Nextcloud: the most active open source file sync and share project!*

[system]

Originally published at: Nextcloud: the most active open source file sync and share project!* - Nextcloud

Nextcloud has been making steady progress, with our Windows Mobile client and the CalDAV Outlook Synchronizer merely the last in a long list of great things coming from our community. Nextcloud 11 Beta is ready for testing, introducing many great features, showing how active our development is.

Statistics

Looking at Github Pulse of last week above you can see that in the 7 days before we wrote this blog, Nextcloud merged an impressive 239 commits to the server code*! To not just look at numbers in a vacuum, we compare with other open source file sync and share solutions:

Nextcloud, Pydio, Seafile and ownCloud

Nextcloud has 5 developers with over 50 commits last month, 2 even over 100. Those who know the company will see four of these five are employees of Nextcloud GmbH, the fifth being the super-active community contributor John Molakvoæ. With 17 engineers on the payroll, the company employs 9 of the top ten contributors in the last half year but it isn’t all about one month, so let’s look a bit further.

Commit graph

Here's a graph of the development the last year, this time with the other projects below.

Top is Nextcloud, below are ownCloud, Pydio and Seafile. The last two were resized to scale.

Nextcloud had a super active week last week, as we had a hackweek just before feature freeze! You can see that Pydio has a rather cyclical development with periods of greater activity followed by less activity. According to the Pydio team (who we’ve asked) their contribution cycles are “not really a pre-established sprint, but more a natural “new major features > bugfixes > community feedback > specs for next features” cycle”. We would guess that the activity pattern in the other two projects probably has to do with the forks (EDIT: see the comments below from a Seafile developer - the German part of the fork did not contribute to the code so it made no difference; moreover, the statistics are skewed against seafile due to how they work).

OpenHub

There is a caveat with the statistics from the Server repository - a lot of work is happening in other repositories in the Nextcloud project. openHub shows statistics from an entire project and while not entirely accurate either (it is hard to keep track of all the repositories being added all the time to Nextcloud) it shows we're doing really well! Here are the commit (until November 15) and contributor graphs (until October). You see the trends visible on github here, too - Nextcloud growing fast since its inception, with a great boost in the last 3 months.

Nextcloud, ownCloud, Pydio and Seafile with a 5 year graph.

We’re especially happy with this spike in new contributors:

All this is of course entirely thanks to our awesome community of contributors, users and customers who have been active and involved in making Nextcloud a super secure, easy to use way to sync and share files, contacts, calendars and more. It shows what an open, collaborative project can do! Of course, it helps that Nextcloud is supported by a healthy business dedicated to real, open source, transparent development. Want to get involved? See what you can do to make a difference!

* All according to statistics on Github and OpenHub. All stats screenshotted on November 21 or 22, links there so you can compare for yourself. Note that not all commits are equal. Some change one line, others introduce whole features! Equally, pull requests can add a big feature or fix a one-line bug, or just backport the same code to an earlier release to fix a bug. In general, don’t take statistics as gospel but compare multiple numbers yourself to get an idea of what is really going on. Also, of course all trademarks belong to their respective owners and being mentioned here means in no way an endorsement for us of them or from them of us!

[sudhirkhanger]

Not sure if 9 of the top ten contributors in the last half year is suppose to be a link but seems to be broken on the blog.

PS:- It’s awesome to see you hire developers to do open source work.

[jonathan_xu]

Hi,

I’m a developer from Seafile. I’m totally alright about you writing about Nextcloud’s active development. But your subjective attitude towards Seafile’s development looks offensive to us.

First of all, you haven’t look at all the Seafile projects. The seafile syncing client and server core are quite stable and just need regular maintenance. So they’re less active. But the web interface (https://github.com/haiwen/seahub) is much more active.

Second, having less commits doesn’t mean the software is worse or is dying. In the seafile syncing client and server core, we usually combines a large change set into a single commit. So the commit number may look smaller.

