John D. Mitchell's Blog

Sun's Phipps rants on Raymond's open-source rant

In an article in the UK's PC Pro magazine, Sun fires back over Open Source Java accusations, Simon Phipps, Sun's Chief Technology Evangelist and java.net blogger, rails about Eric Raymond's recent open letter to Sun, Let Java Go.

Seems pretty clear that both sides make some points and miss their marks. I don't have the time right now into a blow by blow analysis but I will say that I'm personally sick and tired of Open-Source fanatics saying that everything should be open sourced. It's Sun's property and they can do whatever they want to do with it (whether we like it or not). On the other side, I'm equally sick and tired of Sun saying that they do NOT have any extra level of control over the rest of the JCP powers -- that's patently false -- check out my earlier blog on that particular subject, Open, Independent JCP?.

Comments

• Blow-by-Blow analysis
  To save both you & me the time, Charles has done one - http://cld.blog-city.com/read/492917.htm

  S.
  

  Posted by: webmink on February 18, 2004 at 04:06 PM

• Lack of clarity is preventing serious discussion...
  There appears to be a serious breakdown in communication between both 'sides' in this argument.

  First of all, we need to establish what ESR is actually proposing. Easy, some may say. He wants 'open source Java'.

  But what is 'Java', exactly? And to what extent should it be 'open source'?

  Remember, 'Java' refers to two things:
  1) The Java programming language, including the core syntax and the APIs for the various editions (ME, SE, EE).
  2) The Java platform, ie. the VM and to a certain extent the implementations of the APIs.

  The term 'Open Source' can be used in a fuzzy way, too. The BSD license is open source. Is the GPL? It is often described as a 'Free Software' license, but can it be both 'Open Source' AND 'Free Software'?

  The effects of licensing under BSD are different to using the GPL, yet many consider both to be Open Source licenses. Was ESR referring to the GPL in his letter, or suggesting the use of ANY non-proprietary license?

  It would be more sensible to review each license *on a case-by-case basis*, instead of simply lumping all open licenses together. They are not the same, and as such must be discussed individually.

  So, before any serious discussion of the merits of 'Open Source Java' can begin, concrete definitions of terms must be agreed on. Most online forums currently attempting to discuss ESR's comments have failed to do this, and consequently have not reached any kind of useful or even rational conclusion - not because the participants are illogical, but because they are working from different definitions.

  Posted by: philwebster on February 19, 2004 at 02:43 AM

• Lack of clarity is preventing serious discussion...
  When ESR talks open source he mean completely and utterly open, anything goes without any controlling bofy (except maybe he himself).

  So both 1) AND 2) from your list would be open and anyone would be able to change them.

  Maybe the OSF would take stewardship (read "complete control") over the language specification.

  In the end it will mean the demise of Java as a plethora of competing and incompatible JVMs and language extensions flood the net as ever more people and companies start releasing their own version with "enhancements" and "fixes" to their pet (usually perceived) problems and shortcomings.

  Posted by: jwenting on February 19, 2004 at 03:32 AM

• Lack of clarity is preventing serious discussion...
  Both BSD and GPL allow code to be completely 'open', yet have drastically different effects on the behaviour of developers.

  We still need to decide *exactly* which license we are discussing, before claiming that certain things would or would not happen.

  It might also be interesting to have a related discussion on partially open Java, ie. the merit of having a Libre implementation of the Java Platform (VM and JFC APIs) while still leaving the Java Language specification under Sun's stewardship.

  Posted by: philwebster on February 19, 2004 at 04:25 AM

• Free Software / Open Source implementations of the Java Platform
  A few facts need to be established in order to make debate over Open Source/Free Software 'Java' more rational.

  1) The source code to parts of the JFC is available when you download the J2SDK.

  2) The source code to the Java 2 SE platform implementation is available, but under a non-free licence.

