[Biojava-l] Off Topic: Monograph

Brian Gilman gilmanb@genome.wi.mit.edu
Fri, 7 Dec 2001 11:27:08 -0500 (EST)


This is very cool stuff!! I've been thinking about this a lot lately.
Specifically, exposing a database schema as XML and then using
RDF/DAML+OIL to represent the relationships amongst these schema's. I'm
thinking of something like powerbuilder which could query databases
anywhere in the world (UDDI) and shows the relationships between entities
in the database as references (RDF). And finally resolves attribute
relationships though DAML+OIL. Transacation management would be a
nightmare though...

			-B

-----------------------
Brian Gilman <gilmanb@genome.wi.mit.edu>
Sr. Software Engineer MIT/Whitehead Inst. Center for Genome Research
One Kendall Square, Bldg. 300 / Cambridge, MA 02139-1561 USA
phone +1 617  252 1069 / fax +1 617 252 1902


On Fri, 7 Dec 2001, Matthew Pocock wrote:

> Hi all,
> 
> Sory if you are not interested in half-baked home-grown query languages.
> 
> Monograph is an API for manipulating meta-data. It can represent 
> relational-db, object inheritance, ace-db style frames, DAML-OIL terms 
> and most other things using the same unified model. It is language 
> neutral (but this implementation is in Java). It is intended as a 
> sensible language for glueing different data-sources together (possibly 
> with very different query APIs) without the need to push some of the 
> query logic into client-side hand-written loops.
> 
> The exciting features possible in Monograph (but not fully implementated 
> by the current code-base):
> 
> * the ability to compare two database schemas to see if they overlap
> 
> * represent concepts like design paterns as first-class objects that are 
> generative models over code
> 
> * no distinction between meta-data and data - you can turn any 
> data-structure into the rules of a new language - indeed, single 
> data-values are just very restricted languages that only ave one 
> possible value - the data-value
> 
> * high-level optimization of execution paths that can potentialy take 
> into account the foibles of sub-class implementation details
> 
> I've just checked a new version of the core interfaces for Monograph 
> into sourceforge (project name monograph, CVS package monograph-live). 
> It is an almost complete re-write of the original code. It is very 
> un-polished, and there are loads of bits of core functinality that 
> aren't implemented yet (like validating types by schema). I will try to 
> check some documentation into the tree before Monday.
> 
> This is /very/ alpha code, but somebody out there may get the warm 
> fuzzies while reading through it.
> 
> all comments/flames/donations gratefully recieved,
> 
> Matthew
> 
> _______________________________________________
> Biojava-l mailing list  -  Biojava-l@biojava.org
> http://biojava.org/mailman/listinfo/biojava-l
>