How Lutris betrayed the Open Source Community

Contents:

[ Updates: this site has recently featured on Slashdot, NewsForge and Java Developer's Journal. I've also created a new site www.simpledb.org devoted to two things - first the distributing InstantDB 3.26, the last release of InstantDB to which users have a clear undisputed right to use and distribute free of charge, and second the creation of SimpleDB an open source database derived from InstantDB 3.26. ]

Recently Lutris announced (see this NewsForge article) that future development of their product Enhydra Enterprise, which had up until then existed both as a commercial product and as an open source version, was going to be close source. They successfully managed to divert any resulting outrage towards SUN claiming that issues such as SUN's licensing conditions forced their hand - despite the obvious fact that there are a number of other J2EE products such as JBoss and Tomcat (which is heavily backed by SUN) that don't appear to have found a closed source future the only option (Lutris's claim that it was unable to successful negotiate a license for J2EE from Sun naturally caused concern among JBoss users but JBoss has responded reassuring users that "Lutris' decision to close its source seems clearly driven by its own business considerations and not by Sun."). [It's interesting to note that the NewsForge piece looks like it was written by Lutris being little more than a verbatim quote from Keith Bigelow, Lutris Vice President of Marketing, no one bothered to check up on anything he said - like they say "don't believe everything you read in the papers."]

In June of this year Lutris test ran this strategy, i.e. co-opting the help of the open source community in developing a product then announcing its closed source future. The product in question was the very popular embedded Java database called InstantDB and nobody kicked up much of a fuss when it became clear it didn't have an open source future. Emboldened I imagine by this trial run they've now close sourced the future development of Enhydra Enterprise and come up with a better cover story (that it's all really SUN's fault), last time they didn't even bother with any explanation at all - it was just presented as a fait accompli, one day the instantdb.enhydra.org Web site was there and next it had gone, visitors were redirected to a new site instantdb.lutris.com and user's were invited to purchase "low cost" deployment licenses (even the Java package names had been changed to reflect this shift from enhydra.org to lutris.com). There were no links to previous releases - it was now purely a closed source payware product.

At the time I assumed Lutris were just re-organizing and that an open source version would surface soon, I sent a long and detailed letter to the InstantDB mailing list, maintained by Lutris, and to various Lutris personnel (including Yancy Lind the President and CEO) expressing my concerns. Only one Lutris employ, Paul A. Morgan, the Chief Technology Officer, responded and then only to confirm my worst fears. Clearly Lutris hadn't called in their PR team to brief them on how to pitch this to the public (they've obviously learnt their lesson and hence the SUN story this time round) and the one response I received bordered on comical. Paul seemed surprised at my response to Lutris's actions and even seemed to suggest that it was somewhat ungrateful to complain - after all hadn't the user community been able to use the releases so far free of charge (these releases are no longer available and even if you have them Lutris has made clear you must purchase a deployment license for them).

InstantDB is strange in that it was often referred to as an open source product by many (including Lutris employees), but no one outside Lutris ever actually got to see the source. Lutris acquired InstantDB in April 2000, the press release of the time stated "Sponsor of Enhydra Open Source Java/XML Application Server Plans to Deliver InstantDB as the First Open Source, 100 percent JavaTM RDBMS" and that "Lutris has also begun preparing InstantDB for its introduction to the Open Source community, scheduled for early summer" (that's summer 2000). InstantDB had been initially developed by Peter Hearty and he joined Lutris to lead its continued development there. Peter Hearty has since left Lutris but I've talked to him since by e-mail and he told me "I joined Lutris solely because they promised me that they intended to make InstantDB Open Source." The user community was told that the source code was being worked on in house to ready it for its initial open source release.

Despite continued references over time on the part of Lutris employees to this imminent event such a release hadn't appeared a year after its acquisition by Lutris. This was often a topic on the InstantDB mailing list but people like Keith Bigelow (who featured so prominently in the NewsForge article announcing Enhydra Enterprise's closed source future) continuously reassured the ever growing user community that all was well, that Lutris had made a firm commitment and an open source release would definitely appear soon (see highlighted postings from Lutris employees in my tree index discussed below). These reassurances continued right up until the disappearance of instantdb.enhydra.org. In true Soviet style history has been rewritten and what was once a firm commitment to an imminent open source release has now become if mentioned at all a vague promise of an open source release at some arbitrary point in the future. With the old InstantDB site gone no one has anything to point to in order to refute Lutris's reworking of the past.