  3) Incomplete Open Source/Free Software implementations of Java JFC APIs exist, for example GNU Classpath.

  4) Open Source/Free Software implementations of a JVM also exist, and can work together with the Free APIs.

  Following on from 3) and 4): Effectively, there already are Libre implementations of the Java Platform, although they are not at the same level as the current language/API specification.
  

  Posted by: philwebster on February 19, 2004 at 04:34 AM

• Blow-by-Blow analysis - ESR's point on stock prices.
  Raymond's letter states:
  "Open source is hardly a zero-revenue model; ask Red Hat, which had a share price over triple Sun's when I checked."

  Some critics have focused on the latter half of this sentence: the part comparing the relative value of Red Hat and Sun shares. It was probably not a good thing for ESR to say, because it detracts from the very issue he is trying to raise.

  The point is not that Red Hat or Sun is worth more than the other, but that Libre licensing does not imply zero-income. You can still make money from Free Software, either by selling copies of it, or selling services to support it. Red Hat is an example of this, and ESR mentioned them simply to support his argument.

  Perhaps it would have been wiser for ESR to say something like the following:
  "Open source is hardly a zero-revenue model; ask Red Hat, which made (X amount of money) in the last financial quarter".

  That would have communicated the point *much* more effectively, without dragging the discussion of a serious and important topic down to the level of pedantic fallacy-spotting or corporate fanboyism.
  

  Posted by: philwebster on February 19, 2004 at 04:54 AM

• Blow-by-Blow analysis - ESR's point on stock prices.
  "It was probably not a good thing for ESR to say, because it detracts from the very issue he is trying to raise."

  Not to mention the fact that it's a completely ignorant statement. All by itself, current share price is just about meaningless. You can't infer anything from it. Only when you have additional information, such as number of shares outstanding (to judge the total value of the company) or historical share price (to judge performance) does it mean anything. Heck, Sun could do a 1 for 4 reverse stock split next week, and then their price would be above Red Hat's. Would that mean they were suddenly 4 times as "good" as they are now? Of course not, because the number of shares outstanding would drop to 1/4 of the previous number.
  

  Posted by: kdenehy on February 19, 2004 at 06:41 AM

• Blow-by-Blow analysis - ESR's point on stock prices.
  I notice that ESR has now added this footnote to his original letter:
  "Note: This letter has attracted a lot of half-bright criticism to the effect that should have compared Sun and Red Hat by market capitalization rather than share price. Wrong. Market cap measures present value; share prices measure investors' expectations of future revenue. The latter is the correct measure for this context."

  However, my objection to the share price comment is not that is it wrong, but that it is irrelevant.

  Regardless of whether or not Red Hat is a 'better' company than Sun, the fact remains that Red Hat makes money from Open Source/Free Software, and it is that fact - and not share comparisons - which provides some support for ESR's argument against equating Libre with zero-revenue.

  Posted by: philwebster on February 19, 2004 at 07:23 AM

• Blow-by-Blow analysis - ESR's point on stock prices.
  Eric's footnote doesn't make his argument any better. His statement about share price is almost right, but he still misses the mark. First, share price is (according to market ) based on the sum of all expected future earnings, not revenue. More importantly, it's based on expected earnings per share. Calculating earnings per share is, of course, a simple matter of dividing total earnings by the number of shares outstanding. Since market cap is nothing more than the share price times the number of shares outstanding (notice the inverse operation), then market prices is, just like share price, related to the expected future earnings.

  Of course, java.net is not a stock market forum, and I don't intend to split hairs or discuss market theory. What's telling, though, is that Eric's misunderstand of the relationship between stock price, earnings, and market capitalization hint at a greater misunderstanding of the market economy and capitalism in general. That's to be expected, since open source fanatism and mistrust of capitalism do tend to go hand-in-hand.

  Posted by: jimothy on February 19, 2004 at 08:12 AM

• Blow-by-Blow analysis - ESR's point on stock prices.
  "Eric's footnote doesn't make his argument any better."

