In article <PMR.97Mar21093432 at unst.sanger.ac.uk>, pmr at sanger.ac.uk (Peter Rice) writes:
> In article <5gtf2c$ch3 at news.u-strasbg.fr> pingouin at crystal.u-strasbg.fr (Francois JEANMOUGIN) writes:
>> As we understand it (like Fred says), you will not be able
>>to distribute modified SDK source code, but you will be able to
>>distribute home-made programms that uses GCG libraries.
>> You will need to confirm this with GCG before distributing anything.
>> My understanding of the SDK (having only seen an extended version) is:
>> 2.B Use of the SDK API
>> i. lets you create your own software with calls to the GCG libraries.
>> ii. lets you run this (original) software.
>> iii. (not standard) lets you distribute this software. My understanding
> is that 2.B.iii is not in the standard SDK licence.
> ... So it appears that even code with just API calls cannot be
> distributed without permission from GCG. Though I would expect that
> they would give permission in practice ...
I'm not by any stretch of the imagination a copyright expert, but from
my limited (and therefore perhaps dangerous :-)) knowledge, I'd guess
it'd be extremely difficult for GCG to enforce such a restriction. If
I understand correctly, their copyright interest resides in their code
itself, not in the ideas which it embodies. (They might try to patent
their algorithms, UI, or API, but I haven't seen any statements to
that effect. I doubt they could patent the algorithms -- they're mostly
reimplementations of others' work -- but SeqLab is arguably original and
not identical to GDE, and might therefore be patentable.) Perhaps it
could be supported on the basis of non-disclosure of trade secrets.
Whatever its legal standing, I'd argue that such a restriction would
constitute yet another poor business decision on this path. As things
stand, source code distribution of original work is really only useful
to sites with SDKs, since the header files are part of the SDK.
Binary distributions could suffice for most sites with a basic GCG
license, given a distribution point with access to all the necessary
architectures and software versions, but the risk to GCG of trade secrets
or code being lost to competitors is much reduced in this case, so the
justification of restrictions on distribution is that much weaker.
Charles Bailey bailey at genetics.upenn.edu