Luckily for us it's rather hard to completely eliminate things on the Web - after a lot of work pulling things out of one of the Web's largest caches I've managed to almost entirely reconstruct the InstantDB site as it existed just before being removed at the start of June. The result is enlightening - the number of references to open source and in particular InstantDB's open source future are amazing. Trawling through the site you come across many interesting things, such as the fact that Keith Bigelow even presented a session at the O'Reilly Conference on Enterprise Java in March entitled "Dynamic Wireless Application Development with Open Source technologies: Java, XML, Enhydra, and InstantDB." You can find the session outline at the O'Reilly site along with the presentation itself - I notice that the above title is used both in the O'Reilly session outline and on the events page at enhydra.org but in the archived presentation one can see that when it actually came to conference time Keith used the somewhat different title "Dynamic Wireless Application Development with Open Source and Java". Some of it is quite sad, just take a look at a Short History, one of the main links that appears on the left hand side of every page, clearly many people at Lutris really did believe in Open Source and worked hard in the belief that what they were doing would ultimately be returned to the Open Source community.

I've provided as much of the site as I can, most of it's there including the mailing list archives and I've a produced a nice tree index of the site showing every page that makes reference to open source. The only changes I made to the pages I retrieved from the Web cache were to relativize all the links so the pages link to each other correctly in their new location and more interestingly I've highlighted the words open source wherever they appear so they can be easily picked out on any page. The lutris.com site itself still has some interesting pages in relation to InstantDB, the original press release mentioned above announcing their acquisition of InstantDB makes particularly interesting reading in light of subsequent happenings (here's a mirror of the press release should Lutris decide to remove it).

A Snapshot of instantdb.enhydra.org as it once was

Compare and contrast (click on the images)

As it was:

And as it now is:

A tree index of all the pages on the old site mentioning open source.

Notes:

I would also like to have a copy of all the postings to the InstantDB mailing list since I took this snapshot of the site. Lutris used to maintain an archive of the list but it amused me to see when I looked recently that this has disappeared. Take a look at the mailing list page of Abalone (or any of Lutris's other products), note the View button that takes you to an archive of the list, the mailing list page for InstantDB used to have just such a button as well (as you can see in my snapshot of the old site) but it seems to have disappeared - I suspect so that you can't see all the negative e-mails that appeared once it became clear Lutris was reneging on its promise to open source InstantDB. I don't have any postings past June 8th, when I was automagically unsubscribed from the mailing list as a thank-you for my comments on the matter - despite the mailing list charter stating that while the lists are moderated (to remove spam etc.) that "Moderation ... does not, in any case, refer to blocking mails that are negative or otherwise unsavory.").

Keith Bigelow, Lutris VP of Marketing, "reassures" on InstantDB and Open Source

Whenever anyone questioned Lutris's commitment to open sourcing InstantDB Keith Bigelow, Lutris Vice President of Marketing, would be on hand to reassure, again and again and ...
Re: Are opensource plans dropped?Keith Bigelow<keith.bigelow@lutris.com>
Re: InstantDB's Licensing and Open Source StatusKeith Bigelow<keith.bigelow@lutris.com>
Re: open sourceKeith Bigelow<keith.bigelow@lutris.com>
Re: Some Issues about InstantDBKeith Bigelow<keith.bigelow@lutris.com>

[ Whatever happened to those "legal commitments with IBM, Nortel and other global companies" mentioned in the last message? Apparently they should have led us to have "no concern regarding either our commitment to the product or to the open source process for the product." ]

Of course as Paul Morgan, Lutris Chief Technology Officer, tells us the views expressed by Lutris employees on the InstantDB mailing list "are often mis-informed" (it's unfortunate then that Lutris is still allowing Keith to express his opinions as if he was talking on behalf of the company).

