Open core is not a crime

One of the reasons I described the current debate about open core as futile is that there seems to be no hope of it ever reaching conclusion. This is partly because, as Stephen O’Grady notes, the anti-open core brigade have not put forward any potential remedies. Stephen argues that this is because there are no potential remedies. I would go further in arguing that there is, in fact, nothing to remedy.

The vendor controlled open core model sees a vendor offering an open source core project a license approved by the Open Source Initiative, as well as using dual licensing to offer a proprietary version, which also has additional features and functionality, that is not open source.

Simon Phipps has articulated why this strategy does not meet the approval of software freedom advocates, but in doing so, in my opinion, mischaracterises the relationship between vendors with open core strategies and open source.

The description of open core as exploiting loopholes in the system suggests that vendors with open core strategies are deliberately bending the Open Source Definition. However, as we have noted before, the OSD applies only to the license of the underlying code, and does nothing to prevent dual licensing or proprietary extensions.

Simon Phipps has claimed that the “OSI can and will challenge use of “open source” in relation to closed strategies” but so far has offered no response to the obvious question that is prompted by this statement.

Simon also argues that vendors with open core strategies “wrap themselves in the open source flag”. The accusation is that by referring to themselves as “open source vendors” they are misleading world-be users. This is a potentially fair criticism, but it is one that in my experience most vendors with open core strategies have rectified.

As we found with last year’s open core transparency test, most vendors have improved their communication in order to avoid confusion (which is in their own best interests).

Additionally, the core software is open source, and available under an OSI-approved license, and the users retain all the freedoms that are associated with that.

While I fully appreciate why software freedom advocates are uncomfortable with vendors that offer any proprietary software referring themselves as “open source vendors”, I do not believe that there is anything they can do about it for the same reasons noted above. This is why this debate continues ad nauseum. The fact that open core opponents seem strangely unwilling to name and shame the vendors they see as gaming the system doesn’t help matters.

Another of the misconceptions about open core is that something – be it code or freedom – is being taken away from users. It is this assumption that led Henrik Ingo to compare it to theft.

“We don’t approve of stealing, and there are several measures against stealing, in particular laws and criminal punishments. Yet, from this it doesn’t follow that stealing is only a crime if you get caught! So if you put closed source modules into your open source product, and nobody notices, then you’re still not open source.”

UPDATE – As can be seen in the comments below Henrik denies that he was comparing open core to theft, noting that he was “using theft and society’s punishment against it as an analogue, to criticise Mårten’s use of “self adjusting system” as a blanket permission to do anything he wants. But nowhere did I explicitly or implicitly say that open core is comparable to stealing.” He later added, however, that “theft goes against the values of our society, whereas open core goes against the values of the open source community” – UPDATE

This comparison is incorrect on the one hand because open core does not involve putting closed source modules into open source, and on the other because adding closed source extensions to the core does not take anything away from the open source user.

They still have the core. They are still free to run the core, to modify it, to distribute it, and to extend it. They are still free, in fact, to fork it and to replicate the vendor’s closed source extensions. They still have all their software freedoms.

It is true to say, however, that certain features are being witheld from them. The fear is that open core prompts the vendor to produce a deliberately crippled core in order to drive users to its proprietary version, but that would be completely self-defeating.

As we previously noted, “There is no point trying to compel community users to become customers by providing them with substandard software and waving an enterprise version at them. It won’t build a community, and it won’t build brand. That is neither the best, nor the right way to generate revenue from open source.” Some vendors have tried that approach. They didn’t last long.

That all being said, I appreciate why advocates of software freedom are wary of open core. It does perpetuate proprietary software licensing, and it does so via open source. But that does not make it a crime. And a considerable amount of code has been contributed to the commons by vendors with open core strategies. Meanwhile even those that would wish to do something to remedy the situation are without the means to do so. Hence the endless and futile debate.

So what is to be done? I find it somewhat ironic that while some software freedom advocates are demonising open core they are also promoting David Wiley’s excellent recent post calling for more tolerance of others.

With that in mind, I will leave the last word on this matter to David:

“If someone has gone out of their way to waive some of the rights guaranteed them under the law so that they can share their creative works – even if that action is to apply a relatively restrictive CC BY-NC-ND to their content – why aren’t we praising that? Why aren’t we encouraging and cultivating and nurturing that? Why are we instead decreeing from a pretended throne on high, “Your licensing decision has been weighed in the balance, and has been found wanting. You are not deemed worthy.” Why the condescension? Why the closed-mindedness? Why the race to create machinery like definitions that give us the self-assumed authority to tell someone their sharing isn’t good enough? Why isn’t the open crowd more open-minded?”

