JCP Conflict
The current JCP conflict is not about the GPL at all. It is all about Java being competitive Free Software or not. Sacha Lebourey has already made the point on how recent Sun movements are the most optimal way to get all the negatives without getting all the positives:
You own that brand, so do something real that justifies breaking this subtle equilibrium you’ve built over the years, not a quick-and-dirty something under the carpet.
And now, as Mr Schwartz reminded everybody renaming SUNW to JAVA at Nasdaq, a healthy Java is crucial for Sun Microsystems. I’m not sure how the misconception is happening, and even "Sun’s VP for software gets it completely wrong, but the FSF is on the same side as the ASF in this conflict. I have not read or heard otherwise from anyone related with Free Software, either officially or in a personal blog, and I would be grateful if someone corrects me if I’m wrong.
If the FSF would be offered a license under similar terms , they would be forced to reject it, as it violates
freedom 2:
The freedom to redistribute copies so you can help your neighbor.
Not every neighbor is allowed to get a copy under a FoU restricted License. Also, a license specifying such conditions would not be acceptable as an Open Source license by the OSI, as it would violate at least:
6. No Discrimination Against Fields of Endeavor The license must not restrict anyone from making use of the program in a specific field of endeavor. For example, it may not restrict the program from being used in a business, or from being used for genetic research.
of the Open Source Definition. restricting effectively what can the software certified used for (Field of Use restrictions).
This is something that, as far as I can tell, violates the GPL too. In fact the FSF has stated publicly that Field of Use restrictions violate section 7 of the GPL v2:
Thus, these “field of use” restrictions would prevent implementation of W3C standards as Free Software. “Field of use” restrictions are also legally incompatible with section 7 of the GNU General Public License, since it does not allow the user’s freedom to modify to be shrunk to zero in this way.
This is from the official FSF position on the W3 Consortium patent policy. Sun, according to Rich Green recent post quoted earlier, will accept any certified GPL implementation so long as it is the ONE OpenJDK they host, or significant derivatives of it. I quote from Dalibor Topic take on it:
It (rather obviously, as OpenJDK is under the GPL, and the GPL does not allow further restrictions) doesn’t have any downstream ‘field of use’ restrictions on the certified binaries. It requires an implementation to be under the GPL, and to be significantly derived from OpenJDK.
No other GPL codebases will be allowed to be certified. This includes, as far as I can tell, that neither GNU classpath, GCJ or kaffe will get the JCK. Correct me if I’m wrong. Dalibor ends saying:
It’ll take getting a bunch of those companies to figure out they are better off with an ubiquitous, free software, verifiably compatible Java platform in the long run, rather than just one or two of them.
That is the really important battle to fight, in my opinion.
Agreed. The Apache Software Foundation is fighting it. A lot of companies in the JCP are supporting the ASF. The ASF has been a part of it for a lot of time, and pushed for it to be relevant for the free software movement at times where it was difficult to argue against those who said that java was a proprietary platform. We keep pushing for a free java, in spite of Sun’s efforts to sink it (quoting Sacha again:
They are not respecting the JCP rules they have themselves defined.). See Geir Magnussen slides for the F2F meeting that took place recently for an example of constructive proposal.Now, why did I say competitive in the first paragraph? Because no healthy ecosystem can survive without some evolution going on. Evolution, don’t forget it, is mutation plus selection. I don’t believe in a world where we can only have one certified OpenJDK, controlled by Sun, and one proprietary Sun JDK. In such a world BEA’s, Apple’s or IBM’s Virtual Machines will disappear, as well as kaffe, GCJ and harmony. No diversity means no or slower evolution, and additionally makes The Big Brother a single point of failure for the whole ecosystem.
I finish quoting a response that Jim Jagielski gave to Geir’s statement java is too important, in the jcp-open@apache.org list. I think it gives a good summary of the feelings of a lot of members of what used to be called the java community:
(...) In other words, if Sun continues down this path, which will invariably lead to Java becoming less and less significant due to the community finally having reached their fill of all this funny business, another language and “technology” will for sure take its place.No - the way Sun is playing this, we all lose. We ned to fix this. Java is too important.