Mail Exchanges

• George C. Hawkins <george.hawkins@pobox.com> Future status of instantDB
    I express my initial concerns - at this point I'm not too worried, I assume things are being moved about but that InstantDB will resurface as an open source project. It's quite long but well argued and well structured IMHO and makes all the points I thought needed to be made - just in case - pointing out specific postings etc. where Lutris committed itself to open source.
• Peter Hearty <pete@hearty.plus.com> Re: Future status of instantDB
    Peter reports that he no longer works for Lutris, states that "I joined Lutris solely because they promised me that they intended to make InstantDB Open Source", and that "No one will be more disappointed than me at such a decision [not to open source InstantDB]." He then asks Keith Bigelow (who's CCed) to clarify the situation (which Keith does not).
• Marci L Lehman <marci.l.lehman@intel.com> thank you
    At least one or two people thanked me for a "straightforward and well-documented commentary" - much appreciated.
• Jonathan Marcotte <marcottej@grics.qc.ca> RE: InstantDB
    Jonathan Marcotte having developed his open source application against InstantDB in the belief that InstantDB would be free to use reports that Lutris have informed him that he'll have to pay $695 per deployment license.
• Paul A Morgan <paul.morgan@lutris.com> Re: Future status of instantDB
    Paul responds to my original e-mail. He really should have run this past the PR department first. One of the lines I love is "I would suggest that we have acted very altruistically towards the community!" - i.e. wasn't it nice of us to let you develop (or should that be "lure you into developing") against InstantDB free of charge for almost a year before informing you that you'd have to pay for deployment licenses! And in response to my references to the many postings on the InstantDB mailing list from Lutris employees reassuring the user community about InstantDB's open source future Paul says "Such views are often mis-informed and this is unfortunate." (one wonders if Keith and others were been severely reprimanded for propagating such mis-informed views, some how I doubt it).
• George C. Hawkins <george.hawkins@pobox.com> Re: Future status of instantDB
    I respond to Paul's e-mail suggesting that his response isn't very satisfactory at all. This is rather less well argued than the original as I realize the user community really has been duped and react angrily. Paul doesn't choose to respond.
• Paul A Morgan <paul.morgan@lutris.com> Followup on the instantDB move
    Paul in an open letter to the user community informs us all that we'd obviously all been seriously confused (as clearly had Keith and other Lutris employees if their postings to the InstantDB mailing list are anything to go by) about the previous state of InstantDB ("there is certainly some confusion on the previous state of InstantDB"). And comes up with a real cracker informing us that buying licenses for the closed source InstantDB will be good for open source - "The proceeds from such licenses will allow us to further development of this technology and the other Enhydra.org open source efforts for the benefit of all." What can you do but laugh?
• Matthew Ferris <hill1823@yahoo.com> Re: Followup on the instantDB move
    Matthew is the first of I suspect many to react less than positively to Paul's open letter, asking him "to be more specific than 'Various unforeseen factors have forced Lutris to reconsider the timing of such a move [i.e. to open source]'."
• Paul A Morgan <paul.morgan@lutris.com> Re: Followup on the instantDB move
    Paul shows he can't think standing on his feet and resorts to a "I could tell you but then I'd have to kill you" response. At this point you can't but realize that any further interaction is probably pointless.
• Kambiz Darabi <darabi@m-creations.com> Re: Followup on the instantDB move
    Kambiz Darabi, CTO of m-creations, echoes the thoughts of many saying that he'll be willing to contribute to InstantDB again once it is open sourced but "Until that time, some of us who came to InstantDB because of the 'dangling carrot' will choose another solution for their Java database needs."
• George C. Hawkins <george.hawkins@pobox.com> Re: Followup on the instantDB move
    My response to Paul's open letter - at this point it seems clear to me that we're just dealing with con-men (a view that Lutris's recent actions in regard to Enterprise Enhydra only confirm) - the casual use of bad language is far more common in Ireland and Britain than America - so apologies if you take offence at the language used - I was very cross! As a result of this posting I'm automagically unsubscribed from the list!
• George C. Hawkins <george.hawkins@pobox.com> InstantDB, its move to a close source future and your tool?
    Many tools have been built by third parties that contribute to the value of InstantDB. This is my open letter to developers that have developed open source tools that are advertised on Lutris's site and would increase the perceived value of InstantDB to anyone considering purchasing an InstantDB license in the future. I suggest that these developers request that InstantDB no longer advertise their product as working with InstantDB considering Lutris's recent behavior towards the open source community.
• Mark Watson <markw@markwatson.com> Re: InstantDB, its move to a close source future and your tool?
    Mark Watson creator of NLBean thanks me for the heads-up and (like many) informs me that he'll "probably switch over to HypersonicSQL if it checks out OK. I want people to be able to use my NLBean program for free."
• Paul Morgan <paul.morgan@lutris.com> Re: InstantDB, its move to a close source future and your tool?
    Paul in the understatement of the year states "You are clearly bitter." And incorrectly informs me "that all of these tools were built against InstantDB prior to Lutris' ownership when the database was under the same license as it is now." My understanding is that while InstantDB was payware for commercial use before it was acquired by Lutris it was free for non-commercial use.
• George C. Hawkins <george.hawkins@pobox.com> Re: InstantDB, its move to a close source future and your tool?
    My final e-mail to Paul - a not so witty rejoinder.

