01.27.08

Selling Services Without Selling Fear of Licences

Posted in FUD, GPL, Security, HP, FOSS, Rumour at 2:29 am by Roy Schestowitz

Accusations against H-P and Palamida seem baseless

It wasn’t long ago that McAfee and InformationWeek were both harshly (and rightly) accused for spreading GPL fear [1, 2, 3]. This was not appreciated. It is actually worth reminding ourselves of speculations and predictions of a McAfee-Novell tie-up because Novell too was caught using FUD to market itself.

“Empty allegations are used against Hewlett Packard (H-P) and Palamida and we wish to present them here in order to make some clarifications.”On the other hand, some baseless accusations are flying about at the moment. Having been in touch with some of the parties involved, we wish to debunk FUD (or just lies) about FUD that never was. Empty allegations are used against Hewlett Packard (H-P) and Palamida and we wish to present them here in order to make some clarifications.

Let us start with H-P. Just the other day, when H-P introduced a set of services and tools that assist tracking of software and licensing, Dana Blankenhorn accused rather than thanked.

The Hewlett-Packard open source strategy is becoming clear.

Fear the source.

I’m certain HP officials will disagree with that. But when your press release is headlined, ” HP Promotes Open Source Software Governance with New Initiative,” there is no other conclusion to draw.

Your big company can’t go into open source alone. It’s dangerous out there. Here, hold our hand.

PJ disagrees with this, as do I. “HP is trying to do something very good with Flossology. I totally support it,” she says.

Why would anyone try to show just the negative side-effect (and yes, we’re sometimes accused of doing this as well)? Maybe because it stands out from the crowd and because ZDNet bloggers can be rewarded for provocations. Regardless of the issue at hand, H-P did make either an observation or a complaint back in 2005 (maybe 2006) when it said there were too many open source licences. But coversely, In this newer case, there is an attempt to address the issue, not just raise it. We should be happy. We should be thankful. And here were have the latest report from Palamida (published on Friday) which heralds to the world that GPLv3 finds love. This is good news, not bad news. Project evolve successfully.

The GPL v3 growth for this week is consistent with our average growth rate. As of January 25th, the GPL v3 count is at 1579 GPL v3 projects, up 44 projects over the past week. The LGPL v3 list is growing slowly but steadily and is currently at 150 LGPL v3 projects, as compared to last weeks number of 148 LGPL v3 projects.

At least one person claimed to have found flaws in Palamida’s work. Here is what one of our readers had to tell to us before we heard from Palamida (it’s reverse-chronological):

[Anonymised:]

I have been visiting Palamida GPLv3 site and I think they are doing a great job at tracking the license adoption, and their statistics can be very useful to counter the established proprietary software oligopolies’ and the mainstream tech media’s FUD machine.

But today I have been warned by Pieter Hitjens about the following: I copy-paste the conversation about recent statements made in the palamida gplv3 site (gplv3.palamida.com -which redirects to –> gplv3.blogspot.com)

[Pieter:]

http://gpl3.blogspot.com/

This site looks like it’s promoting GPLv3 but in fact it looks like subtle anti-GPLv3 FUD. E.g.:

“In the case of putting a GPL v3 project under a commercial license as well, there is high potential to violate the terms of the GPL v3. This is not to say that any of the aforementioned projects are or are not
in violation of the license, since our analysis of the terms are not yet complete, but caution should be used if a project is under both the GPL v3 and a commercial license.”

What they are saying, I think, is that GPL projects that do not have a clear copyright centralization cannot easily be re-licensed. However they don’t state this clearly, and they are not publishing my comments on the blog.

-Pieter

[Anonymised:]

as somebody who has gotten note of Palamida very early after GPLv3 was released and I’ve got a bit of contact with actual GPLv2->v3 conversions, I can say this:

Palamida, the owner of this blog (it’s advertized in the banner on the top of the blog) is a company who’s business is software risk management, so it’s the business of marketing at this company to show what risks may be there and that risk is increasing.

It is increasing, because GPLv3 makes things indeed a bit more complicated by the simple fact that it is a successor of GPLv2.

The only long-term solution to that which I see is to convince as many free software developers that licensing under “GPL v2 only” is a __very__ bad idea.

I think you guessed right that they may suggest that companies might want to buy services from Palamida, to improve legal security in software distribution.