Tags: , , , , , , , , , , , , , , , , ,

20 comments ↓

#1 Benjamin Reed on 07.02.10 at 10:58 am

I agree it’s not a crime, and they are welcome to pursue their business plan as they like, I just think it is “less” open than a true open-source company, and as someone who works hard for a company that is truly open, it is important to me that users know the difference.

> While I fully appreciate why software freedom advocates are
> uncomfortable with vendors that offer any proprietary
> software referring themselves as “open source vendors”,
> I do not believe that there is anything they can do about
> it for the same reasons noted above. This is why this
> debate continues ad nauseum.

This is where I disagree. The debate that you see as continuing ad nauseum is, from my point of view, customer education. It’s no longer a “debate,” in that I think I’m going to convince an open core company to be open source, but I can let users know there’s a difference between what is at it’s heart a proprietary business model and an actual open source company.

I could care less if some open core company continues to be open core, it’s up to them to have a working business plan. Like most businesses focused on fast growth with a buyout at the end, 90% of them will die on their own under the realities of the VC way of doing business.

Proprietary software is a perfectly legitimate business, and they’re welcome to pursue it. But it is a good and useful thing to continue to point out that open core is just that, even if they are magnanimous enough to provide some open source value as well.

#2 Matthew Aslett on 07.02.10 at 11:26 am

Thanks for the comment. Actually that’s a great point. There is of course nothing wrong with advocating in favour of a 100% open source approach and pointing out the problems associated with the open core model – and there are significant problems that we have previously identified. I have no great love for open core, but I have much less love for campaigning that is anti-open core, rather than pro-open source. I have even less love for campaigning that relies on vague accusations. In other circumstances that would be called FUD.

#3 Andreas Kuckartz on 07.02.10 at 12:06 pm

That obviously is what OSI is considering. Also benchmarking is possible for Open Source projects. Interestingly companies with an open core business models do not like to see that term applied to them: the education already seems to work!

#4 Tarus on 07.02.10 at 11:23 am

Heh, I think you missed the boat on this one, but I’ll try to keep my comments short.

First, “open core” is not a crime, but can you point me to one website where a fauxpen source company refers to themselves as open core? No, they drape themselves in the “open source” flag whenever possible. I’m happy with the market distinction between open source and open core but now we just need to get the market to use it.

You write: It is true to say, however, that certain features are being witheld from them. The fear is that open core prompts the vendor to produce a deliberately crippled core in order to drive users to its proprietary version, but that would be completely self-defeating.

Really? If my business model is based on selling software licenses, why would that be self-defeating? To use an example from my industry, Hyperic chose to put the ability to set, say, a universal threshold, in the enterprise version. So if you used the community version you had to set each and every threshold manually for each instance. Any client of any size would then be forced to buy the enterprise version or pay someone to manually set thousands of thresholds.

But we can sit here and navel gaze all we want. I think the best take I’ve seen on the whole open core debate came from Brian Prentice at Gartner. In his “Emperor’s New Clothes” post he points out that enterprises looking at open core need to treat it like any other commercial software buying decision. He also states that open core does not bring any innovation to the table – unlike pure open source. This comes from a guy outside of open source who advises enterprises on buying decisions, versus those of us who have a vested or personal interest in the distinction. From here on out when anyone asks me about open core I just point them to his link: http://blogs.gartner.com/brian_prentice/2010/03/31/open-core-the-emperors-new-clothes/

#5 Matthew Aslett on 07.02.10 at 11:45 am

Here’s one: http://www.opsera.com/dotAsset/7632.pdf

As for crippleware. The success of the open core model depends on having a ubiquitous core from which a percentage of users will choose to purchase the proprietary extensions. If the core is crippled ubiquity will never be achieved.

Some vendors have achieved this balance. Others haven’t. The difference is in execution, not the strategy itself.

#6 Jack on 07.08.10 at 8:36 pm

Of course the strategy is flawed. The open part of their product is either stagnant or dead. their’s no community around it. Any investment goes towards the proprietary parts. Any advantages of true open source projects don’t exist.

Total failure. Result is crippleware and dishonesty, presenting their products as “open source” when they clearly are not.

Your article was condescending towards those who criticize “open core” as deceivers. And congratulatory to the “open core” vendors.

They want to be proprietary, but adding “open source” to their products is better than free advertising, so they created this misleading expression “open core”. Falsehood is not a value!