I would like to point out that while much of the software I use is open source I am not an open source fanatic, I use MS Win2K on my desktop system but my server system runs Linux. If I submit a bug report to MS, even though I've contributed to the company and helped them improve the product, I don't expect to get a copy of the source in return - MS made no such deal with me. This is however part of the deal I and the rest of the open source community had with Lutris or thought we had. Lutris said "help us to develop a top class product by incorporating it into or using it with products you're working on, whether they be open source or otherwise, see what issues arise and through your bug reports and contributions on the mailing lists together we'll develop a top class Java database, and an open source release will be made soon."

In retrospect you may think users were naïve - what good reason could there have been for not making the CVS repository public from the start - but the InstantDB site and Lutris employees provided much reassurance - the source was just being readied in house and when people pointed out every so often that this seemed to be taking a long time, Keith Bigelow or someone similar popped up with further reassurance.

I guess we may see more of this in future, I hope not though, as fair weather open source enthusiasts who jumped on the band wagon when the money looked good and all talk was of a new economy see their business dreams come crashing in less exuberant times. Open source was around long before anyone had heard of the new economy and despite the odd hiccup like this I'm sure it'll keep going from strength to strength. However companies like Lutris who take advantage of the open source community must be named and shamed and shouldn't be allowed to damage people's confidence in and support for open source development.

If you have any comments about this issue or what I've said please send them to me - George C. Hawkins <george.hawkins@pobox.com>. Your thoughts and comments would be much appreciated.

Terry Steichen's exchanges with Yancy Lind

Terry Steichen wrote to me recently and I took a look at his postings to the enhydra.org mailing list. Terry is far more eloquent than I and I recommend you take a look at his postings on the list, they make very good reading. Particularly the following exchange. Terry Steichen takes the role of tough political journalist grilling Yancy Lind (President and CEO of Lutris) playing slippery politico for whom reality warps to meet his needs. Unfortunately unlike a TV interview where the interviewee can't escape and a tenacious interviewer can seriously embarrass the interviewee by repeatedly asking the same blunt question that the interviewee clearly doesn't wish to answer Yancy is able here to bail with the classic line "All of these questions have been previously asked and responded to." - responded to maybe, answered no.

Enhydra: Lutris' J2EE Change - What are the Core Issues?Terry Steichen<terry@net-frame.com>
Re: Enhydra: Lutris' J2EE Change - What are the Core Issues?Yancy Lind<yancy.lind@lutris.com>
Re: Enhydra: Lutris' J2EE Change - What are the Core Issues?Terry Steichen<terry@net-frame.com>
Re: Enhydra: Lutris' J2EE Change - What are the Core Issues?Yancy Lind<yancy.lind@lutris.com>
Re: Enhydra: Lutris' J2EE Change - What are the Core Issues?Terry Steichen<terry@net-frame.com>
Re: Enhydra: Lutris' J2EE Change - What are the Core Issues?Kenny A. Chaffin<kenny@kacweb.com>
Re: Enhydra: Lutris' J2EE Change - What are the Core Issues?Bill Karwin<bill.karwin@lutris.com>
Re: Enhydra: Lutris' J2EE Change - What are the Core Issues?Terry Steichen<terry@net-frame.com>
Re: Enhydra: Lutris' J2EE Change - What are the Core Issues?Yancy Lind<yancy.lind@lutris.com>
Re: Enhydra: Lutris' J2EE Change - What are the Core Issues?Terry Steichen<terry@net-frame.com>
Re: Enhydra: Lutris' J2EE Change - What are the Core Issues?Alexander Schatten<alasan@gmx.net>

Unfortunately the enhydra.org mailing list is not the ideal place to discuss such matters, inconvenient postings, such as this one from Terry Steichen, tend to get "lost", i.e. you won't find them in the Lutris maintained list archive - the lists and their archives are clearly censored despite any Lutris claims to the contrary. As Terry put it in this e-mail to me "Yancy and others tend to give self-serving responses and, when the questions are particularly discomforting, they either stonewall or worse, 'lose' the relevant (read: critical) messages."

N.B. I must point out that Terry Steichen has a far more favorable attitude towards Lutris than I do, while I see them as little short of con-men Terry believes that positive ongoing engagement with Lutris is worthwhile, this can be seen in his postings to me such as the one above (that the quote comes from) and this one and in his other postings to the enhydra.org mailing list.

How you can help

While suggestions for changes are welcome actually making the changes and sending me the result is even more appreciated! You can contact me, George C. Hawkins, at <george.hawkins@pobox.com>.

This site is dedicated to all those in the open source community who gave their time and effort during the development of InstantDB and were then cheated out of the open source product they thought they'd been part of - George C. Hawkins <george.hawkins@pobox.com> September 2001

since October 10th, 2001