[Biojava-l] File formats

[Rhett Sutphin]

Hi Rich, Matthew,

I am also not a lawyer.  That in mind, here's my understanding of the topic:

First, the org.biojava.bio.program.abi package is based on a paper by Clark Tibbetts (available online: http://www-2.cs.cmu.edu/afs/cs/project/genome/WWW/Papers/clark.html ).  That paper was published in August 1995 and is a fairly thorough technical description of the ABI 377, including its means of operation, communication protocols, and (of course) data files.  I am unaware of any legal action taken against him or Vanderbilt (his apparent employer at the time).  And, of course, the paper remains available.

Second, the Staden io_lib library can read ABI-formatted chromatograms.  I am unaware of any legal action against its makers and it is currently available (and has been for a while).

Third (and this is, again, just my understanding of current US law), reverse engineering for interoperability is legal.  The only area where this is not true is if the material is (a) copyrighted and (b) protected by an "access control."  If these conditions are met, then the material falls under that most unpleasant of IP laws, the DMCA.  However, (a) whoever wants to read the ABI files with your software will probably own the copyright to them (if they are even copyrightable -- they might just be lists of facts and hence uncopyrightable in the US); and (b) I don't think a proprietary file format rises to the level of an "access control."  An ABI file isn't encrypted -- you just have to know what offsets from which to read the bytes.

Rhett

Matthew Pocock wrote:
> Hi Rich,
> 
> We should check this out. This is one of the bizar things about digital 
> IP right now - the data in the abi file is obviosly yours, but 
> potentially you are not alowed to access it in non-blessed ways because 
> the encoding is proprietary. I have a feeling that we would have been in 
> trouble if our code was based upon their serializer/deserializer code 
> (which it is not) due to copyright issues. SW pattents don't work in the 
> EU/UK (yet). Further than that I don't know. Oh, and IANAL.
> 
> Matthew
> 
> (goes to speak with someone who may know more)
> 
> Rich Heath wrote:
> 
>> Hi,
>> I am a software developer based in the UK that has
>> been asked about producing a piece of software that
>> outputs data from the files in ABI sequencers in a
>> more human readable format. I hope the
>> org.biojava.bio.program.abi package will let me do
>> this, but I have some concerns about the legal
>> implications of using and contributing to this
>> package.
>> Does anyone know what the legal position is with
>> regards reverse engineering the Applied Biosystems
>> file format (and any other file formats come to that
>> matter)? I would imagine this file format is the
>> property of Applied Biosystems and they would not like
>> me producing applications that read from it unless I
>> provide them with a sizable licence fee (although I
>> guess I am not reverse engineering it if I just use
>> the above package, just if I contribute to it?).
>> Many thanks in advance for your help,
>> Rich
>>