Neary: Rotten to the (Open) Core? [LWN.net]

Dave Neary steps into the open core debate on his blog. Part of the problem is that people have divergent definitions of open core, he says. "There is another name for this which is even more pejorative, Crippleware. Deliberately hobbled software. And that's what I think gets people riled up — if you're releasing something as free software, then there should at least be the pretence that you are giving the community the opportunity to fend for itself — even if that is by providing an "unofficial" git tree where the community can code up GPL features competing with your commercial offering, or a nice forum for people to share templates, themes and extensions and fend for themselves. But what gets people riled is hearing a company call themselves "an Open Source company" when most of the users of their "open source" product do not have software freedom. It's disingenuous, and it is indeed brand dilution."

to post comments

Neary: Rotten to the (Open) Core?

Posted Jul 19, 2010 23:52 UTC (Mon) by SEJeff (guest, #51588) [Link] (7 responses)

Sounds a lot like syslog-ng and syslog-ng "enterprise"

Neary: Rotten to the (Open) Core?

Posted Jul 19, 2010 23:57 UTC (Mon) by rahulsundaram (subscriber, #21946) [Link] (4 responses)

Part of the reason why pretty much every distribution has moved from syslog to rsyslog instead of syslog-ng.

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 14:00 UTC (Tue) by frobert (guest, #62734) [Link] (3 responses)

Seems they have also realized this and going for an LGPL+GPL combo license: http://bazsi.blogs.balabit.com/2010/07/syslog-ng-contributions-redefined.html.

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 16:19 UTC (Tue) by nix (subscriber, #2304) [Link]

Excellent! And Baszi used this as an excuse to make an architectural improvement as well :)

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 18:39 UTC (Tue) by SEJeff (guest, #51588) [Link] (1 responses)

Very cool, but features still in syslog-ng enterprise that you pay for are in rsyslog for free.

Neary: Rotten to the (Open) Core?

Posted Jul 21, 2010 6:25 UTC (Wed) by frobert (guest, #62734) [Link]

That's right. But as it goes, people have to make a living somehow. I think Bazsi would be the happiest if he hadn't have to maintain two different versions of the same stuff.

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 9:28 UTC (Tue) by nix (subscriber, #2304) [Link]

Except that enterprise features migrate fairly rapidly into non-enterprise, and people do sometimes (not often) contribute features to non-enterprise which then go the other way.

syslog-ng's case is *much* more like what Artifex used to do with ghostscript.

Neary: Rotten to the (Open) Core?

Posted Aug 8, 2010 11:22 UTC (Sun) by bazsi (guest, #63084) [Link]

First of all, I want to make it clear that I'm biased on the syslog-ng case, but still wanted to express my opinion here. I'm biased as I'm the primary author of syslog-ng.

I think syslog-ng is a completely different case from the one described by Neary. The GPL version is not crippleware, it was never published for marketing purposes only and for the majority of syslog-ng's existence only the Open Source stuff existed. The Premium Edition is only about 3 years old and syslog-ng started in 1998.

We never removed features from the OSE version, the Premium Edition only included _additional_ features, and a lot of those are already available in the OSE.

Some examples:
* TLS support (became available in 3.0, almost 2 years ago)
* SQL destination (became available in 2.1, 2.5 years ago)
* performance improvements (3.0)
* etc.

In the other direction, we usually receive bugfixes and it is a pure technical reason that we used to require copyright assignment: I wanted to keep the two branches as close as possible (which if not done is the reason #1 why Open Core products become crippleware fast). _And_ since we heavily invested in automatic testing and our customers report bugs directly to us, we fix way more bugs in the OSE version than the community.

But anyway, I didn't think that the dual license model was so problematic at the time we made this decision 3 years ago. Our efforts have never been "Rotten to the Open Core". If you don't believe that, check out the git repository or read the mailing list archive and see it yourself.

And this whole mess is the past, OSE 3.2 has been relicensed, and it is true that we're going to publish non-free plugins, but anyone else is welcome to join and do the same.