What I see, rather looks like research which gives great information of the GPLv3 adoption, and no clear FUD.

[Anonymised:]

I see clear FUD, in this respect.

Dual-licensing is in fact a very strong argument for using GPLv3 but it depends on clear centralization of copyright. Projects like 0MQ - see www.zeromq.org - are careful to demand copyright assignments and/or MIT licensing from all contributors. For these projects, dual licensing is essential. This statement:

“This is not to say that any of the aforementioned projects are or are not in violation of the license, since our analysis of the terms are not yet complete, but caution should be used if a project is under both the GPL v3 and a commercial license.”

Is really bad. It suggests that we have to wait for Palamida to give the green light on whether it’s safe to use 0MQ. That’s very misleading and designed to create business for Palamida by exaggerating the complexity of the GPLv3 and ignoring the key role of copyright ownership.

If a company owns its code, how can it be in violation of the GPLv3 by dual-licensing its own code? That’s pure FUD, and worse, it brings into question one of the key business models for new smart FOSS businesses.

[Anonymised:]

Care if I forward your message to Pamela Jones (groklaw) and Roy Schestowitz (boycottnovell) so they alert about the issue. Think the palamida guys, who are doing a great tracking of projects adopting the GPLv3 should be aware as well. And of course the FSF/FSFE

[Pieter:]

Forward away, of course. Tracking GPLv3 usage is fine. Throwing fear and uncertainty onto other businesses to try to create extra business is not fine.

-Pieter

Shared with implicit permission, the above is intended to at least show the arguments that were thrown into this debate, which we believe is resolved by several factors.

For starters, PJ says: “I don’t agree they are doing that [spreading fear]”. Further: “They want business, so they highlight problems without telling you the solution, because they want business, but that isn’t, to me, exactly the same thing as FUD, although it can have a similar effect.”

Our reader adds: “Up to now, their work at tracking GPLv3 project has proven nice and useful to counter quite a lot of FUD […] I think Palamida at least should publish Pieter’s comments. If they don´t do it after a while, “someone” should be pointing at the problem. Of course making clear that the tracking of GPLv3 projects is nice and useful.”

We received a response from Palamida quite quickly and it was very convincing. Judge for yourselves however:

I can say with 100% honesty that no, Palamida does not resort to FUD to sell our services. However, we do point out what can happen if you don’t know what you’ve got in your code base, which is a reality, and it’s what drives a lot of lawsuits and insecure apps. It’s just something people want to avoid and we’re here to help organizations figure it out so they can get it right. There is a subset of folks (including you) that know what the heck is going on and would vet and check you code, versions, and licenses ahead of time. Funny though that very large organizations often do not, or possibly can not, because of their size and geographically dispersed team of developers. These are the folks who have the Top 5 Most Overlooked OS vulnerabilities (and many more but let’s stick with 5) and don’t know it.

So in general, our message and mantra has always been “Know What’s In Your Code.” It’s a message that shouldn’t be considered FUD, because not knowing has very real consequences (can anyone say Busybox?).

Since H-P came under similar unjustified scrutiny we brought up this issue, which quite expectedly revealed sympathy:

In general, we like HP but here’s something to think about. Back at the beginning of Palamida, folks used to ask us, “Why wouldn’t I just use Google Code Search instead of paying for Palamida?” Our response was always that
they certainly could use Google if they only wanted a skim the surface view of what was going on in one single segment (say, JBoss code). However, our expertise coupled with the depth and breadth of our code base (which weighs in at 3 Terabytes) could give you a little more (to put it mildly). So I personally feel the same about FOSSology. This is my singular opinion, it’s a fantastic tool but it answers only one of the many, many questions people need to be asking (take a look at the blog we just posted Friday) about: what code are you using? What version? What license is it under? Is it secure?

How often is the FOSSbazaar updated? What does it include? What are its rates of false positives or irrelevant search matches? How comprehensive is it? Who has tested it? Would you bet your eBanking system security on it?

That sort of thing.

This hopefully resolves the issue, at least for those who were involved in a blame game. Censorship (aka “selective approval”) of comment was probably the main reason for going this far. We never delete comments in this Web site and only a single abusive reader has his comments flagged (still truly visible) for repetitive abuses even against other readers. Transparency brings better answers than censorship, which we last complained about just an hours ago (ODF/OOXML). █

Permalink  Send this to a friend       View as PDF