In political terms, this is an oligarchy, a governing system where the final decisions are in the hands of an elite. This is further reinforced by the selection model. Sun gets one seat. Five seats are elected. The other ten are nominated by Sun and ratified (or, theoretically, not) by members. In other words, these ten are chosen by one-candidate elections.

Or to put it another way, if the individual members rose up in unison, we could elect, at most, five of sixteen members. For the rest we would have to vote down company after company, or person after person, until Sun nominated them someone we wanted.

Or, more simply, it's "You discuss, we decide."

As individuals we should join, but the JCP needs a more open system that allows all stakeholders, not primarily the largest, a part in the final decisions. More members should be elected. Equally important, it should be people on the EC, not companies. Yes, people should be elected to represent commercial interests. A person with good understanding of the telecommunications industry, say, could be elected, and while they might be employed by one company, they should be expected to represent a set of commercial interests that elect them, not just that one company.

The Jini Decision Process has many different requirements so is probably not simply transplantable. But it has a system that has individuals elected by both commercial and individual blocks. This gives it a structure that represents more stakeholders at the top.

Politically, the technical term for this is "representative democracy".

Or, as it is commonly known, "The worst form of government. Except all the others." It is time for a less worse form of governance in the JCP.

Comments

• Seconded. One of the reasons why I am NOT on the JCP. Could you members in there get together and make a JCP 2.7?

  cheers, dalibor topic

  Posted by: robilad on May 18, 2006 at 02:12 PM

• Well, one could join and then agitate. You know, be part of the proletariat rising to throw of the yoke of the petit bourgeoisie or something. Although we aren't really talking about petit here, but still the principle may apply.

  Posted by: arnold on May 18, 2006 at 05:13 PM

• Here's my question, can anyone submit a JSR? For instance I want the inclusion of a few new keywords in the Java language. Can I just submit it via one of those forms? Or do I have to be part of some group, paid a 3K USD membership fee etc...

  Posted by: ilazarte on May 19, 2006 at 10:35 AM

• Entering the JCP requires to sign a 10 page license plus one or two additional "Exibits". The stuff is written in terrible lawyer speak. If you are old-fashioned, like I am, and actually want to understand what you are signing, you need a lawyer by your own to interpret it for you.

  
  Further, the JCP is a buerocratic system. It takes time to manipulate it so you get what you want. Time which company representatives have. It's their job, that's how they make money. But time which many individuals don't have, because they need their time to make a honest living.

  
  Third, it is not in the interest of the established JCP members to have more JCP members. More members means more discussions and make it more difficult for the established members get what they want. So don't expect a process change from within the JCP. The current JCP secures their power.

  Posted by: ewin on May 19, 2006 at 03:02 PM

• So you're suggesting indirect action instead of direct action?

  Posted by: robilad on May 19, 2006 at 03:28 PM

• I'm just saying that those who are in a position to join can help influence from the inside, and others can work from the outside. Whether you are in a position to join is up to you: Maybe you can't stomach the legal contract even if you could technically sign it, so you're an outside agitator. If you can join, then you can make your comments as a member.

  Or, more simply, I'm saying that you should consider whether to work from the inside instead of assuming that the outside is the right place.

  Posted by: arnold on May 19, 2006 at 03:34 PM

• Pardon my ignorance but.... can we actually be members of JCP?
  In particular, what's the difference between a member of the JCP and an outsider (like me as of now) who occasionally clicks into one of the JSRs and look at the comment section of the EC votes, scorn at IBM's repetitive (and personally, I think they are not doing themselves any favour) message, and giggles at more random ones by Doug / Sun / Google etc.? :-)

  Posted by: alexlamsl on May 19, 2006 at 05:54 PM

• You can not work the JCP from the outside. It is by intention a closed system, well shielded and tuned in favor of companies. E.g. from the outside one does not see when work on a new JSR starts. JSRs just magically appear. You can't see in advance when decisions are about to be made. You can't even find a particular spec about a subject of interest unless you magically happen to know the JSR number.

  As I already mentioned, the company representatives can work the system full-time, because that's what they are paid for. As an individual one has to find the money/time to work on a JSR and to outsmart the full-time workers on their home turf.

