[Biojava-l] Re: This tangled web on which I'm laid intwined

[Thomas Down]

On Sat, Feb 02, 2002 at 06:26:14AM +1000, Tony Obermeit wrote:
> I'm interested in SequenceDB and a port to MS-Access and Oracle, I've got 
> years of Oracle experience and am happy to review what would be necessary 
> to get it working in either of those environments.  I was not able to 
> download the draft document at 
> http://www.technophage.com/open-bio-database.pdf  The document was 
> blank. 

Could you try it once more, just in case there was a transient
server problem?  I've just re-downloaded it and didn't have
any trouble.

(If you're still having trouble, I can mail you a copy or
something).

> At the moment I'm new to BioJava and have only just joined the 
> list.  Do I need to join any other area of BioJava to access the cvs 
> SequenceDB code?

No, the BioSQL SequenceDB implementation is part of the main
CVS module: biojava-live.  If you're going to be checking much
code into the CVS repositiory, you'll want to contact either
myself or David to get an account created.  But if you
want to look at the code and experiment with it, you can just
follow the instructions at:

   http://cvs.biojava.org/

The BioSQL code can be found in the package:

  org.biojava.bio.seq.db.biosql

The complication with BioSQL is that the current schemas don't
yet have a proper home on the web.  As a /very/ temporary
measure, I've placed them in the BioJava download area:

   http://www.biojava.org/download/

But please don't expect them to stay there for too long.  Note
that the mysql version is currently `authoratative'.  PostgreSQL
was generated by an automatic script.


I'd be very interested in seeing ports to other databases.  The
job ought to be pretty simple:

  - Edit the MySQL schema so that it loads into your database.
  
  - Do a quick check through the SQL in the code to make sure
    it doesn't do anything really stupid.  If it does, please
    let me know -- I'd like to work out a set of queries which
    will work in as many popular databases as possible.

  - Implement the DBHelper interface for your database.  This
    is the database-abstraction layer I'm using for the BioSQL
    code.  Right now, it just has one method, which is used to
    discover auto-generated IDs after an INSERT query.  I don't
    have any Oracle experience myself, but I showed the interface
    to someone who does, and he said it would be no trouble
    to implement.


Let me know how this goes!

    Thomas.