  Perhaps not, but it doesn't make it any worse either. In fact it makes no difference at all, since the share comparison is irrelevant to the point ESR was making - that Open Source / Free Software can be used to make money.

  Is this point even worth discussing further? I'm suprised that so many people seem to focus on this one aspect of Raymond's letter, when it is quite clearly only a very minor side-note.
  

  Posted by: philwebster on February 19, 2004 at 08:25 AM

• Blow-by-Blow analysis - ESR's point on stock prices.
  "open source fanatism and mistrust of capitalism do tend to go hand-in-hand."

  There is no need to resort to thinly veiled personal attacks in this discussion. Whether or not you consider ESR to be a fanatic is irrelevent to the topic of whether or not Java sould be Open Source / Free Software. And even if he *was* a fanatic, that wouldn't stop him being right if evidence and reason happened to be in his favour.

  Secondly, it is worth noting that not all advocates of Open Source / Free Software are fanatics, nor are they all mistrustful of capitalism. Capitalism and Open Source are not mutually exclusive ideologies.

  Judge ESR by his arguments, not by his beliefs.
  

  Posted by: philwebster on February 19, 2004 at 08:42 AM

• Blow-by-Blow analysis - ESR's point on stock prices.
  "Secondly, it is worth noting that not all advocates of Open Source / Free Software are fanatics"

  I wasn't making these statements of advocates, but of fanatics.

  Posted by: jimothy on February 19, 2004 at 11:36 AM

• Blow-by-Blow analysis - ESR's point on stock prices.
  Sort of off topic: Don't people realize that share price doesn't mean as much as *market capitalization* in many respects? Sun's market cap is 6 times that of Red Hat. Does that say any more or less about either company's business model or revenue sources than share price? Don't think so. Leave the stock market out of these discussions - unless some M&A activity is going on...

  Posted by: brintoul on February 19, 2004 at 03:23 PM

• in case sun goes bust
  Perhaps it's because everyone is woried that
  with sun's financial performance the platform
  ( and our personal investments in adopting it )
  might be put in jeopardy if someone were to do
  takeover on sun.

  And a lot of people think that that is a posibility .

  By open sourcing the java product it would be
  like an escrow mechanism that would mean
  we couldn't be held over a barrel.

  Whether anyone would want to take over sun is another question.

  Posted by: huntbryan on February 19, 2004 at 05:12 PM

• Meaningful discussion

  One of the biggest problems I have when the concept of open sourcing java or opening up java in some way is that it seems the discussion is very much one way. I see many comment about how the open source advocates are at it agian along with a few stock responses to the subject.

  1.) MS will fragment java.
  2.) Java will fragment in general
  3.) Its sun's technology do with as they please.

  1 and 2 are really speculation no one can predict the future and its not clear since Sun has never proposed a potential open java license it is impossible to even consider how likely the fragmentation is. As far as I'm concerned the license could exclude any company or individual which broke previous java licenses.

  My point is its pure speculation yet treated as gospel.

  As far as Java being Sun's technology to do with as they please I really disagree with this statement. Sun has presented java as a "open" standard technology for wide adoption since it was first released. A lot of people and companies have supported java based on Sun's commitment to openess. If you said the same about .NET outside what MS has offered to standards commitee's then sure I'd agree. Oracle's database products sure. But java was never presented or sold as the exclusive IP of sun.

  Agian there are certianly a number of good arguments and even more important looking at Sun's own actions to view java as community property. SCSL stands for Community Source License right ?

  Finally the name calling "open source fanatics" I advocate vigrously for Sun to open up J2SE because I think thats the best way for J2SE to succeed yet I've faithfully licensed sun source ever since it was avialable and also paid thousands of dollars for testing and certificaiton. Does this make me a fanatic simply because I believe that opening up J2SE is the best thing to do ? I hope that one can be a SCSL licensee and yet still advocate for change without being branded a fanatic and ridiculed and flamed by Sun employees all the way up to Suns top executives. Can't we instead attempt to engage in honest discourse on the possibility of opening up java.