Dalibor,
Given that we are useless in your “JCP Future - Dream world” :) So ASF should withdraw from the JCP?
thanks,
dims
Not at all. The ASF is doing a good job inside the JCP, and I hope it continues to do it in the future. But from the discussions on jcp-open, I think it’s clear that the ASF will not and can not fight for the radical idea of ‘open source everything’ without tearing itself apart between the diametrically opposed interests of its membership.
It’s not the right tool for that particular job.
Posted by Dalibor Topic atDalibor, first I wanted to thank you for the extensive comment. I agree with most of it.
quoting (my strong markup):
FSF has been instrumental in voicing the need for open source Java loud and clear, and getting people on writing the code for it, while it was still unfashionably “idealistic” to do so :). The ASF has been instrumental in making the JCP suck a lot less, while that work was still seen as unfashionably “pragmatic”. :) OpenJDK has been instrumental in removing all sorts of legalese from the JCK license, and I think it will continue to serve the community well in that respect, among many other things.
I agree that the poles you mentioned, idealistic and pragmatics, were the way we all were playing two or three years ago. I tend to agree that OpenJDK is being useful to clean legal terms and ease the use of Sun’s JDK in linux distributions. When I find a problem, and from the things I quote in the main entry I guess you agree, is when Sun uses OpenJDK to get all people staring at the screen, waiting for the film to start... for several years, and they even work hard to stop harmony, which was not waiting for OpenJDK to happen. Getting to the point of bending hard the JSPA rules.
What I’m trying to stick in the wall with this post is a message to avoid the framing of “Java is free, those ASF people is just jealous” which seems to be arising continuously and is clearly wrong. Division in the FLOSS (as people here in Europe uses to say) is not a way forward, and I think that stalling harmony, kaffe or GCJ development goes against the very principles of evolutionary software, and is very bad.
With regards to your answer to Davanum, I think the ASF has enough problems trying to certify Harmony to go into more radical fights right now. After all, and as you just said, we are^Wused to be the pragmatics. Though it looks like the roles are somehow reversing. Interesting times. :)
Dims, welcome, and thanks for passing by.
I have no doubts (but I’m not even officer currently) that the ASF will get out of the JCP in the moment that the perception that it is not useful for getting things done to the membership and users of our software gets prevalent. For the moment it is only a minority thinking this way, and there are (optimists? realists? ) that think that there is room for progress inside the JCP.
As an example, two years ago I installed blojsom for my English blog, in a failed attempt. 4 years ago I used JSPWiki for my wiki/blog in Spanish. Now I’m migrating all to mombo, which is mostly python but, the way it is designed, could morph into a mixture of python, ruby, erlang, php, whatever in little time.
Posted by Santiago Gala atI think ASF’s problem in the definition of the conflict is that it has framed the issue from the beginning as an extremely complicated issue of interpreting some long & weird contract between Sun and the ASF.
From there, to ‘I don’t have a clue what the ASF is complaining about, they are probably just jealous’ is not a long way, and does not require Sun to do anything to frame the issue.
Posted by Dalibor Topic atBut that aside, I don’t think that the ASF has anything to gain by
going after OpenJDK, in particular. People will just see that as the
familiar BSD vs. GPL mudpit at best, and frame it as jealousy at worst.
On the contrary, it would be quite interesting, for example, if the ASF
went ahead, and asked to certify a Harmony/OpenJDK hybrid under GPLv2.
That would be creatively gaming the system and stretching the boundaries
of what’s possible in a “pragmatic” way. But due to the way the ASF
works, that’s not going to happen, and if an ASF director were to
propose such a thing, I’m sure some people would call for his head on a
pike.
ASF’s ability to explore such options is limited by the structure of its
membership, and the more or less exclusive focus on one license, and one
way of doing things. Classpath has a bit of an advantage there, since we
tried to spread out the diversity of ways of doing things, licenses and
communities as far as possible (more than two dozen different JVM
projects are using it, from proprietary ones to GPLd Kaffe, JamVM, gcj &
cacao), so we can play both the “pragmatic”, and “idealistic” roles,
regardless of the situation. With the ASF, as soon as the GPLv2 is
involved in any form, the option for the foundation to play the
pragmatic role is gone, due to its membership structure. And Sun surely
knows that.
