’s better to be conservative with Nextcloud as the apps that are developed usually don’t continue to be supported. Nextcloud does what Google does, tries a bunch of new things, but they do it publicly (ie. you know when people have stopped working on it.)

Unless there’s a community around the extension, or you’re willing to support it, just think of it as someone pumping up an app to try to get momentum for a side-project.

Nextcloud is primarily focused on enterprise users; they don’t even have a business model around making money from consumers (none of the hosting providers listed on their site pay Nextcloud for development for example.) There’s also no way to donate. They have some major clients (the German federal government, some universities, corporations, etc.), but as part of that business model, they 1) don’t go out of their way to make things easy for consumers because, why, and 2) if they made it too easy, enterprises wouldn’t need them, they could spin and maintain it themselves. It’s actually a big debate currently happening in the FOSS world, how to make money sustainably.

If you’re looking for something really tight and focused just on file-sharing, check out Seafile… but their transparency and development process is lacklustre IMO. For my personal use, Nextcloud hosts and syncs my files without data mining, and it syncs well enough. I don’t bother with most of the extensions.


I want to spin up a bunch of Ubuntu web servers that mirror a bunch of folders of static HTML content on my Nextcloud.

you do not need nextcloud for that - any network-enabled server will do (even a shitty CIFS-server (aka “Microsoft Windoze”) or FTP-server will do). ssh/rsync might be a nice solution.
if you do not want to run it yourself, i recommend rsync.net .

Actually, it’s the community that releases a lot of apps. If they find many users and manage to build a community, they can live rather long. Some of them become officially supported apps (contacts, calendar, …).

Yes for their official support. They did a few things to improve Nextcloud for the community which is less interesting for enterprise users (e.g. web-updater).
There is bountysource to give a bit of money for community developers. There have been discussions for a kickstarter or similar, but there would be a lot of administrative overhead.

There are a few enterprise features that are not documented very well. In theory, the community could do it and share experiences but I haven’t seen many companies that use Nextcloud for free that they share such information or contribute to documentation.

For me NextCloud is very different to solutions like DropBox (which I also find more and more unpleasant to use).

  1. It is your own server to run, adapt and configure the way you want it - if you don’t want any apps then don’t install them.
  2. They use open standards wherever it makes sense (and publish any original APIs). Nothing is hidden, needs to be reverse engineered, accessed via a binary blob API, or will change unexpectedly.

Point 2 is essential because Nextcloud is entirely WebDAV compatible, and the NextCloud client is just one of many that can be used to access a NextCloud server, it is a desktop focussed client, but others aren’t. You can mount WebDAV shares on your Ubuntu boxes and then use rsync as someone else suggested, or use a cloud sync client like rclone which is intended to do what I think you want.

Actually Nextcloud apps are very well developed, at least the ones I use, most of them indeed not by payed Nextcloud GmbH developers but by the community. Having hired developers focus on the business side and the business model is just a total common strategy for open source projects. Free users benefit however as well from that, in terms of security, stability and core features. That is IMO much more attractive than any closed source project that either pays out through advertising, user data usage or that is not free of cost.

For free home (or small office) users there are plenty of official and 3rd party/community appliance solutions that allow a very easy setup of a Nextcloud server, NextcloudPi, Docker, Snap, VM or DietPi (which I develope) and many more. The official docs provide all general info that allow any bid more experienced to set it up oneself and there are guides around, even from official NC developers, as well.

So AFAIK there is really no need to make it “too easy” to setup from official dev side, especially since any more automated/easy solution limits the use cases (network setup, business-related needs, range of instance user counts, …), what is a no-go for any advanced users or businesses. Offering payed customer setup and support is thus not a reason to make things unnecessary complicated, but setting up a server system/network around Nextcloud for business and large scale demand IS simply complicated, not because of the Nextcloud software, but the webserver/balancer/PHP/network/OS/database/caching etc etc setup around it.

EDIT: The main thing that is probably misunderstood when reading the criticism is that with Nextcloud you administer a whole own server where your data is stored under your control, while Dropbox is a client only. And an own server is simply a totally different thing!


Why don’t you simply mount your nextcloud instance as a directory with davfs2, then sync your files in there? No need for a specific Nextcloud client.