Third, Seafile project’s development is not affected by disengaging Seafile GmbH. They’re actually not developers, but just a business partner. We have recently released Seafile 6.0 and the innovative Drive client. Do you think a dying project can do all these?

– Jonathan

[nickvergessen]

@jonathan_xu

First of all, you haven’t look at all the Seafile projects.

Same applies for all products of all other projects. We didn’t look at our apps either, even not at those which are bundled and enabled by default.

Regarding your second point, see the disclaimer at the end of the blog post:

@system

Note that not all commits are equal. Some change one line, others introduce whole features! Equally, pull requests can add a big feature or fix a one-line bug, or just backport the same code to an earlier release to fix a bug.

I don’t think there was any offense intended towards you guys.
But that are just my 2cents from me as a little dev.

[tflidd]

It would be interesting to see the comparison to closed-source sync solutions.

Besides of these stats, it’s way better to get the latest beta version and check out the latest features if they are convincing. It wouldn’t help to have hundreds of buggy features.

[jan]

@jonathan_xu

we usually combines a large change set into a single commit

@jonathan_xu I would strongly advise against doing this. It’s what pseudo-open-source projects do to »release« their code all at once to be able to say »we’re open source«. Open source means being open the whole time, not only when something is ready to release.

[jonathan_xu]

@nickvergessen

Same applies for all products of all other projects. We didn’t look at our apps either, even not at those which are bundled and enabled by default.

Yes, but the web interface is certainly part of the server. Seafile’s project organization is different from NextCloud. The server has C part and Python part.[quote=“nickvergessen, post:4, topic:5600”]
I don’t think there was any offense intended towards you guys.
[/quote]

@system

We would guess that the activity pattern in the other two projects probably has to do with the forks.

I think this sentence is very subjective and better be removed. Since you don’t have enough information about others internal status, it’s not good to make judgement. If it’s just a personal blog I would say it’s somehow okay. But it’s your official blog.

[jonathan_xu]

@jan

I would strongly advise against doing this. It’s what pseudo-open-source projects do to »release« their code all at once to be able to say »we’re open source«. Open source means being open the whole time, not only when something is ready to release.

I’d better make myself clearer. We’re not working in that pseudo-open-source way. What I meant is that some features are complex and can take long to finish, some of our developers are more used to committing that to master branch after it’s done. That explains why the commits are not frequent.

[jospoortvliet]

Hi Jonathan,

I wrote the article. I did in no way mean to offend other projects, just wanted to highlight what statistics show. And yes, as I pointed out (multiple times) statistics are very flawed. So let me respond to your points (and I will edit the blog to address your feedback):

1. I tried to compare the core of each project, but you’re right that that doesn’t paint an accurate picture. That is why I also used OpenHub, which takes all related repositories into account.
2. Combined with 1, that does lead to an under-representation of the work you do, you are right - while Github mentions lines of code, OpenHub looks at commits and contributors.
3. I did look for Seafile GmbH development efforts and could indeed not find any but heard today that they do have developers. I can’t confirm either way so I will just point to what you said about this.

Again, thanks for commenting! While we’re at it, perhaps Seafile is interested in working with us and Pydio in implementing the Federated Cloud Sharing protocol? That would rock. If you are, shoot me an email on jos @ the nextcloud.com server.

[jospoortvliet]

@tflidd

It would be interesting to see the comparison to closed-source sync solutions.[/quote]

Absolutely. This is actually an advantage of open source: you can see if you’re perhaps betting on the wrong horse. With closed source, it is easier to make unsubstantiated claims about upcoming features or developer resources.

[quote=“tflidd, post:5, topic:5600, full:true”]Besides of these stats, it’s way better to get the latest beta version and check out the latest features if they are convincing. It wouldn’t help to have hundreds of buggy features.

Indeed, a thing not shown is how capable and experienced developers are. Of course most of our team members have been working on this code base for 3-5 years so we’re a very experienced team, this isn’t the same for all other projects.