#7 Matthew Aslett on 07.09.10 at 1:24 am

The strategy may well be flawed,but for none of the reasons you mention. Sorry if that sounds condescending…

In order for the open core strategy to succeed the open source version has to be a viable project so accusations of crippleware reveal a fundamental misunderstanding of how the strategy works.

And the problem is not that they call themselves “open core” but, apparently, that they continue to refer to themselves as ‘open source vendors”.

Besides, pointing out the deficiencies of the anti-open core argument does not make you pro-open core. The world is not black and white.

#8 Henrik Ingo on 07.02.10 at 11:30 am

Hi Matthew

If you reread my post that you quote, it should be apparent that I do not compare open core to stealing.

I’m using theft and society’s punishment against it as an analogue, to criticise Mårten’s use of “self adjusting system” as a blanket permission to do anything he wants. But nowhere did I explicitly or implicitly say that open core is comparable to stealing, on the contrary, the whole point of using an analogue is to take an example *out of a different domain* than the one being discussed.

I would appreciate if you can add a small correction about this into the main article, since your accusation is rather grave.

Now back to our regurlarly scheduled football match… (other comments another day)

#9 Matthew Aslett on 07.02.10 at 11:49 am

Hi Henrik,

Point taken. Although I think the comparison was inferred it is not worth arguing about. Post corrected accordingly.

Matt

#10 Henrik Ingo on 07.02.10 at 5:52 pm

Thanks. (You gave me much more space even than I was asking for.)

To clarify (great matches today btw, totally worth being offline 🙂 some more: The connection between them is that theft goes against the values of our society, whereas open core goes against the values of the open source community -> hence the analogue. Beyond this, no further similarity is inferred.

I fully understand the temptation to misunderstand it though, I gave an easy target to attack perhaps.

Also congratulations to your own football team that you tweeted about yesterday!

#11 Matthew Aslett on 07.04.10 at 2:55 pm

I’m not convinced, incidentally, that open core in and of itself does “go against the values of the open source community”. As Mark Radcliffe notes (see below) it does not violate the Open Source Definition, either literally or in spirit.

#12 Henrik Ingo on 07.02.10 at 6:05 pm

Re-reading my own text, I see now how the theft example was rather lumped together with the main storyline. I’ve made a small edit in an attempt to clarify for future readers.

#13 Kim Weins on 07.02.10 at 4:11 pm

I have no problem with a company choosing the open core model either. I do however disagree with those who say open core is the only viable choice for building an open source company. Just because there are some companies that have had success with open core models does not mean that it is the only approach that works or ever will work.

#14 Matthew Aslett on 07.02.10 at 5:36 pm

I don’t think I have ever seen anyone claim it is the only viable choice.

#15 Jack on 07.08.10 at 8:38 pm

See? Even you is confused. They are not “open source” companies!

#16 Mark Radcliffe on 07.03.10 at 1:09 pm

I largely agree with Matt’s comments and I think that Simon is wrong in his assumption that end users can only use the “commercial version” of open core products. However, I think that there is a danger that has not been mentioned: if “open core” is demonized and determined not to be open source, then venture capitalists, who like the open core model, will be less interested in investing in open source companies. Although not always necessary, venture capital investment has been important to make many open source projects more valuable to end users. To see more, you can go to my post: : http://lawandlifesiliconvalley.com/blog/?p=485

#17 Matthew Aslett on 07.04.10 at 2:22 am

Mark,

Thanks for the comment and insight into the conversion rates of the companies with which you are involved. Anecdotal evidence suggests that there is no significance difference in conversion of open source to paid subscriptions and open source to proprietary extensions. If it were true that the open core model forces users to adopt the proprietary extensions then we would expect to see a notable difference.

Matt

#18 ☞ For A Topic That’s “Futile To Discuss”, A Lot Of People Seem Interested « Wild Webmink on 07.05.10 at 1:25 pm

[…] Open Core Is Not A Crime “I appreciate why advocates of software freedom are wary of open core. It does perpetuate proprietary software licensing, and it does so via open source. But that does not make it a crime. And a considerable amount of code has been contributed to the commons by open core vendors. Meanwhile even those that would wish to do something to remedy the situation are without the means to do so. Hence the endless and futile debate.” […]

#19 451 CAOS Theory » Do customers want open core? on 07.09.10 at 1:28 am

[…] good insights and arguments: Simon Phipps, Mark Radcliffe, Stephen O’Grady and our own Matt Aslett to name a […]

#20 How Open Core is like Coitus Interruptus (from Control and Community) | OpenLife.cc on 11.27.10 at 1:26 pm

[…] So to once again throw in an analogy here (after all, it worked great last time, didn't it?): […]

Leave a Comment