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 0:21 UTC (Tue) by ncm (guest, #165) [Link] (8 responses)

If "open source" had any objective meaning at all, it would be easier to get worked up about this. They can call themselves an "open source company" all they want, and nobody has any legal basis to complain. They aren't a "Free Software company", and they don't call themselves one, so there's no conflict there. The question is, why pay them any more attention than any other proprietary software shop?

Their paying customers might reasonably feel misled. Surely that's a matter between them and their customers. A company that sets itself up as adversary to its own customers is not long a threat to anyone.

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 0:33 UTC (Tue) by rahulsundaram (subscriber, #21946) [Link] (7 responses)

Bit of self contradiction there. There is no legal basis whatsoever for complaining about it if companies called themselves Free software companies. If the community can discourage one, it can discourage the other as well. Why do you believe that "Free software companies" have a objective meaning but not "open source companies" I doubt everyone has a single definition for either in mind. It affects customers, community and companies which genuinely are in the business where free and open source software is key.

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 5:13 UTC (Tue) by ncm (guest, #165) [Link] (5 responses)

It doesn't matter. The fact is that they _don't_ call it Free Software, and aren't tempted to call it Free Software. Whatever anybody thinks Free Software is, what they have ain't it, and whatever disagreements people have about Free Software, they don't disagree about that.

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 13:29 UTC (Tue) by rahulsundaram (subscriber, #21946) [Link] (4 responses)

That's because open source as a term has won over Free software in popularity years back. If Free software had won, we would have similar problems with distortion of names. It is naive to believe otherwise. I would say Free software would easier to distort in several ways.

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 14:28 UTC (Tue) by Zack (guest, #37335) [Link] (3 responses)

>If Free software had won, we would have similar problems with distortion of names.

I doubt that. I suspect the often blamed double meaning of the word "Free" would work against that.
Whereas a loosely knit community of many individuals has no problem explaining the two different meanings over and over again to different people (using the speech not beer catch-phrase, which is quite humourous the first time a person hears it, and catches their interest), for a commercial company it would require a real expense to make a differentiation between gratis and libre. What's more is that they need to establish that first and from there on distort the term "Free Software" for their own purposes, but by that time they have already educated their customer, and it would be a lot harder to mislead them.

With the term "Free Software" you cannot skip on explaining why "Free" doesn't mean no cost, especially in a context where you would try to maximise profits by distorting its meaning. But once you explain what it doesn't mean, you can't just leave it dangling there; you'd have to give an explanation what it *does* mean. It's probably possible to distort the meaning there, but all in all the process to distort would be more laborious and error prone and as such less profitable.
The term "open source" is far more monetizeable, which I think is one of the reasons it was adopted quickly, but for the same reason it is unfortunately also more vulnerable to corruption.

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 14:32 UTC (Tue) by rahulsundaram (subscriber, #21946) [Link]

We can all continuously second guess history. Companies can give free crippled trialware and call it Free software too and most of the users wouldn't know any better. My opinion is that since open source is the term that got popular, time is better spend protecting it rather than trying to rewrite history.

Neary: Rotten to the (Open) Core?

Posted Jul 21, 2010 5:24 UTC (Wed) by jmalcolm (subscriber, #8876) [Link] (1 responses)

I am not sure why you think companies would corrupt the "Free Software" brand by explaining lots of things. Rather, I would expect them just to run with the common expectation of what "free" would mean from a corporation. That is, they would just say that they have a "free" community version and a commercial "enterprise" version without saying anything about "libre". The corruption would come when companies would offer "free" software for which the source is not available and talk about it as part of the "Free Software" trend. This would significantly dilute the brand of the political movement that is more interested in libre than in no cost. We see this already and "Free Software" is a much less successful brand than "Open Source" outside of the community of people that care about the difference (ie. most of the world).

Neary: Rotten to the (Open) Core?

Posted Jul 21, 2010 11:09 UTC (Wed) by Zack (guest, #37335) [Link]

>I am not sure why you think companies would corrupt the "Free Software" brand by explaining lots of things.

