Comment by nisa | Hacker News Reader

Comment by nisa | Hacker News Reader

As an admin of a 500 user instance I'm having mixed feelings about Nextcloud. Sure it's GPL and free and it got a lot better than the old owncloud mess that broke always but it's still somewhat a gamble. My install is just a no-warranty, non-commercial use case and I wouldn't use it for any large commercial installation without having at least one dedicated and skilled PHP-dev on-site.

I came to resent their marketing blog posts. If you look it's just an alpha version that just got hacked together, often with problems, often not fixed long after the release. There still lot's of serious bugs in end to end encryption for instance.

There does not seem to be a lot of work to improve the base features. Syncing, ACLs, User Mangement leave a lot to be desired. It's not easy but they are in a position where they really could be a perfect solution for a lot of people and probably also charge good money for it - but at the moment it feels like they just build alpha-grade fancy addons.

Me and probably most users would love a no-feature, we just fix the most serious bugs on our github issues release-cycle. I'm fine when it's called Nextcloud 16 and takes a year, hell take your time, but basic functionality should be rock solid.

I also don't get why they ignore the existing open-source ecosystem. Make it possible to integrate SOGo for Calenders and Mail - it's all webdav in the browser anyway. SOGo at least has somewhat working ActiveSync. Same with SAML/Shibboleth/Kerberos - some things got fixed, very late. No deployment story for Active Directory or even Linux.

User management is a mess - the Circles app is a good idea that is crippled by usability and functionality bugs... that possibility to have self-selecting groups and decentralized management is a huge win for them, but they don't seem to realize that this is an important use-case.

Nextcloud is very close to what would be ideal for a lot of smaller organisations, but they somehow are unable to polish what they have but decided to pile alpha-quality on alpha-quality code (and ignore the bug reports)...

That beeing said, it mostly works but your users won't be happy.

Let's be honest, if we did a release with only stabilization, it might get us more users. Not more customers. And it is customers who pay for the development.

Everyone asks us to focus "on the basics" and then claim their pet feature requests are, of course, basics. ACL's and user mgmt in your case, perfectly fine features, but features nonetheless.

In the end, our development is guided by the community and customers. If you're not either, we might not do what you want... That's the reality of any open source project.

And you're right, for a large installation, you should obviously get an onsite engineer, or, much better, a contract with us which is both cheaper and better. That's what most larger installations do, and that is how we earn the money to make Nextcloud better.

WRT the E2E, that sadly took a whole lot longer to get working than we had expected. Of course, if you think it is important or need it, time and money can speed it up, we have only so much of those.

Why do you say that a release with only stabilization brings you more users but not customers? Wouldn't the quality word-of-mouth convince more businesses to switch? And I assume they need professional support anyhow, don't they?

Another Nextcloud enthusiast and (very small) contributor here. Thank you for the awesome work you and the team do :)

GitLab seems to suffer from a similar conundrum, and they also seem to have gained a reputation for being unstable. Which is sad. In my experience, GitLab is very stable, nowadays, but it will probably never fully recover from its first impressions for some people.

Thanks for your comment, we really appreciate your feedback about GitLab. Our team is working hard on improving the availability and stability of the platform. Our goal is currently to achieve 99.95% availability on GitLab.com. According to Pingdom, over the last year our availability has been 99.81%, although this includes the large (multi-hour) maintenance window on 11 August 2018 when we migrated GitLab.com from Microsoft Azure cloud to Google Cloud Platform (GCP).

Since the migration, our availability has improved greatly (caveat: we obviously have much less data than from Azure).

There are multiple reasons for this improvement. We chose Google Cloud Platform because we believe that they offer the most reliable cloud platform for our workload, particularly as we move towards running GitLab.com in Kubernetes. It is worth pointing out that we also used the migration as an opportunity to improve our infrastructure, simplify some components and otherwise make things more stable and more observable. Finally, we've also been focusing on building the infrastructure team up, having hired many new team members over the past few months. This means that the team has been better able to balance the job of running GitLab.com with making it more stable.

I use Nextcloud and it is an integral part of my life, but your current set up does not allow me to pay you, as it focuses only on enterprise use case.

If I may make a suggestion, could you make a page for donation, and then you can have a "put your donation to Nextcloud to the following uses: - New features - Stablilty - Android Client development"

I think you get the idea. I would very happily donate if I was given the option to.

I like your 'vote with your wallet' idea for open source. Want to get that issue to the top of the ticket queue? Pay up! It makes perfect sense.

