Same old, same old
|
| Author | Content |
|---|---|
| chappaquachap May 23, 2005 5:45 PM EDT |
I worked on Jikes, a Java source-to-bytecode compiler, from early 1996 through the end of 1999. It was released as IBM's first open source project in December 1998. By the end of 1999 I and my colleague Philippe were told it was time to work on something else at IBM Research. I could have continued working on Jikes part-time on my own time, but I realized then that the odds that Sun would ever "get it" when it came to Java and open-source were low indeed, in part because the "ECMA Java standardization effort" fell apart around then. So I stopped working on Jikes. Five and half years later, and Sun still doesn't get it. All this nonsense about "forking". The "fear of fork" is not what makes open source bad, but one of the main forces behind its success. There is no better incentive to do the best possible job you can to improve the code and strengthen the project than knowing that anyone else can fork the project and try to do a better job. I can recall waiting MONTHS to get an answer from Sun about a Java language specification question. Would that happen were Java open source? I doubt it. Spoken as a Jikes guy and not speaking for IBM, dave shields |
| peragrin May 23, 2005 5:53 PM EDT |
And people wonder why no one trusts Sun, is because of comments like this. OSS people asked for Java to be opened Sourced, What does Sun do create a new license and Open Solaris. IBM grants OSS developers access to a large number of Patents What does Sun do? releases their patents to their License and says GPL doesn't make up all OSS programs. Sun Pushes Open office to use java tightly, forcing one to install Java to use Open Office. Then complains when OSS users start making their own version of the supposedly 'Open" Java framework. Sun then gets upset because they don't understand why no one likes them. They contribute the most code right? Of course they contribute it in such a way as to piss off anyone who tries to use it. |
| cjcox May 23, 2005 7:04 PM EDT |
Most of the people at Sun fully believe their anti open champions Schwartz and McNealy. They believe with all of their heart that open source is wrong and bad for business as a whole. To summarize their position: "Duke, I am your Father" (in a deep voice with heavy breathing) Seemed appropriate. :) |
| PaulFerris May 24, 2005 2:43 AM EDT |
I'll say it again, Sun talks a lot of the talk, but they don't appear to be walking the walk. The politics of their downward slide make them an extremely dangerous entity. Things like this only make me more sure of the situation. --FeriCyde |
| incinerator May 24, 2005 3:29 AM EDT |
Aye, but I would not worry to much. Sun's going down, anyway. If not this decade, then the next. |
| sbergman27 May 24, 2005 8:28 AM EDT |
I predict that in the next few years, the old stand by section on the LWN Weekly Edition front page will change from "The SCO Problem" to "The Sun Problem". Sun is on their way down, but you can bet that they will go down clawing. I'll go further out on a limb and predict that McNealy will be one of the first casualties. He'll be hated at the time of his resignation (perhaps to spend more time with his family), but not nearly as much as the next CEO will be. Like Ransom Love, we'll wish he was back. Sun and SCO have a lot in common. (And here, I should probably explain that while I quite understand that from a legal standpoint, The SCO Group is really Caldera, and Old SCO is really Tarrantella, Inc., the Old SCO OS division was the heart and soul of Old SCO, and that, under new "management" is what the SCO Group really is.) Oh, and I almost forgot. Expect Sun to pull an "AOL" on Open Office, and spin them off into "The Open Office.org Foundation". The resulting slimmed down "OfficeFox" will thrive once cut from its (once) primary benefactor. They'll finally learn how to market themselves instead of treading lightly so as not to offend Sun by competing openly with Star Office. OK. I'm past the "limb" and out into the twigs, now. But I'll bet I'm right. Tune in 3 years from now, at which time I will either say "I Told You So" or totally disavow any connection with this post. ;-) |
| TxtEdMacs May 24, 2005 9:01 AM EDT |
"Tune in 3 years from now, at which time I will either say "I Told You So" or totally disavow any connection with this post. ;-)" But we will track you down, and then will you pay!! [;-))] |
| PaulFerris May 24, 2005 9:57 AM EDT |
sbergman: You forgot the obvious: What about tainted Slowlaris code in the Linux kernel... |
| sbergman27 May 24, 2005 6:43 PM EDT |
Paul: I thought that was kind of implicit, though I did not specifically point it out. My crystal ball advised me to leave a certain amount of wiggle room to allow for Heisenberg variances. |
