In reply to Sean Eddy (btw: Hi Sean. I'm your host at GPC IV in February and
thank you for agreeing to speak [I had to get a plug in ... see
http://www.dl.ac.uk/CCP/CCP11/conferences/gpc4.html ])
My own experience in the UK (I run one of the national services) is that its
a mixed bag. Whether or not the software is free or easy to get depends to
a large extent where you are based and how literally people interpret the
`rules'.
Some universities insist that any software or data written (even at home)
whilst in their employment belongs to them. An extension of this at some
universities is a requirement to investigate possible commercialisation.
Other universities are more relaxed.
If you are funded by a research council you also have to read the small print.
Similar restrictions frequently apply. Muddy waters can occur over IPR if
your salary is paid by the university but the university gets money for you
from a research council!
There is good news here. Many people get around this by just ignoring the
restrictions either through not reading them or, having read them, realising
the they're at the end of a grant so cannot be sacked! Some research councils
(e.g. BBSRC ... I think it stands for Biotechnology and Biological Sciences
Research Council although we all lovingly refer to it as the Bioinformatics
Beer and Spirits Research Council ... I suppose I ought to know as they pay
my wage) make it a requirement that any sw/data produced from their grants
should be made freely available. Some pharmaceutical companies
also realise that it is in their interest to make software freely available,
very refreshing indeed.
If you have a permanent position, which I'm fortunate to have, and wish to
avoid waves then my own experience is that, before writing any sw/data, its
best to clear this with all parties. You don't give an ultimatum that you wont
write the software unless its GPL or GLPL but impress upon people the prestige
your site will get by providing the data for free. I speak as a confirmed GPL/
GLPL addict who's source code and data is available on anonymous ftp servers.
I'll end on a related note which has not really been addressed in this
discussion. It hinges on the definition of "making the data available". I
would regard this (in the spirit of science) as making everything available.
This should include full source code. There seem to be a plethora of plain
and licenced binaries mentioned in publications recently. It is easy to get
them but obviously, without source, not possible to either work out whether
they're an unbugged reflection of the paper or to easily further them. I
have similar worries about people "hiding" behind JAVA.
Rgds
Alan Bleasby
SEQNET (& BIONET manager Europe, Africa and Asia ... I mention this just in
case :-)
Daresbury Laboratory
Daresbury
Warrington WA4 4AD
UK
SEQNET is the UK national EMBnet node