  Posted by: memmel2 on February 19, 2004 at 07:03 PM

• Some modest proposals
  Sun has every right to do which its software what it wants. Eric S. Raymond has of course the right of free speech -- which must not be necessarily well prepared speech.

  But maybe Sun could think about following proposals:

  1) Modify the J2SE SDK license in a way that it allows redistribution together with open source implementations of the license.

  2) Modify the media label requirement in a way, that allows the distribution of the SDK in platform dependend packages.

  3) Update the documentation for the Java language, VM and the new JVM Tool Interface to reflect all the changes of the 1.5 "Tiger" release.

  Currently there is not even a single document specifying the Java language as supported by the 1.5 "Tiger" release.

  I'm quite certain that these steps could help to grow the application and development space for Java more.

  Posted by: kunitz on February 20, 2004 at 02:17 AM

• Meaningful discussion
  " I hope that one can be a SCSL licensee and yet still advocate for change without being branded a fanatic and ridiculed and flamed by Sun employees all the way up to Suns top executives. Can't we instead attempt to engage in honest discourse on the possibility of opening up java."

  That is what Java.net is for. Let's discuss some scenarios and see where it leads.
  

  Posted by: philwebster on February 20, 2004 at 02:27 AM

• Scenario 1: Sun continues along the closed source route. Discuss.
  If Sun does not 'open' Java any more than it is today, what will happen?

  Bear in mind that Open Source / Free Software implementations of Java VMs and APIs are in development, and will continue to improve, regardless of what Sun does with their own implementation.
  

  Posted by: philwebster on February 20, 2004 at 03:04 AM

• Blow-by-Blow analysis - ESR's point on stock prices.
  ESR's footnote has changed. The text now reads:

  "Note: This letter has attracted a lot of half-bright criticism to the effect that I should have compared Sun and Red Hat by market capitalization rather than share price. Wrong. Market cap measures present value; the relative prices investors are willing to pay for shares reflect their expectations of future return per their invested dollar (that last qualification is what normalizes share volume out of the equation). I'm informed that this effect is particularly visible in IPOs and derivative trades, though my informant thinks prices of established stocks are too noisy for it to be reliable there."
  

  Posted by: philwebster on February 20, 2004 at 07:54 AM

• Have you actually read the letter?
  The author comments of being sick of Sun's control over Java is exactly the point Eric strssed on his open letter. So, what is this fuss about? This discussion is a lot of noise, each one defending his own position without understading the other arguments and replying to then.

  Open source software is not about the ability to see the source (as you can do with Java or Delphi libraries). It's not about being zero-cost (as many mistakenly concludes from the associated term "free software"). It's about control. Who controls the evolution of the software.

  You see that, although IBM, Oracle and etc have limited interest on marketing Java (because of Sun control of the brand and plataform) the JCP got some success on creating viable, vendor-neutral standards. You can also see that, although the Java plataform itself is not open source, there are lots of successfull open source Java projects like Jakarta and JBoss.

  There's even open source (free software as the GNU definition) Java implementations, being GCJ and Kaffe the most successfull so far. But these are far com being drop-in replacements for Sun's implementation. And, if they ever become such, it's very easy for Sun to render them incompatible with the next Java language, VM or SDK release. The same risk faces Mono and Dot.GNU developers when trying to "clone" MS C#, CLR and .NET.

  It's not enough to be compatible with the specs. The implementations must be interoperable. Sun JavaVM and SDK themselves do not follow Sun's own specs for them on some parts (these are documented bugs).

  The control issue is the heart of disagreements on the Java community itself, the reason companies involved on the JCP won't allways work together and the reason open source advogates ask Sun to open source Java.