  Asking for more individual JSR participation is a PR trick if the conditions are not changed. This includes a much more transparent management of the JCP, but also a human-readable license. As long as the JCP is designed and run to put individuals off, it doesn't make much sense to ask for more individual participants. Change the conditions. The current conditions have been proven to not attrackt individuals.

  
  @alexlamsl An individual can become a JCP member, if you are willing to sign over your first-born to Sun in the JSPA. A company can become a JCP member if they are willing to shell out $5000/year.

  Posted by: ewin on May 21, 2006 at 11:41 AM

• ewin, I was talking about creating pressure from the outside, not about changing what JSRs get approved.

  In fact, you have roughly restated what I said: The JCP is not designed for individual participation, so telling folks to sign up is not enough -- they have to change so it makes sense to do so. Right now it's skewed not just to companies but to the biggest companies.

  The question was raised about where you can create this pressure, and I say "both ways". Members can point this out -- corporate and individual members -- and press to have it changed. Outsiders can make a fuss, point the fingers, and continue to pressure Sun to fix this. That's what you can do usefully from the outside.

  I'm not the political activist here, I don't have (yet anyway) a plan of action to pressure Sun. But that doesn't mean it can't be done. It just means that Ken Arnold doesn't have a plan. It's still better to act in ways that we can rather than giving up, because Java is going to be around a long time.

  I'd better stop here or I'll start throwing out all sorts of ideas about how to address this, and that will take more thought than a random idea explosion. ;-)

  Posted by: arnold on May 21, 2006 at 02:56 PM

• Ken, I think a third of the individuals who put their names in last election won a seat. Too bad that's just one person. The fact that so few ran is why there are so few individuals. Unfortunately, they don't get paid anything, and the office looks like it takes a lot of time for dead boring meetings plus time reading specs they might not care about. Have you thought about putting your own name in?

  Posted by: dwalend on May 22, 2006 at 10:56 AM

• Now that is truly unfair -- asking me to put my money where my mouth is. :-)

  No, I haven't thought seriously of that. I couldn't possibly afford the time, it isn't important enough to my startup company that I do that. But I don't see these problems as being universal, just very common. Two folks are somehow managing.

  Although I think you have it backwards. Or sideways. It's not about whether more individuals running would get more individuals to join or vice versa. It's really a matter of attitude and behavior in the community as a whole, and the leadership (mostly Sun) in particular. Individuals will matter when Sun decides they should make it so, and the question is how to help them reach that good decision.

  Right now it's hard to join as an individual and it's hard to be an EC member. Until they make it easy to join, it won't matter how hard it is to be an individual member on the EC. Until individuals (and smaller companies) are represented in the process with respect and power the motivation to join will be low. These are merely symptoms of the real problem: The JCP doesn't view the little folk (individuals and smaller companies) as very important. If that changes, those problems will be solved because they won't be acceptable.

  What form the solution might take is an interesting question. It's one people often get caught up in. "How could you make a standard with 1.5 million JCP members?" they ask. While it is a smart question, it will only actually matter when the JCP's question is: "Given that we need to have smaller companies and individuals, how will we make a standard with 1.5 million JCP members?" When posed that way, it will be solved.

  Posted by: arnold on May 22, 2006 at 11:19 AM

• Pardon me, but I strongly believe we are treating individuals with respect. Please point out where not. Individual members have the same standing as corporate members; the same rights and obligations. Show me another standards body where that is the case.

  Currently of the 1050 or so members, over 65% are individuals. This doesn't seem to indicate a reluctance of the JCP (or Sun) to welcome individual members.

  With regards to whether individual members are appropriately represented on the EC I would point out that it seems an oversimplification to treat throw all individual members on one heap and all corporate on another and make assumptions that the desires, needs and interests in each heap are uniform. Slicing and dicing the membership is a rather tricky exercise. What about the balance between vendors and ISVs, what about end-user companies vs hardware and software companies, and so on and so forth. The world is a spectrum of many shades of grey in between black and white.

  With regards to the Jini Decision Process. That's a bit apples and oranges. It's a mechanism to decide on a final specification. It is not a mechanism of how you get to that point. It would be interesting to see how the JDP would hold up if Sun's competitors would get engaged in it as they are in the JCP.

