Hacker News Reader

Comment by Etheryte | Hacker News Reader

At the end of the day, the money has to come from somewhere. There's only a handful of companies that make the open source profit game work and last I checked, most of them make their money from B2B deals with support, SLAs, etc. If you want to deploy something at nation scale, surely you can afford to pay for it.

View on HN

While it's absolutely true that businesses need to make money, the idea that this must therefore logically conclude with a FLOSS and non-FLOSS version is not correct.

If you look at companies like Red Hat, and Nextcloud, neither of these companies sold proprietary versions of their software. Red Hat sold licensed versions of their software, but they were always FLOSS.

Nextcloud's CEO talks about how selling a proprietary version of your code is bad for your customers and puts your company into a Catch 22 with itself. "Do we add this feature to the 'Community' or 'Enterprise' version?" or worse "What do we do when someone makes a FLOSS plugin that competes or is better than our internal version of the same functionality?"

There's a huge need for a good chat system to compete with the likes of Slack, but Matrix has bitten me several times and I don't trust it. In an alternate universe, they could have made Matrix work well for all levels of users and then sold customization and B2B deals with their customers leveraging the community, which has always wanted Matrix to succeed.

Would you be able to point me to the talk or blogpost of nextcloud that talks about their business model in detail? I would be very interested in how it works in practice, how they differentiate with their enterprise versions and so on. I agree that their business model sounds like having the best of both worlds, so any material on the details that you could share would help me a lot figuring out how to replicate it for other solutions.

Last I checked they ratelimit push notifications using a proprietary push gateway, unless you’re a paying customer.

The feature exist in both versions but you pay for additional capacity which is not what they are talking about when they said deciding whether the feature should be in community or enterprise.

If you look up Frank Karlitschek and "I forked my own project and my own company", he's given this talk many, many times.

As for it being the best of both worlds, it depends what best means.

A company that wants to make a lot of money won't do well with this model. It's a slow growth model requiring a lot of hands on work that requires customers who are paying for features and custom installations.

It's labor intensive work that's quite the opposite of the proprietary software model, which has traditionally had a lot in common with the publishing or entertainment industry. Write a book, get paid for it for years.

> There's a huge need for a good chat system to compete with the likes of Slack, but Matrix has bitten me several times and I don't trust it.

I thought this was Zulip?

Zulip is excellent for geeks. I don't think it's for the broad public. We use it in our company with great success, but our marketing people still have not understood how topics are used (both by their own comments and by the evidence I see regularly).

That's a shame! What sort of problems arise?

Topics only help to keep things organized and better than other chat systems if they are reusable and long-term identifiable. If people create a new topic like e.g. "Wednesday meeting" or "The printer does not work" every time they want to say something you end up with same mess as in other chat systems. Also it's quite human that discussions get offtopic. Zulip has nice support to continue in or move the discussion to the right place. But to my experience only a couple of geeks uses them.

RedHat sold to IBM and neutered it's free CentOS variant recently, so not so sure it's a good example.

CentOS is no longer a drop in replacement for RHEL, but both are still entirely open source other than trademarked logos and the like AFAIK.

More importantly, Red Hat was very profitable even when CentOS was a drop in replacement.

I think the only reason Red Hat’s model works for them is that Linux distributions are obviously massive sprawling complex things, and maintaining LTS distros which you then sell support for is an incredibly complex and valuable thing that no intermediary can plausibly claim they can do. Even if the underlying code happens to be FOSS.

Matrix servers (and web servers and mail servers etc) are not as complex - and if anything you want incentives to make them less complex, not more. And the better job you do of making them easy to admin and support (ie the more scalable and stable they are), the less reason there is to pay the upstream for support but some system integrator instead.

> Linux distributions are obviously massive sprawling complex things

Just my perspective but I think this was also perhaps a product of the time when Red Hat really started growing, back when linux was a little more rough around the edges. These days it seems linux has enough polish that some (many?) corporations are happy to just use off the shelf ubuntu and call it a day and hope nothing breaks too badly.

I think another thing that has changed is that big business is less suspicious of open source and are less likely to feel that all the software they use has to be licensed from someone.

I think it has dawned on people that it is impossible to avoid using open source.

This is definitely true. It’s also completely normalised to build a large commercial service on open source and not even contemplate if or how the upstream project is funded, or consider funding the upstream a serious cost.

[0] has many more details but CentOS was never a drop in replacement and the new CentOS model is a big improvement for anyone wanting drop in replacements. I agree with everything else you said though, they were independent of IBM for nearly 3 decades and always were and still remain among the most active linux contributors (just see any of LWN's development statistics articles).

0: https://medium.com/@gordon.messmer/in-favor-of-centos-stream...

They still don't sell a proprietary version of their software.

CentOS started as a fork, and new forks exist now.

Reply on news.ycombinator.com