Developer perspective

[Matus]

Here is my perspective as a developer of ClearFlask (open-source Fider alternative) and I assume others have the same take as well:

• I don't see the right incentives for me to contribute to Cloudron
• I see Cloudron as a for-profit, open-core (at best) organization making money from repackaged open-source projects integrated into Cloudron by volunteer developers.
• I'd prefer to make my own project easier to deploy, rather than helping Cloudron
• At the same time I am not hindering Cloudron, I stated I would help and support anyone wanting to put ClearFlask on Cloudron. I licensed ClearFlask under Apache license and Cloudron is more than welcome to use it, including repackaging and re-selling.

This is an unfortunate situation because:

• I can package ClearFlask for Cloudron much quicker and be able to provide ongoing support much better than any other volunteer

My suggestions:

• Open-source Cloudron
• or, pay developers to help you from the funds you receive

A crazy scheme I just came up with:
Take a percentage of Cloudrons' sales and re-distribute back to either:
a) existing apps on Cloudron based on popualarity
b) or as a bounty for apps not already on Cloudron. For best quality, add preference for the original developers of the project to bid on the bounty and provide support.

[marcusquinn]

@Matus All fair comments. It's certainly a problem, bottleneck and really now becomes a gatekeeping issue that comes up more than any other.

I also had some thoughts on a possible solution:

• https://forum.cloudron.io/topic/7061/why-cloudron-s-docker-only-how-about-vm-containers-with-generic-docker-compose-scripts-s

Ultimately, no-one can wait indefinitely on another, so attention will spread to alternatives, as people will have no choice but to seek lower barrier to entry methods of self-hosting new apps.

Often we don't need all the Cloudron convenience features, since VPSs offer backup snapshots, email is easy enough to setup, we've moved to preferring Univention for Active Directory & LDAP, we fire up small separate VPSs for other apps and use Cockpit to oversee admin, Portainer keeps calling as a quick and scaleable alternative.

Having a business model based on generating income from open-source software, but not also being open-source software is certainly going to remain a challenge for securing any voluntary contributions from those in exactly your position of having donated all your work to the community already, specifically for non-proprietary setups.

[timconsidine]

@Matus interesting ideas but not for me to comment on really (it's for @staff).
Having said that ....

• I don't care that Cloudron is not 100% open and free.
  I pay for other software/services, why not for Cloudron also.

• I would just like to say the root of the problem is that Cloudron users love Cloudron so much that they want every available app to be packaged !
  OK, gross exaggeration but the principle remains true.
  There's too many good apps for this issue ever to be "solved".
  Realism and demand expectation needs to be considered.

• It's not viable in practical terms for @staff to be expert enough on all apps to package them quickly.

• And every new app needs to be maintained with Cloudron updates.
  We all tend to forget this. Initial packaging takes time, but a resource allocation needs to be made also for updates and support.
  Of course, the original app developer does the core of this, but we must remember it doesn't stop there.

• My solution to the Cloudron user "hunger problem" (not the core issue of course) is simple.
  Get a separate VPS and install app with Docker / DockerCompose until it becomes available on Cloudron.
  Having a 2nd VPS does not invalidate the relevance, usefulness or cost-effectiveness of Cloudron.
  Yeah, I know, it's not a real solution.
  But it does get you functional with that app you just live without.

As always, just my 2p.

I'm interested in Fider but interested also to take a look at Clearflask, and investigate a self-package. May have some questions for you. But time is scarce for next 2-4 weeks.

[jdaviescoates]

@Matus said in Fider:

Open-source Cloudron

For a while it was fully open source. I agree that it would make more sense for it to be fully open source again and so far I've yet to hear any good reasons from @staff or others about why it isn't, see:

https://forum.cloudron.io/topic/2862/why-not-make-cloudron-fully-open-source-again/

The only reason I can fathom is a fear that making Cloudron open source again would somehow mess up their business model due to the potential for copycats to offer the same service for cheaper.

Personally I don't see that fear as having that much risk as Cloudron has a strong community and that is something much much harder to simply copy.

But also, as you mention yourself, I think there are a bunch of people out there who don't contribute to Cloudron because it's not open source, and a whole load of people who don't subscribe because it isn't too.

@Matus said in Fider:

A crazy scheme I just came up with:
Take a percentage of Cloudrons' sales and re-distribute back to either:
a) existing apps on Cloudron based on popualarity
b) or as a bounty for apps not already on Cloudron. For best quality, add preference for the original developers of the project to bid on the bounty and provide support.

Interesting proposal. Cloudron does already support some upstream projects with both code and money, but I'd love to see it do so systematically!

I'd also love to see more transparency about the finances, how much is coming in and going out and on what. Buffer are a really great example of this, see https://buffer.com/transparency