  With regards to last question: I argue that we are making a standard with 1.5 million developers. Even more, with 5 million developers. If you are participating in java.net on the various projects such as JDK community and Glassfish, you are helping forming the standard. If you are participating on the many Apache projects related to JSRs then you are helping forming the standard. Just as no open source software project can tolerate millions of committers, neither can the Java community survive with a 1.5 million spec leads.

  Posted by: onno on May 22, 2006 at 01:06 PM

• 
  Pardon me, but I strongly believe we are treating individuals with respect.

  
  in other words, you don't see a need to change anything?

  
  we are making a standard with ... 5 million developers

  So, how come that so many developers were caught by surprise when they finally saw generics in 1.5? Were these some other 5 million developers, the ones the Java community can not tolerate?

  Posted by: ewin on May 22, 2006 at 03:17 PM

• There are two possible meanings for "respect" in what I said. You took the one I did not mean. I did not notice the ambiguity when I wrote it, sorry. I did not mean that you were disrespectful to individuals.

  What the JCP is not is fully inclusive. As an individual or small company I can discuss anything I like, but the final decision happens in a place where my interests as a stakeholder are not sufficiently represented. That is true whether I speak of myself as an individual or as a member of a small company.

  Beyond that, as others have noted, becoming an individual member is actually quite burdensome. It is not something that one ought to do lightly -- the legal document is hard to understand and even if understood it is not a trivial commitment. And as an employee, I have to convince my company to sign an agreement to allow me to become an individual member. (Not to mention that there are two different versions I may have to sign.) This is not a user-friendly system.

  This second issue also reduces the ability of individuals to be represented as stakeholders in that fewer of us will join in the first place.

  Note, for example, that you are saying that about 700 individuals have joined the JCP. 15,000 people attended JavaOne, and the vast majority of Java programmers cannot attend. Doesn't 700 seem like a pretty small sample? I know most folks won't want to be standards jockeys, but the drop-off from 1,500,000 to 700... And how many of them participate? If they don't, is it because they don't care about the issues, or because they don't see the point? If I had a product that was reaching 0.047% of my potential audience, I wouldn't be feeling terribly successful. Even if we suspect that only 1% of Java programmers might want to join, you're only picking up 5% of that 1%. Why isn't it more?

  I said myself that the Jini process wasn't directly transferable, but it might be informative. Although since you ask what might happen should competitors get engaged, it is something we discussed long and hard. We believe that if competitors get involved who are not hostile to Jini the system will work fine -- if they want to move it in a way that Sun doesn't, they may very well win. That's what an open system is about. Sun has some special recourse for now but not much, and that's by design. The idea is to protect Jini, and not necessarily Sun's vision of what Jini ought to be at any given moment. (Hostile attack was also discussed, but this is starting to turn into a discussion of the Jini process.)

  Becoming a member is not the same as becoming a spec lead. Becoming a member should mean that you have a voice that respects (at some approximation) your value in the community. You can't do just what Jini did. But I think you can change to allow people more voice in discussion and approvals. A large reason Java is successful is that individual programmers insisted on using it, sometimes over managerial objections, sometimes not, but it was largely a pull from the developer side. So it seems to me like they are cumulatively a big stakeholder with a small voice.

  Posted by: arnold on May 22, 2006 at 03:30 PM

• Obviously, as the individual who just got elected onto the EC, I'm biased. I do think however that there are a lot of misconceptions in this thread.

  There is nothing from stopping individuals joining, beyond the fact that it's a PITA to fax 11 pages. What you gain from joining however, is entirely up to you.

  The JCP isn't an organisation that 'gives' you many tangible rewards. You don't get a free wiki, an open source project, commit rights, or anything fun like that. All being a member does is allow you participate.

  Participating however is often not that easy. Sure, you could just vote in elections, that bit is certainly easy. The meat of it though is in working on specs and expert groups. That job requires time and effort (assuming your track record says you have something valuable to say in the first place). How many people are willing to put in that effort? Not very many, it turns out.

  Sure, the vendors all jostle for position, power, and to get their agenda pushed through. Chuck in a whole bunch of them though and you'll find that no vendor gets very far when they push an agenda that does not aid the community at large.