| chappaquachap May 24, 2005 9:30 PM EDT |
To follow up on my earlier post. The idea of "forking" -- that someone can take YOUR code and do what they will with it -- seems alien to many, a cause for concern. But it's really a cause for celebration, one of the many factors that make open source such a powerful phenomenon. I first appreciated this in the fall of '99. I had decided to run the Jikes project as the sole contributor for the first few months. IBM Research wasn't really set up to allow others to log in, and I also decided I would make all the initial contributions personally, so if any problems came up then all the blame would fall to me, not to some pore sod who had committed something that brought IBM attorneys into the frey. However, around Sep '99 the folks setting up developerWorks came to me, as they wanted Jikes to be the first project in their "Open Source Zone". They had acquired sourceforge-like environment (as I recall they paid for commercial version of sf) so others could have access to CVS, there would be mail list, mail archives, bug tracker -- all the stuff we take for granted. So I had a chance to set up a core team , one with more folks from outside IBM than from within it. Then I realized I had a practical ethics problem: 1) IBM was only a minority member of the core team; 2) Jikes was run from a web site with a url ending in 'ibm.com'. It was very visible (that was the idea all along with Jikes, by the way: to show IBM could release a significant amount of code under an open source license, and run it out of the IBM domain to show we were willing to play by open source rules). 3) As an IBM employee I had to follow IBM's Business Conduct Guidelines. So though I trusted the folks on the core team, I had to put together something that would survive a rogue core team. That was the conflict: suppose a "rogue" core team in the majority told me to do something that violated the guidelines. I could do ask they asked, and then get fired by IBM, or I could refuse, and then get fired by them because I had not respected their decision. So I spent a couple of days trying to draft the appropriate language. There would be a core team in the "usual sense" except there would be a plenipotentiary (me) from IBM with special powers that would allow intercession in case IBM's Guidelines would be violated. Thing is, I could never get the language right: it all smacked of corporate control, even though this was due to a real ethical issue. Then the penny dropped. All I had to say was that if the core team ever voted for something that would put one of us in conflict with the guidelines then we would fork the code ourselves, and continue working -- as we always had -- to make Jikes better and better. It was along about then that I realized that open source was no passing fancy, but a real phenomenon. It was also in those early Jikes days that a chap named Dave Whitinger provided valuable education and guidance about open source. I have never forgotten that, which is why I'm writing this at lxer.com, and not in the blogo-sphere or at some other site. By the way, as someone who has spent most of the last four decades writing code, open source is unique. The available open source artifact is worth at least two billion dollars: if you started from scratch, hired the best people available and paid them the going rate, how much would it cost to create the gcc tool chain, the Linux kernel, KDE, Apache, Tex, and all the rest? Now all that code is available for free to anyone who is interested and who is willing to accept the license terms. Which means that every person who has helped put all this together has made a valuable contribution. What other profession can make the same claim -- that it has donated the efforts of its best members to make a freely-available artifact of great value? Another lesson I learned around this time is that open source is much more about ethics that most folks realize. It's not just a bunch of high-school kids punching out the code. To do it right means you must respect the intellectual rights of others, honor their licenses, give credit where credit is due, and try do this every day you engage with the community, and still write code worthty of its attention. It's indeed difficult, but it is so much fun -- programming at the highest level. dave |
| PaulFerris May 25, 2005 3:32 AM EDT |
Another lesson I learned around this time is that open source is much more about ethics that most folks realize. A lesson some seem to miss; and a comment that is extremely profound and fundamental. How very true to the core. |
Posting in this forum is limited to members of the group: [ForumMods, SITEADMINS, MEMBERS.]
Becoming a member of LXer is easy and free. Join Us!