I want to say Ubuntu does that?

Its a larger frustration with a lot of open source projects though. I'd love to throw some money to say thanks, but its either really hard or convoluted. The easiest way I have found so far is with android ones, I just buy the paid versions on there.

That's really a nice, sympathetic idea.

The main issue with that is that it simply doesn't provide anywhere near enough money to pay a real team.

Our forums, for example, have about 17.7k users according to help.nextcloud.com/about

Let's say we get ALL those users to pay 10 bucks per year. Not going to happen, of course - but let's say we put in a lot of effort there. That can pay for 3-4 developers. We have 25. Should we really spend time on that? Or spend that energy on getting a few big customers who pay 100K each?

We have explicitly chosen not to even TRY to monetize any installation under 50-100 users.

That means we don't try to squeeze money out of small businesses and home users (for example by not fixing issues for those users to try and get them to sign a contract). Instead, our incentive is to make Nextcloud easy and good enough that anyone <100 users can just use it without needing us. Because then 1. they don't ask too many questions on our forums and github and twitter etc and 2. they advocate for how great Nextcloud is to our real potential customers: businesses with 500 to 50 million users.

Of course, if community members want to donate or if contributors want to earn a cup of coffee, there's always https://www.bountysource.com/teams/nextcloud But our employees aren't allowed to take money out of that, they have to donate it back to issues they/we care about.

Hello!

I understand and very much respect your prespective. If I was in a company that could use your software, I would point them to you.

If you don't mind from a single home user, my prospective was: I really like Nextcloud, I want to at least send some money your way. I know it won't sustain you, but for all your software does for me, I feel like I can donate. I go to your website, I look at pricing, and the lowest pricing I saw was 1900 euro for 50 users. I didn't even know bounty source was an option.

Now that I know bounty source exists, I will send money on that. However, may I suggest to put bountysource on your main website so folks like me can at least support you (or if your employees can't take money out of it, give folks like me a way to give you money)? I get the fact that running on that won't sustain a business, but I would like to think that if users saw that on your website, you would get more money that way.

Or maybe for a thought too, I brought up Android on the previous one. I forget which app I had, but they had a paid and a free version. I looked at the paid version, and it said something along the lines of "this version is no different than the free version. However, if you would like to support us, please consider buying our paid version." Their monaterization was not more intrustive than that. I bought the paid version, because I knew the money would go to the developers.

If perhaps you doing any monaterization is too much if a pain that it isn't worth it (or you don't want to alienate anyone, or perhaps if people pay then you have to support and its too costly), I understand. However, I would like to think you would get some return on investment by some simple non-intrusive ways to monatize your product.

And since I didn't say it before, thank you for the software. I know I can't give enough back to sustain your company, but your software is in my daily workflow and is an amazing product. I have also been responsible for getting several other people to use it on a daily basis, and they depend on it as well.

Thanks for your work on Nextcloud! I really appreciate what it does. Also thanks for taking the time to answer my rant ;)

On a positive note: Just upgraded to Nextcloud 15 without any issues in a few minutes. I feel it's getting better with every release. You put in a lot of work and I recently logged into an Owncloud 8 install and it's a night and day difference.

I'm also aware that you owe me nothing. I'm fine with that. It's a discussion forum here and I just wanted to write down my experience. I'm sure you'd rather see solid pull requests instead of unwanted advice. Mea culpa.

It would be more fair to compare the LATEST Nextcloud with the LATEST ownCloud. Both projects are going strong! Nextcloud is more community-focused, ownCloud is more enterprisey.

Hi Markus,

I'd argue the opposite, actually, ownCloud is betting on providing hosting for small and medium businesses, while Nextcloud targets large businesses.

And that is what reality shows. Nextcloud has far larger customers than ownCloud, which breaks down at instances over 50K users. The only large-ish oC setup is actually a bunch of separate universities.

Nextcloud is far more scalable. Thanks to Global Scale, our largest installation has tens of millions of users on a single instance, and customers like the German government (300K users) wouldn't be able to work with oC unless they split up their users and create silos.

>logged into an Owncloud 8 install

welcome back to the February 2015 :)

Spot on. As an admin the saml situation was a nightmare: it took so long for them to update the saml library, so for a time it wasn't even compatible with their docker images because the library version they were using required mbstring, while many moons ago an updated version of the lib was released upstream without the need for mbstring. I still can't get ss-logout to work either.