Because without any explanation, "free software" would simply default to software at no cost for most people, as it is today. If a company wanted to co-opt the libre meaning "Free Software" for nefarious purposes, they would have to at least explain how free has two meanings, and that the gratis one isn't the one they refer to by "free".

If a company would put out a shareware version as free software, it would simply mean that, a version of the software at no cost; they wouldn't even be lying.
If they would want to use that to subvert the libre meaning, they would have to position it as "free, but not just as in no-cost", leaving open the question "but then free as in what as well ?"

Please note that I don't think open-source is a term not worth saving or guarding, nor do I suggest it should be dropped wholse-sale in favour of Free Software. I am responding to the statement that the term "free software" would inevitably have suffered from the same problems as the term open-source, and that to believe otherwise amounts to being naive.

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 6:03 UTC (Tue) by epa (subscriber, #39769) [Link]

For ages now RMS has promoted a clear meaning for 'free software company' as a company which sells free software and only free software. I suppose the other poster was thinking of that.

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 2:19 UTC (Tue) by pabs (subscriber, #43278) [Link]

I would echo Benjamin Otte's comment from that post.

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 4:27 UTC (Tue) by butlerm (guest, #13312) [Link] (2 responses)

In my opinion, it is legitimate to make _add on_ products and utilities that are proprietary. But the open source software package released as part of an "open core" project should be based the same source code as the core of the proprietary version.

Otherwise the company is essentially forking the code base and making a "hand-me-down", flaky version for the outside world. If the open version of the core package isn't genuinely useful, stable, and up-to-date as a standalone software package, the claim to "open core" is a sham. Why would any outside developer want to contribute to a project like that?

And if the "open" core doesn't actually share the same source base as the "proprietary" core, the claim to "open core" is arguably fraudulent, because the core of the proprietary version isn't actually open at all.

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 4:39 UTC (Tue) by dlang (guest, #313) [Link] (1 responses)

I don't even mind the proprietary fork, except that the next step is always to prevent the opensource version from gaining some of the features that the proprietary version has to protect the company income.

Neary: Rotten to the (Open) Core?

Posted Jul 20, 2010 6:16 UTC (Tue) by butlerm (guest, #13312) [Link]

If a company wants to produce a proprietary version of a core package, certainly the legitimate way to do it is to create hooks and/or plugin support so that others can do the same. No fork required.

Personally, I am skeptical of the merit of any substantial open source software project that multiple companies can't each participate in and add value to in their own way. At best the open source character of a single company dominated project (like MySQL) is as an insurance policy against the day when the project management by that company becomes intolerable or fails entirely. Otherwise it seems more like a PR move with marginal side benefits than a game changer for the software industry.

Earning money

Posted Jul 20, 2010 9:12 UTC (Tue) by aristedes (guest, #35729) [Link] (4 responses)

I write and publish software for the education market (http://www.ish.com.au/oncourse ) We give away that software for free (but not open source) and charge for additional services (which are maybe only 10% of the product).

We've long thought about the idea of open sourcing the core application, but even under GPL, we'd have to remove a great deal of functionality from the application or risk having no customers at all who'd need pay us for our services or extra features. Yes, I know we could go for a pure services model (support contracts, etc like RedHat) but our customer target would not support that arrangement (in my opinion).

I understand why you'd go for an approach of pulling out some parts of a product, release it as open source and hope to attract a community (maybe even for all the right open source reasons). But still want to hold something back to create a market for your company to survive on. Not every software product suits a pure 'support services' model of open source.

I don't understand the antagonism towards companies that are trying this approach. They may well understand the open source 'way' but are also looking for an approach to earn a living which isn't always easy if your customers grow to expect 'open source' == 'inexpensive'. I want to give our customers freedom to extend our product, customise it into their business and not be locked in artificially (that is, I want them to stay because the product is good not because of a contract or because they can't get their data out). But I also want to stay in business and pay the developers who make all this neat stuff.