  IBM created a huge market by not controlling the PC architecture. The Internet is such a big success precisely because there wasn't a sole controller of it's specs, protocols, or infrastructure.

  Java has (had?) the potential to have the same success as the PC and the Internet but the limiting factor is Sun's control over it.

  You see the whole IT industry today supports Linux, with very few exceptions. That's becaus Linux is open source, so no one is afraid of being double crossed on their investment on Linux. Has Java got such a broad acceptance? Risking being flamed, i'd tell no. Java does not have the same broad support from ISVs, hardware manufactures, and both corporate and end users as Linux has today.

  Open sourcing java is in the best interest of everybody involved, including Sun. Learn from IBM: which one did made more money for IBM: The PC or the PS/2 architecture? The open one, or the proprietary one?

  Maybe someone remembers the attempts made by Novell, Lotus and AOL to create a "global" network, before the boon of e-commerce on the Internet.

  The Internet is so successfull because it's open source. Java could also be and enjoy the same status of being everywhere.

  Posted by: flozano on February 20, 2004 at 08:24 AM

• Have you actually read the letter?
  "Open source software is not about the ability to see the source (as you can do with Java or Delphi libraries). It's not about being zero-cost (as many mistakenly concludes from the associated term "free software"). It's about control. Who controls the evolution of the software."

  What open source software "is about" depends on the audience. For me, as a developer, it is, partially, about having access to the source code. I may fix bugs, or I may just study the source to figure out what is going on.

  For a business or a department, being zero-cost IS often what it's about. Even nicer than not having to shell out dollars to acquire software is not having to stall a project while I get an appropriation signed.

  How much control does my employer have over, say, Apache? In reality, not all that much, and they proably want less than they actually have. It's not their "core competency".

  I think that to pragmatists, who make up a larger part of the software development community, the factors you feel are unimportant are more "what it's about" than "who controls the evolution of the software".

  Posted by: jimothy on February 20, 2004 at 11:46 AM

• Have you actually read the letter?
  Recall that IBM did not choose to open the PC architecture. The small piece that they controlled, the BIOS, was reverse engineered. The opening of the PC platform happened against IBM's wishes.

  People are free to reverse engineer Java; it's essentially what the open source JVMs that others have pointed out have done.

  Posted by: jimothy on February 20, 2004 at 12:08 PM

• ...
  One thing Sun could do to make me happy is to stop funding SCO...

  Posted by: pulse1014 on February 22, 2004 at 08:16 PM

• Funding SCO... off topic. Go elsewhere to discuss it.
  That has no relevence to this topic. We are discussing the arguments for and against Sun open-sourcing Java, as suggested in ESR's recent open letter.

  To discuss the relationship between Sun and SCO, go to a GNU/Linux forum. This is a *Java* thread, and posts about anything other than Java are off topic.
  

  Posted by: philwebster on February 23, 2004 at 01:42 AM

• 'Open source' is not subjective.
  I don't think the term 'open source' is quite as subjective as you suggest.

  Zero-cost is one aspect, but it is not always the case that 'Open Source' implies 'Zero-cost'. It is possible to buy Open Source or even 'Free' software. The flip side of this argument is that it is also possible to get zero-cost software which is not Open Source or Free Software.

  Open Source should be defined by what it does *differently* rather than by criteria which apply equally well to other development methods.

  Posted by: philwebster on February 23, 2004 at 01:52 AM

• Funding SCO... off topic. Go elsewhere to discuss it.
  Forgive me - I thought this weblog was concerning Raymond's "rant" about Sun and open source. I guess you didn't read the letters =)

  Posted by: pulse1014 on February 23, 2004 at 09:41 AM

• Lack of interest in Java in the O-S community
  One blog author wrote about Raymond's letter, and said that from a search on SourceForge, one could see that Java was a very popular language for open source projects.