We use SOGo as well, and at some point I will get around to integrating SAML into sogo, so at least when users click on the "external sites app" icon for SOGo, they'll be signed in automatically. Either that or add support to use external contacts and calendar dav servers.

I feel like in general there will come a point where PHP's execute and die model won't do. I really wish they'd implement some sort of backend daemon to carry out long running tasks.

My knowledge is a year or two out of date, but I think “it’s all just webdav” is part of the problem, and they sort of painted themselves into a corner by using it as a protocol. They ended up with a bunch of extra API hacks on top of WebDAV for things like multiparty uploads, and things still go wonky if you try to connect a random WebDAV client to it.

I built an ownCloud file storage replacement that worked with their front end client, while it was all technically WebDAV, syncing issues were a nightmare. I have a great amount of respect for how much effort they put into getting WebDAV to work, I just think it’s not not as solid/specified a protocol as one would hope.

The synchronization is a nightmare. The C++ client is hard to figure out, and even harder to maintain as a customized solution.

Worse, they don't get the usual benefit of C++ clients being lean and resource-friendly because they now ship with an embedded QT Webbrowser.

You should try the ownCloud client instead. It works just fine with Nextcloud as server and has less bloat than the Nextcloud-branded client.

https://github.com/owncloud/client/releases/tag/v2.5.1

Disclaimer: I'm one of the developers of the the ownCloud desktop client.

Of course, you lose support for 2FA, have crappy notifications and activity feed and so on, plus it is entirely unsupported so I'd recommend against that...

While there is certainly a point about electron apps wasting resources, an electron app would be a lot more maintainable and the GUI could be improved a lot.

For example it's impossible to know which errors are actually a problem, and if synchronization was successful or aborted. Some of the progress information is in a tooltip which changes a couple of times per second.

I had to deploy Nextcloud at work for a couple dozen users and it was a nightmare. There was zero response to any bug reports / questions I had. I even tried to get commercial support, and they wouldn't answer.

The client is effectively a black box, and we can't even get synchronization to work correctly. Some of it certainly is the fault of me or the users. But some of it probably isn't, and it's hard to tell the difference, what with nobody responding to bug reports.

Hi,

Sorry to hear about the bad experience, that shouldn't happen. We explicitly design Nextcloud to be seamless for small installations like yours - but if you need support, you can simply purchase it directly on https://nextcloud.com/enterprise/buy/ so you don't need to wait for sales people. We'll fix your issues for sure.

Keep in mind: using a bog-standard LAMP stack is your best bet, but you can always add the 1-day installation help to let an expert guide you through the decisions on how to integrate in your infrastructure. We have hundreds of customers with literally tens of millions of users deployed, so by now pretty much every problem has been seen and can be fixed ;-)

Oh, I can totally recommend to my boss forking out at least 2000€ for commercial support by people who take two weeks to respond to any sales requests without knowing if these problems are actually fixable...

Don't use the old ownCloud mess, use the new one. I'm currently running ownCloud 10 and it works great. The stability and architecture of the project have really improved in the last years.

It seems to be scalable too, when you look at the big organizations running ownCloud. All the know-how and feedback of the big instances is flowing into ownCloud, not nextcloud. Speaks for alpha-quality.

I have not tried the migration from nextcloud to ownCloud yet, but it should not be too hard.

ownCloud seems to be going for hosting small and medium sized businesses, I'm not sure what that means for big organizations in the future.

Our customers tend to be bigger, actually. Like the German federal government with 300K users, companies like Siemens, german Radio & TV stations (transferring over 1TB/day!). Our biggest deployment (in a single instance) is over 20 million users - see nextcloud.com/globalscale for how we pull that off. ownCloud has no equivalent solution for such installations and breaks down at about 50K users.

For home users or very small businesses, the difference might not be that big but the migrations we do for customers show time and again that there is a big improvement in performance and reliability, especially on large installations.

... and now I see I probably replied to a sockpuppet account. Sigh sorry for the noise.

Similar situation with LDAP as well. The first login of a new user results in a crash. They haven't fixed this issue in 3 months.

Sorry to hear it broke.

Try logging a support ticket on portal.nextcloud.com. LDAP is an enterprise feature so it is an area where we strongly focus on customers and their issues first, and if none of our customers experience this problem it can take a long time to get resolved as we're always very busy...

My guess he is referring to https://github.com/nextcloud/server/pull/12693, which sits there as a pr for 16 days now (even though it was opened from one of the nextcloud core team afaik).

Reply on news.ycombinator.com