Re: "First" PoBoL BBF RFC 31 "Provisional BioBrick Language (PoBoL)"
Michal Galdzicki
Thank you for your help. I incorporated many of your comments into
RFC31. Critical issues remain and need discussion as there are
important solution choices to be made. PoBoL must evolve to meet the
needs of the community. Writing the RFC has helped us identify many
of these issues, which we should now discuss and address. Below is a
list of those that i have remaining now in my notes. I probably missed
something so please add and elaborate.
(I posted all these on Google Moderator
http://moderator.appspot.com/#8/e=6e5fa so we can address each one
individually. Try it, if its not generally liked we can abandon it
and go back to the mailing list.)
CRITICAL issues to solve:
1*. BioBrickComposite should be an ordered sequence and requires more
information about the composition: (eg connecting scars) (from Raik)
2*. The DNA sequence of a BioBrick could be described independent of
the composition format (from Alec)
3. A mechanism for attributing authors to BioBricks, "Samples"
(DublinCore "dc:" v. FOAF) (from Raik,Mac,Jason, me)
4*. Clear cardinality restrictions (OWL class definition restrictions,
range restrictions, logic that would verify consistency of data)
(Jason,Mac,me)
5. Modular potential of PoBoL (eg. laboratory information management
(tubes,plates, stock cultures) v models for simulation) (Raik, Sean,
Deepak)
6. Overall idea about what is"minimal" or "core". Are all the
REQUIRED properties etc really needed?
THINGS to consider
1. Should PoBoL be described in OWL, RDFS, or should it be independent
of a specific implementation? (UML)? Right now PoBoL is almost OWL
Lite (uses Union and Disjoint but are they necessary?)
(BTW OWL2 has some important improvements, should consider.)
2. Adopting conventions for naming of classes, properties, instances
(Camel case, verb like properties)? (Jason,Raik)
2.a. Should we use unique identifiers for instances? (me)
3. Namespace issues, should we have some URL that returns pobol.owl
(RFC30 suggested it)
4. We should get some real use cases described on the site, its best
to develop solutions to well documented problems.
5. What is a BioBrickFamily anyway (no one noticed, but we didnt
really talk about it oops)
6. Turtle is nice but updating the turtle version, diagram, and
separate example file while writing the RFC was a pain! (me)
Thank you again for your help, :)
-mike