  Most java developers do not (and should not, to be fair) care much about the JCP. It's the exact same reason that 1.5 million people do not go to JavaOne, or download Mustang betas, or post in any online forum. Does this mean that the Java community is broken?

  There is absolutely nothing stopping anyone from submitting a JSR, provided they can convince enough people on the EC that it has some merit. I talked to a bunch of people during JavaOne and encouraged them to do just that. These aren't corporations, these are open source developers you've likely heard of.

  It's a lot of fun to whine about a process you're not involved in, hell, I do it all the time about practically everything, but when you do so, at least try to ensure that your complaints have some basis in fact!

  Finally, I think an EC full of individuals would be a terrible thing. The reason that Java specs are taken seriously has nothing to do with individuals, and everything to do with corporate backing. I shudder to think of what sort of mayhem a '12 wise men' EC would cause!

  Posted by: hani on May 22, 2006 at 03:55 PM

• Hi, Hani. Thanks for being on the EC.

  
  I beg to differ: It isn't faxing 11 pages. It's reading 11 pages, comprehending 11 pages of daunting legal text, accepting the contract described in those 11 pages, and getting your employer to read, comprehend, and accept the part that they must sign. I know from my work in the Jini community that each of these is a real challenge, and the bigger these challenges the harder it is for folks to get involved. It's a real drop-off in participation.

  
  The employer one is particularly bad because it means that my company has spend legal resources and business risk on something that brings it no obvious benefit. I might be willing to risk resources based upon an amateur understanding of the JCP requirements. My employer is not likely to be so careless.

  
  Oh, and I might have to do this with two different agreements depending on which JSRs I want to get involved in.

  
  The rest of what you say rests on two assumptions: That few people care, and that their involvement wouldn't add much value. The first assumption is partly rebutted above: That people aren't signing up can be proof that it's too hard, not that they don't care.

  
  That first assumption is also rebutted by many counter-examples. Major open source technologies attract many folks to their design discussions. Apache, linux, tomcat, emacs, the GNU project, Perl, etc. Java is core to many people's lives. Why wouldn't they care? Further, they comment an awful lot on the JSRs. They just don't do it in the JCP. Why not? It isn't because they aren't interested in having an influence, it's that they find it hard to have an influence.

  
  The second assumption I reject out of hand. I think that it matters to have these people involved. I think it diminishes the community to have them on the outside. I think it hurts the standards themselves. I think it hurts the position of Java in the market: As people get alternatives in any market, they tend to reject major players that they feel don't listen to them.

  Finally, I think an EC full of individuals would be a terrible thing. The reason that Java specs are taken seriously has nothing to do with individuals, and everything to do with corporate backing. I shudder to think of what sort of mayhem a '12 wise men' EC would cause!

  I don't know what you think the "mayhem" would be. Pray elaborate.

  The Java specs are taken seriously because people follow them. People follow them because standards are good and valuable. Their adoption is a complex web of many factors. Sure, corporations can sometimes impose standards on others, but they can also fail to do so no matter how hard they try. It just isn't this simple. The history of standards adoption is full of examples of failed corporate-backed standards.

  
  By the way, IETF did pretty well with a strongly individual-friendly governing structure. Good thing, too -- otherwise we couldn't have this debate via an internet that wouldn't have existed.

  
  It's a lot of fun to whine about a process you're not involved in, hell, I do it all the time about practically everything, but when you do so, at least try to ensure that your complaints have some basis in fact!

  Hani, I think you are barking up the wrong tree here. You present many opinions ("mayhem", "should not care", "it turns out") but do not contradict specific facts. What facts did I get wrong?

  
  I suggest you look around at many other communities. Gather facts about how other people actually get things done. You might see how much it benefits a design process to have many smart brains involved, including those who aren't wondering if a feature that helps company X but not company Y should be supported even though it costs company Y money to implement, but instead wonder if the feature is good or bad.

  
  When we did this research as part of designing the Jini process we saw a lot of successes. And our designs were better, and also further improved by comments from folks who chose to get involved because they cared. Even if they didn't get paid for it (some did, some didn't). It works. That's a fact.
  