That, again, is something I don’t see the ASF as being able to overcome,
without tearing itself apart between the anti-GPL “idealistic” fraction,
and the rest of it. So I understand why Geir has picked the strategy he
is using, his other option is to do nothing, and the majority of the
membership seems to be too frustrated with the situation for that to be
a viable option.
Unfortunately, I don’t think that the strategy of taking that particular
issue head on, now, with the ASF as the tool to do it, is working, and I
don’t think it can work, as Sun can simply choose to continue not to say
anything, denying the ASF a platform to base their public campaign on,
and wait it out, until they and IBM make whatever deal they need to make
to make it all work out.
And I’m pretty sure they won’t ask me or Davanum to hold the candle
while they make their deals. ;)
If you want the JCP to get out of that dark age of backroom deals, there
is only way: make them all play by the same rules, to open source
everything in a community maintained standard by default.
A Couple More Minutes
Today my 15 will be complete when my interview with Kuldip gets published on SDN-TV . You will get an update on OpenJDK and the work the community has already accomplished towards a completely Free runtime in creating IcedTea . Currently on SDN you...Excerpt from Tom Marble's Weblog at
Dalibor , first I wanted to thank you for the extensive comment. I agree with most of it. quoting (my strong markup ): FSF has been instrumental in voicing the need for open source Java loud and clear, and getting people on writing the code for...
Excerpt from Boxes and Glue's Comments at
What is limiting the competitive aspect, though, is that the JCP has always been discriminatory in the access to the TCKs, since they’ve rarely been published under open source licenses. When that fundamental component of a community initiated and maintained standard is not available to that community under an open source license, you don’t really have a community of equals among implementations. You have a walled garden, where the vendors leading the specifications are managing the toll booths at the gateways and discriminating implementations based on their implementors financial power, as they do since the TCK scholarship was instituted, or based on their current business goals, as the latest conflict of interest shows.
You can get rich and famous within JCP’s walled garden, and that’s OK, a lot of people have gone that route, and emulated the business models used by Sun, IBM, Intel, BEA, Oracle, et. al. based on exclusive control of vital elements of specifications, combined with discriminatory access to them. But if you want to turn that system on its head, you need to chose the right tool for the job, and the right point to apply the pressure on.
FSF has been instrumental in voicing the need for open source Java loud and clear, and getting people on writing the code for it, while it was still unfashionably “idealistic” to do so :). The ASF has been instrumental in making the JCP suck a lot less, while that work was still seen as unfashionably “pragmatic”. :) OpenJDK has been instrumental in removing all sorts of legalese from the JCK license, and I think it will continue to serve the community well in that respect, among many other things.
I don’t think that FSF, ASF or OpenJDK are the right tools for turning the JCP around, though, and forcing the vendors to give everyone equal, non-discriminatory access to the specifications, RIs, and TCKs without NDAs, or other limitations. Only a mandate in the JSPA for open source RIs, TCKs, etc. could guarantee that the community can enjoy the benefits of its standardization work freely. And none of the groups above is the right tool to push that through, due to such organizations always having to represent a complex set of opinions and interests, rather than being able to do the obviously right thing to do and no longer let any JSR with a proprietary component, NDA, or similar nonsense through.
The right tool for that job is a community that will demand equal access to the fruits of its labor for everyone, regardless of their corporate affiliation, size of their bank account, etc, guaranteed by the requirement that all output of community’s specification process be open source, available publicly at no cost, under a patent covenant for every implementation of community standards. A radical thought, I know, almost as radical as talking about open source Java in public was just a couple of years ago. :)
That community is not represented on the EC yet ... but it will be, sooner or later. I think it will be a community of individuals, rather then organizations, that will save Java and the JCP from being driven into the ground by squabbling vendors.
Posted by Dalibor Topic at