Hi, Cees, JDBM certainly looks interesting!
One clarification about process is the project would still enter incubation, but sponsored by Apache DB. On graduation, it would then join Apache DB as a subproject -- or possibly even as its own top level project. (For example, Apache Derby and JDO were sponsored by Apache DB, and graduated as DB subprojects. Apache Cayenne was sponsored by Apache DB, but graduated as a TLP. ) Exact destination gets determined at graduation time -- but talk of graduation is leaping way too far ahead for this discussion. :-)
The Incubator proposal template will help guide discussion on this list: http://incubator.apache.org/guides/proposal.html#proposal-template You can add a JDBM proposal to the Incubator wiki: http://wiki.apache.org/incubator/ I'd especially like to hear more about your thoughts regarding this comment:
There have been some concerns about the continuity of the package, although the original contributors are still active.
Apache is about community -- and no project graduates from the Incubator without demonstrating that it has a community diverse and strong enough to manage that code base moving forward. You stated a commitment to keep JDBM alive -- how about the other contributors? --I'm keying in on this risk mentioned in the template:
http://incubator.apache.org/guides/proposal.html#template-orphaned-products
I sincerely hope the Apache DB project will consider adoption of the code base.
Again, Apache is about communities. We just need to make sure that everyone understands that.
Since I pulled out a couple comments, possibly without adequate context, I'm including your entire original post down below.
I look forward to hearing more about JDBM. regards, -jean Cees de Groot wrote:
(Currently: http://jdbm.sourceforge.net/) JDBM is an 8-year old open source library currently under a custom (original-BSD-style) license created by me in 2000 as the result of some chat with folks from Intalio and extended by a number of persons since then, most notably Alex Boisvert. It mainly aims to give similar functionality as GDBM but it adds transactions and btrees/htrees. As its goals were (and are) simplicity and robustness, after an initial flurry of development the package has been mostly stable and is in use/has been used by a number of high profile projects (http://jdbm.sourceforge.net/JDBM-Powered.html). There have been some concerns about the continuity of the package, although the original contributors are still active. ApacheDS already includes a (temprary) fork in the code base, and the main question the mailing list receives is along the line "does it still work?" (answer: yes, it does - that fact that the stable release is 3 years old means that we haven't received any serious bug reports since then, I figure ;-)). While we were discussing the future of JDBM on SourceForge (people want to move to Codehaus because of the better tools), Emmanuel Lecharny (ApacheDS) suggested that we'd consider Apache DB as a new home, and I think it is an extremely good idea for obvious reasons (continuity, exposure, peace of mind for Apache DS and other users). JDBM still fills a good niche, IMO, and would be a worthwhile addition to the Apache DB project. At the moment, I'm busy modernizing JDBM: - Move from Ant to Maven (mostly done); - Move from Junit3 to Junit4; - Move to JDK1.5+, most notably by adding generics. - Move from jdbm.* to a package name that is in accordance with the rules of ibiblio et al. for release. - Re-license under ASL/MIT/BSD ("something in that particular corner of the open source licensing jungle"). The first three points are underway, the package rename depends partially on its new home, and re-licensing requires some discussion and a vote but the contributors so far are quite relaxed on licensing (how refreshing!). Personally, as OC, I am very much committed to keeping JDBM alive. I have a bit over 2 decades of experience with working in Free Software/Open Source Software environments (starting with an fast ANSI.SYS replacement, distributed over Fidonet, in the '80s :)), among others I contributed to early Linux kernels, helped with the port of Java to Linux as a member of the Blackdown team, ran the Linuxdoc/SGMLtools projects and I have been active in various Smalltak-related communities, most notably the Squeak community (squeak.org). So I hope I qualify the "open source experience" requirement. As to the other main contributors, I leave it to them to chime in. I sincerely hope the Apache DB project will consider adoption of the code base. Regards, Cees