  Posted by: arnold on May 22, 2006 at 07:35 PM

• Ken, your maths don't hold up. The 15000 JavaOne attendees are not all individual developers. The majority of them work for IBM, BEA, Oracle, Sun, Fidelity, Schwab, Walmart etc etc etc. Most of these companies are members of the JCP and thus these developers are included. Same argument for the 5 million developers.

  With regards to the JSPA, its complexity and that your employer has to sign something if you want to join on personal title: yes that is difficult and hard and it is a tradeoff. If we wouldn't have that paperwork then many, if not all, of the corporations in the JCP will not want to have you on their expert groups because of fear that your employer may sue them because you contributed something that your boss really feels is theirs.

  So the IETF has some individuals too. Good for them. Last time I checked they didn't have millions of members either. Nor does the Jini community for that matter. I like the Jini technology and its community is a delight to work with. However, my remark about the JDP was not with Sun in mind. The Jini community is quite homogenous with not many direct competitors. It has not have to deal with much controversy about its technical direction (everybody basically follows what Sun does). The diversity in the JCP is much greater and the JCP has shown that it can work through controversy that comes together with that diversity. And yes, it can surely be improved but I don't believe that an inability for individuals to join and participate is present nor that that would be JCP's current biggest challenge.

  I must also just disagree with your opinion on the value of corporations in a community like the JCP. I hold that Java (in the EE space, the ME space, and the SE space) would not have been the success it is without the commitment and contributions to the community by IBM, Oracle, Nokia, Motorola and the many other big corporations.

  Posted by: onno on May 23, 2006 at 08:29 AM

• Ewin: you were surprised by a JSR that took five years!?

  Posted by: onno on May 23, 2006 at 08:30 AM

• Many, many developers, including me were indeed very surpised about the clusterfsck that generics turned out to be.

  Many probably not even knew that there was something like this in the queue, and many probably don't know what the JCP or a JSR is.

  Among the ones who knew it, people like me trusted the JSR members to get it right, and they got it terribly wrong. And like many others I did not monitor the JSR, because it is very difficult to monitor a JSR in general, and that one in particular, if it is not you full-time job. Like many others I didn't and still don't know the JSR numbers by heart, and the fscking brain dead JCP web site does not even provide a simple search function to find things.

  And once you find some JSR you still have to guess if that one is really relevant or if it is just a brain fart of some corporate drone. With several hundred JSRs it is impossible for an individual to monitor them all just for good measure.

  Your five year figure for the generics JSR is misleading. People who were aware of the JSR were probably guessing that generics will never come, because "nothing" happened The JSR page didn't change for three years somewhere during that five year duration. People do not monitor a JSR web page over five years, especially, when there are probably only five to ten changes of the provided information during such a time, as it is typical for a JSR web page.

  The issue got warm the moment someone decided to really include the change in a Java release, not the moment the JSR started. The issue got hot for developers when the first 1.5 betas were available. I would even guess that this was the time when many developers heard about generics for the first time. It was the first time developers saw that the generics stuff was ment serious and when they could evaluate it in practice. But that point in time was fare to late to get things changed.

  The generics JSR was extremely bad also because someone apparently decided to just provide the language changes in terms of changes to the JLS. The proposed draft was even worse. It was just a changed JLS copy, with piecemeal changes all over the place. The JLS is a difficult read, and it becomes a horrible task if you have to pick generics changes out of a full JLS and try to interpret them. The committee did not publish a rationale, the committee did not publish information about the implications of their design decisions. The committee did not publish progress information. One had to painfully infer what was going on from language description fragments.

  The generics JSR is a fine example of the closed-door policy of the JCP. Decissions were made in the back room. The absolute minimum information required by the JCP was officially made public to non JCP members. The results from the JSR work were thrown at the community, without background information. A few people voted on it, and BAM!, five million developers had to swallow it.

  Your statements indicate that you find this ok and changes are unlikely to happen. But next time McNealy calles on developers to get involved and to join Java-related organizations someone should tell that guy that things are not fine for the garden-variety developer when it comes to participation in the corporate-infested and controlled JCP. Someone should tell that McNealy guy about the action and word thing.
  

