Login | Register
My pages Projects Community openCollabNet

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Catacomb] [PROPOSAL] Catacomb joining Apache



Gianugo Rabellino wrote:
OK, I had some time to think about how to proceed, with mixed
results. Bottom line first: I think that there are quite a few
things that need to be sorted out before submitting a proposal to
Apache. Catacomb, while being good software, has severe lacks from a
number of perspectives, and before having a chance to enter Apache I
think that such problems should be solved here.

Are there other issues besides licensing/copyright related ones?

The first and foremost is licensing, since unfortunately, it's not
as easy as "OK, let's use ASL". Catacomb is in a very strange
situation, being currently "unlicensed", the only snippet stating
that it's Open Source stuff being a phrase on the website.

Catacomb has been distributed using the ASF 1.1 license since it was first made available publicly, and all but a few* of the Catacomb source files include it at the top of the file. The core development team initially chose to use the ASF 1.1 license over other common OS licenses due to the dependency Catacomb has with the Apache server. In addition, there is a UCSC copyright notice, which is more of a legal disclaimer meant to divorce UCSC from any legal liability.

IMO, this
problem can be solved only by the original copyright holder, which I
understand should be UCSC (but correct me if I'm wrong). Also, all
further contributions from non-UCSC people should be chased and
sorted out, with a proper code donation abiding to the chosen language.

Yes, I believe that UCSC would be the original copyright holder. There are some funny issues surrounding this, however... As far as I know, work done as part of a class project (the bulk of Catacomb development) belongs to the student(s) who did the work. On the other hand, work that is funded by UCSC, whether through stipends, grants or facilities, technically belongs to the university. Of course, the classic IANAL disclaimer applies to the above. Jim, can you confirm or deny my understanding of the above?

The issue of other peoples contributions to the Catacomb codebase is interesting. We can certainly track down all the various contributors and ask them if they are comfortable with their work being moved into the ASF incubator and ultimately the ASF itself. However, I am unclear about whether this level of effort is actually necessary. Given that the source code is distributed with the ASF 1.1 license, it is reasonable to assume that patches were submitted with full knowledge and acceptance of that license. Again, IANAL...

[...] licensing issues are probably the most important no-no in ASF-land.

Yeah, I've seen several lengthy license related discussions on a couple ASF projects. Most recently the JBoss / Geronimo debacle, but I digress... Let it suffice to say that I think most of us recognize the importance of licensing related issues and would agree that they should be resolved.

Now, can someone (Jim?) from UCSC - assuming that this is where
Catacomb came from - find out how hard would be having an "official"
statement from the organization about open sourcing this software
with the ASL 2.0 license?

Yes, please? Or, if you're too busy to spend cycles on this, I'd be willing to once I know who to contact. :-)

Sorry for the length of the above, these are important issues that must be fully considered. I'm also somewhat sleep deprived right now, as my pager went off at around 3am after only about 2 hours or so of sleep...


Regards,
Elias

* The following files are currently missing license information: dbms_mysql.h, dbms_mysql.c, table.sql, init_lock.sql, data.sql, configure.in, and config7.m4.in. Additionally, lock.c is missing the UCSC disclaimer.