I haven't open sourced our product mainly because I don't have the resources to support a fork. But the idea is an attractive one if I did.

Ari Maniatis

Earning money

Posted Jul 20, 2010 9:40 UTC (Tue) by gevaerts (subscriber, #21521) [Link]

That model as such might be OK in theory, but in practice there often seem to be several problems:

- The "GPL" version as it's often called (it's usually GPL) is marketed as some sort of demo. This on its own will create resentment, and will stop many people from even considering looking at your software. Selling additional propriteray components is fine, (subtly) trying to imply that the GPL version is somehow inferior is insulting to the people who worked to make the GPL ecosystem great.

- What happens to outside contributions? Will you refuse a patch because it reimplements a proprietary feature? If so, prepare for some bad feelings.

- If these companies accept outside contributions at all, they usually require some sort of copyright assignment or unlimited rights to use the contributions in the proprietary version. The outside contributor often does *not* receive the right to also sell a proprietary version. This is fundamentally asymmetrical, and enough reason for many people (including myself) to not work on this software. (aside: I don't necessarily *want* to release my own proprietary version, I just want equal rights for equal effort. If you don't expect (at least) equal effort from the community in the long term, why do you even want this community?). Using a BSD style license would solve this of course, but one gets the feeling that this was a deliberate choice.

Earning money

Posted Jul 20, 2010 10:44 UTC (Tue) by tzafrir (subscriber, #11501) [Link] (1 responses)

There are a number of advantages people have in mind with respect to "Open Source" or "Free Software". One of them is independence of the vendor. If Oracle goes down, or loses interest in OpenSSO, someone can take over. If Sun/Oracle has slightly conflictling business interests with Novell, the latter can maintain its own semi-fork of OpenOffice.org.

Will the users of the software be indeed independent of you?

If you retain vendor lock-in, then maybe it is a bit misleading to call the resulting product "Free Software / Open Source". "Inexpensive" is an expected by-product of the lack of vendor lock-in (in most cases), and not the other way around.

Earning money

Posted Jul 22, 2010 11:50 UTC (Thu) by mjthayer (guest, #39183) [Link]

> There are a number of advantages people have in mind with respect to "Open Source" or "Free Software". One of them is independence of the vendor. If Oracle goes down, or loses interest in OpenSSO, someone can take over. If Sun/Oracle has slightly conflictling business interests with Novell, the latter can maintain its own semi-fork of OpenOffice.org.

That is rather hard to reconcile with producing software when you don't expect to get back your costs by charging for service. If the only way you have of making money from it is by selling licences, then vendor independance is pretty fatal for you (and may not help the user in the long term if no one is paying for maintainance). Something like the Qt promise, like that if you stop maintaining the software then it will all be released as GPL or whatever, might provide a good middle ground though.

Out of interest, do you know of many non-niche companies that are fully FLOSS and make a profit? Or pieces of software for that matter?

Earning money

Posted Jul 20, 2010 17:21 UTC (Tue) by iabervon (subscriber, #722) [Link]

I think that, in order to have an "open core", you really have to have a "core" as an entity. If you do, it's perfectly reasonable to have proprietary plugins, and there aren't the same conflicts over whether you'd accept contributions that would compete with them, because it would be a layering violation to accept plugins in the core. Open source plugins, whether or not they compete with proprietary offerings, are a separate project from either the core or the proprietary plugins, in much the same way that mod_python is separate from apache.

The issue is that a lot of companies seem to be claiming an open core when they don't have a division between a core and non-core in the first place, but rather have a single project with more and less functionality in branches. This puts the maintainer in a conflict of interest, rather than cooperating on one front while competing on a different front. It only makes sense to make something open source when you and the community would both be motivated to make the part you share as complete as possible within the scope of that part.

"the open core debate"

Posted Jul 21, 2010 8:12 UTC (Wed) by dmk (guest, #50141) [Link] (2 responses)

Does someone care to explain what this "the open core debate" is about?