  Posted by: ewin on May 23, 2006 at 03:25 PM

• OK, ewin is being pretty pungent, but that's the basic gist of things with the generics JSR. Even when I was at Sun and signed up as a member I could hardly follow several of the JSRs that interested me without sending personal mail to the spec leads, who basically were nice to me.

  The JCP is a relatively closed system, both by culture and by rules. I know that there are legal issues here, but once I've signed the relevant agreements I don't know why specs are so shut down. People seem more worried about the possibility of pointless feedback than about missing valuable feedback. That's just weird. I've lived in many public design situations, and the occasional problem child is many times overweighed by the value of the feedback of passionate people you didn't think to involve, or the massive feedback of casual users who, as outsiders, can tell you when you aren't doing things that serve them.

  As for the legal issues, I'm sure there are hard things to solve, and major stakeholders to work with. But right now the system is very overweighted to these players. The legal issues seem more like an afterthought. The overall attitude seems to be that it's a good thing that the process is walled off the way it is because a lot of individuals or smaller companies contributing would be a pain. Therefore it's OK that it's hard for them to be involved, to contribute, to have a final say.

  Notice the order. If the legal issues caused the situation, wouldn't we hear something more like: "The legal situation forces us, but it isn't as bad as it sounds because..." And you certainly would hear "...and we're trying to counterbalance this problem by ..." But those inside the process don't think it's a problem at all. The legal situation follows from the attitude. The legal situation doesn't make JSRs closed to member discussions. The contract doesn't make it hard for individuals and small companies to be represented on the EC. Those are policy decisions that come from the same approach that the legal documents come from.

  So while I accept that there are difficult legal issues, I think that the fundamental attitude of oligarchy can be changed independently. And the legal issues would lessen because it would matter to those hiring the lawyers. I don't believe the legalities are the cause of the problem. Change the attitude and the lawyers will follow as far as they can.

  Posted by: arnold on May 24, 2006 at 04:10 AM

• So what really, really bugs me in the things that Onno and Hani have said is that it isn't hard for individuals to join. This is a pretty major disconnect from the facts on the ground. It is hard. It is burdensome. It is well nigh impossible for most folks. You may think that that's a good thing, you may think that it's an inevitable thing, but it's how things are. To brush this off as a triviality -- that folks who care are really just too lazy to do a pretty easy thing -- is just nuts.

  Second to this, the next idea I'd nominate is that IBM employees (say) are included in the process because IBM is on the EC and expert groups. Employees of IBM have their own opinions, and may disagree with the technical and marketing decision makers at IBM. They may even be more correct on some issues. If IBM employees, through some democratic system, chose their representatives you could make such an argument. I'm pretty sure that's not how it works. (This reminds me of one argument against women having the vote: Their husbands/fathers/brothers had the vote so they didn't need an independent vote because the males could take their views into account.)

  In some sense, I feel more comfortable with the arguments that people outside the process just aren't relevant and would be a nuisance. As wrong as I think this is, at least it seems to actually represent the guiding principles of the process design. We can discuss whether this belief is correct or not and why. If the right people were convinced that this was not correct, the process would change. Arguments like those above are really a form of brush-off that flows from the belief, so let's stick to the belief itself.

  Posted by: arnold on May 24, 2006 at 04:33 AM

• I didn't say it is easy for individuals to join. I said it is possible and that there is currently a comparatively high level of participation by individuals. I explained why some of the mechanisms are there. I believe we disagree on the need of such mechanisms and that is fine. I said several times that individuals are important and that their time dedication is appreciated. I also find it important and appreciate the corporate participation. I am not willing to choose between these two categories in importance. Nor do I think this is the only way neither necessarily the most important way to look at the membership and its diversity. Equally important for me is a balanace between US and non-US, vendors and end-users and so on.
  With regards to the ability (or the lack thereof) of IBM employees to speak with their voice in the JCP or not: sadly (or luckily) that is not my problem to solve but something for you to discuss with Palmisano :-) The JCP does not prevent it (to give just one example: Geir is both an Intel employee, ex-IBM employee and ASF's rep in the JCP).

  Regarding JSR 14, Generics: Ewin, thanks very much for the frank feedback. Just on the issue of search: that feature will be brought back in a web site update we're planning for July.
  