  I disagree, and think Raymond is right on this one. Apache/Jakarta hosts a number of interesting, sometimes useful open source packages for Java, and has a popular development community. But Jakarta shies away from Java desktop development, for example, and outside of Jakarta, the Java open source development efforts are scattered and rather sparse.

  If you search on SourceForge, check out how many Java projects are actually production-grade, and then check out how many are actively developed. You will find relatively few compared to other languages. There are many many projects on SourceForge that are just ideas, prototypes, first stabs.

  A better gauge of this, I think, is to check out FreshMeat, which updates its page as project updates are released. There is a dearth of Java development, it's a tiny percentage of the total number of projects announced.

  Not only that, but PHP, which was developed after Java, has a vigorous, committed following despite the widespread belief that Java scales more reliably in the server environment, and despite the large range of libraries and kits available for Java server development--not even including J2EE packages. The most popular bulletin board/discussion groups are not written in Java, JSP or servlets. Now why is that?

  From what I can see, being an avid reader of various technical discussion groups on the web, Java is not widely popular outside the business-programmer environment, and is not widely popular in the open source space.

  Sun made a strategic decision to pitch Java as a business-friendly language, which has helped create an ecosystem for Java development in IT. But at the same time, they have lost out by not inspiring the many thousands of open source programmers who could extend, enhance, and improve the capabilities of the language. And Sun has something to do with that, and I think, should work to remedy it.

  Patrick

  Posted by: pdoubleya on February 25, 2004 at 07:00 AM

• Lack of interest in Java in the O-S community
  If what you are suggesting is true, then people who have countered Raymond's 'Java is not popular among Open Source/Free Software developers' argument have made a rather large error.

  To summarise your argument, you are saying that simply counting the total number of Java projects is not giving a truly representative estimate of popularity, since a large proportion of those projects are either dead or proof-of-concept prototypes.

  I would agree with you that a comparison of *active* projects is much more useful.

  It seems that you have already done some research on the subject. How about posting a URL, so people can see your evidence without the need to trawl through Sourceforge and effecively duplicate your research effort?
  

  Posted by: philwebster on February 26, 2004 at 02:31 AM

• Use of Java in Web-BBS systems
  "Not only that, but PHP, which was developed after Java, has a vigorous, committed following despite the widespread belief that Java scales more reliably in the server environment, and despite the large range of libraries and kits available for Java server development--not even including J2EE packages. The most popular bulletin board/discussion groups are not written in Java, JSP or servlets. Now why is that?"

  It may simply be because a BBS is quite a simple application which doesn't require any of the powerful features offered by J2EE or the J2SE libraries you mention.

  Most discussion groups on the web are hardly well implemented, and it isn't especially difficult to implement a subject-threading system anyway.

  BBSs are not complex enough to warrant the use of Java. It would be overkill.
  

  Posted by: philwebster on February 26, 2004 at 02:35 AM

• Open PC vs. closed PS/2 architecture
  Regardless of whether or not IBM opened the PC architecture, the question remains: which made more money for IBM - the open PC platform or the closed PS/2 architecture?
  

  Posted by: philwebster on February 26, 2004 at 04:07 AM

• IBM weighs in...
  IBM VP Rod Smith weighs in with IBM's support for creating an independent body to control Java.
  

  Posted by: johnm on February 26, 2004 at 11:16 AM

• Open Source / Free Software implementation
  He doesn't actually say anything about an 'independent body', the letter talks about Open Source *implementations* of Java.

  Or am I missing something?

  Posted by: philwebster on February 27, 2004 at 02:15 AM

• Open Source / Free Software implementation
  Rod Smith:
  "IBM would like to work with Sun on an independent project to open source Java."
  

  Posted by: johnm on February 27, 2004 at 08:12 AM

• Open Source / Free Software implementation
  "independent project" != "independent body"
  

  Posted by: philwebster on March 22, 2004 at 11:45 AM

This work is licensed under a Creative Commons License.