"the open core debate"

Posted Jul 21, 2010 20:18 UTC (Wed) by hingo (guest, #14792) [Link] (1 responses)

I think I'm partly responsible, so for once a link to my personal blog may be a legitimate way to answer your question. If you read from this post onward: http://openlife.cc/blogs/2010/june/open-core-not-open-source ...and also take into account links to other blogs (the interview of Mårten Mickos, Matthew Asslett's blog, SugarCRM, etc...), then you'd be pretty well up to speed.

"the open core debate"

Posted Jul 22, 2010 5:39 UTC (Thu) by dmk (guest, #50141) [Link]

thx! this indeed gives some context to the news.

"Open core" overloaded

Posted Jul 21, 2010 8:44 UTC (Wed) by job (guest, #670) [Link] (7 responses)

Someone keeps trying to market his own pet term "open core", but Open Cores is one of the larger free hardware projects and is quite successful in it's niche. Google the term and you'll find that project as the number one hit, and a couple of references to that project.

Why in the world are they still going on about "open cores" as another word for open-source-as-crippleware? They haven't hijacked the term yet, and I don't they ever will. It would be downright nasty to the people who have invested time and money in Open Cores.

There are perfectly good other terms or just make up a new one ("open crippleware" anyone? or "open demo"?). I keep misreading those headlines every time they pop up and aren't actually about open hardware.

"Open core" overloaded

Posted Jul 21, 2010 21:55 UTC (Wed) by hingo (guest, #14792) [Link] (6 responses)

"open core" seems to have been adopted by FOSS business people and analysts in 2008 and is quite established by today. I personally knew about "open cores" (and wrote about it in a book, even :-), but the term is in wide use now, what can you do.

FWIW, when I first heard about "Open Cores", I had no idea that a "core" referred to chip design. It's an easy mistake to make.

"Open core" overloaded

Posted Jul 25, 2010 6:48 UTC (Sun) by job (guest, #670) [Link] (5 responses)

If by "wide use" you mean a handful of blogs. Just Google the term. They have just now made a dent in the first page of results with a mention at Gartner, despite the fact that their blogs are highly interlinked to skew search results in their favor. I think it's downright rude.

"Open core" overloaded

Posted Jul 25, 2010 12:24 UTC (Sun) by hingo (guest, #14792) [Link] (4 responses)

There is no absolute truth in Google anymore, the results are much more personalized than they once were. I get 6 search results for the business model term, 1 unrelated Java result and 3 for opencores.org.

If you want people to use some other term, why not contact those who use it. Personally, I think it is just too late.

"Open core" overloaded

Posted Jul 31, 2010 20:56 UTC (Sat) by nix (subscriber, #2304) [Link] (3 responses)

That might not have been personalization. You could just have had a different set of machines win the race to provide you with answers. Google is a massively distributed system: in such systems, cross-node consistency is the first thing to go.

"Open core" overloaded

Posted Aug 1, 2010 11:32 UTC (Sun) by hingo (guest, #14792) [Link] (2 responses)

"There is no standard google any more." 57 personalization signals. Ditto FB. @elipariser of moveon talks about "filter bubbles" at #pdf10
twitter.com/timoreilly

Since "job" is clearly active in open hardware, and I've been active in the open source related debate, I would still guess on personalization as that makes more sense than assuming that we got well personalized results just due to chance.

"Open core" overloaded

Posted Aug 8, 2010 13:13 UTC (Sun) by nix (subscriber, #2304) [Link] (1 responses)

But you *do* get different results in consecutive search queries due to chance. Several systems with slightly different datasets race to give you results, and whichever happens to get there first is the one you see the results from (very roughly).

Don't assume that all variability is due to some sort of evil personalization. Some of it is an efficiency hack.

"Open core" overloaded

Posted Aug 11, 2010 7:31 UTC (Wed) by job (guest, #670) [Link]

Interesting. I never knew your Google results were so very much personalized. That will give the SEOs something to loose sleep over.