  Posted by: onno on May 24, 2006 at 11:03 AM

• As for JSR 14, count me in - I was caught by surprise too, even though I reckon Generics is a good thing. I have turned to use Java as my major language since Tiger, mostly due to its new features.
  But still, I have never noticed that this Generics feature has actually been planned "in the public" and introduced in accord to a 5 years plan. For me it might as well have been an over-night decision.

  Posted by: alexlamsl on May 28, 2006 at 04:21 AM

• I think the pretty low quality of the specifications (at least in the J2SE area) coming out of the JCP are a very good indicator how flawed the current process is. When I see stuff like http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=4350444 in something that's a so-called official specification, I have to shake my head at the lack of professionalism at the J2SE JSRs. The specification testing team needs to file bug reports because the J2SE JSR has been too lazy to actually specify how an API works? And it takes only 4 years for someone to reply to the bug report? With a WONTFIX?? In an official specification??? That's astonishing.

  My impression is that all of the legal clusterbombs piled up before someone wishing to participate in a JSR serves for one thing alone: giving the proprietary Java vendors unfair advantage over people & companies who don't (want to) belong to the Java oligarchy. By making external oversight and participation non-trivial, the powers that be can ram through whatever they can get away with, rubberstamping half-finished, amateurish work as official 'standards', while at the same time securing a time-to-market advantage for licensees of the RIs over independent implementations.
  I don't blame that on malice, though. I blame it on incompetence. The major problem Java has, is that due to the incomprehensible lack of a way to have something akin to package & dependency management, every Java vendor desires to bundle whatever useless thing they can come up with into the core platforms. Did Java really need a Swing HTML widget? Does it have to stay largely unspecified for almost 10 years now? Who's responsible for that and why don't they fix it already? What's taking so long? Why doesn't JCP have any sort of QA, afaict?

  The most depressing thing about the JCP, for me, is that even though Sun is great at chest-thumping about transparency, when one looks at Sun-led J2SE JSRs, they are not transparent at all, with no apparent desire to correct that. There are very few JSRs that actually are transparent beyond the token gestures required by JCP 2.6, and they are in general led by people outside Sun, like Doug Lea, funny enough.

  And even the JSR 277, the widely hailed one to actually bring some vague notion of modularity and dependency management into Javaland, is completely mismanaged by Sun. There is no information whatsoever on its progress since the JSR has been announced to PR fanfares. No public mailing lists. The folks on it must not talk about it. It's just some more vapourware, as far as anyone outside the inner circle at Sun can tell.

  I have no doubt that JCP could be great and useful, and worthy of participation. But in its current form, it's pretty far from that ideal. And unfortunately, there is no JCP 2.7 under way to fix its many, many issues.

  Posted by: robilad on June 02, 2006 at 07:01 AM

• It is not directly JCP related, but I'd like to add to robilad's comment above about the effectiveness (or lack of) in the Sun Java camp.

  Back in 1999 a RFE was placed on the sun.misc.BASE64Encoder to try and open these, basically very generic, utilities up for the public. Today, some 7 years after the item is still unresolved on SDN:

  http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=4235519

  Incedently, there are multiple scattered implementations all over (Sun even made another one in javamail). For the love of god Sun, java.sun.codec would be a nice place for such a generic thing no?

  Posted by: mrmorris on December 11, 2006 at 06:30 AM

• Nice...
  http://www.impact-fellowship.org/_cusudi/0000007a.htm
  http://www.chaco.gov.ar/meccyt/subsecyt/_act1/0000060b.htm
  http://www.chaco.gov.ar/meccyt/subsecyt/_act1/00000353.htm
  http://www.wapug.org.uk/_CoP_discussion/00001ca9.htm
  Harvard - Harvard

  http://www.wapug.org.uk/_CoP_discussion/00001ca8.htm
  Stanford - Stanford
  http://washington.uwc.edu/about/faculty/rybak_c/webpage/_disc10/00005663.htm
  http://orgs.salisbury.edu/fishing/Discussion/0000737f.htm
  Yale - Yale

  Posted by: jamesdalton on January 21, 2008 at 01:13 AM

This work is licensed under a Creative Commons License.