This file is referenced in linear-ordering.xml and nofinal.xml

linear-time-ordering.clif is referenced in linear-ordering.xml

owltime2orderings.clif, a mapping ontology that we need for the example in the 
DOL paper for CICM, exists twice:

colore$ find . -name owltime2orderings.clif 
./complex/owltime/owltime_instants/mappings/owltime2orderings.clif
./complex/owltime/owltime_interval/mappings/owltime2orderings.clif

The second one has one more axiom.

colore$ diff `!!`
diff `find . -name owltime2orderings.clif`
3a4,8
>       (iff    (leq x y)
>               (or     (before x y)
>                       (= x y))))
> 
> (forall (x y)

The second one is referred to from 
complex/owltime/owltime_interval/mappings/owltime_*.dol, whereas the first one 
doesn't seem to be referred to from anywhere.

Secondly, I wonder whether "owltime2orderings" is the right name.  The pattern 
of the axioms is "iff <some situation in terms of orderings> then <some 
situation in terms of owltime>", and they way it's used in the DOL files is 
also "interpretation NAME : <ordering> to <owltime>".  OK, the "iff" is 
bidirectional, but overall the direction of _writing_ the mappings seems to be 
"ordering2owltime".

Further investigation is not easy for me, as e.g. the symbol "before" seems to 
occur both for intervals and for instants.  Maybe it would help if we used IRI 
identifiers for it, for disambiguating "interval-before" from "instant-before".

cd <repo>
grep -r  '(di(x,y) & d(y,z))' .

There is currently no single consistent module naming convention within COLORE. 
 Some modules use the file name for the module name (e.g. see actocc.clif).  
Other modules use the filename less the extension (e.g. see psl-core).  Some 
use a module name that differs in case from the filename (e.g. see rt.clif).  
Some use the Camel Case naming convention (e.g. see 
orthocomplementedlattice.clif) while others use the hyphen convention (e.g. 
near-linear.clif). 

The CL standard recommends that URIs be used within CL dialects intended for 
use on the web.  Whilst CLIF is not specifically targeting the web, it stands 
to reason that importations within a CLIF file may reference named texts and 
modules that need to somehow be resolved.  URIs are an accepted mechanism for 
doing this.  Also, the intension to eventual deploy COLORE within an "online" 
repository that publishes its content through a collection of RESTfull web 
services.  Therefore, I recommend that the following actions be considered:

1) Obtain an Internet domain name for colore (colore.ca is currently available).

2) Adopt a COLORE file naming convention that uses all small letters with 
underscores separating words (e.g. psl_core.clif).

3) All modules are named according to best practices for designing URIs.  For 
example, http://oor.colore.ca/ontologies/psl_core.  All CLIF texts importing 
the psl_core ontology would refer to this URI which is universally locatable. 

Cameron.

Unique Point Plane - imported files

1st import: weak_tripartite.clif

A summary of guidelines/issues discussed during the Lab meeting on Jan. 17th 
2011:

1)  As per Cameron's suggestion for file naming and module naming: 
- Adopt a COLORE file naming convention that uses all small letters with 
underscores separating words (e.g. psl_core.clif).

2)  In the clif files, include the full URL for module names:
- Ex.  cl-module http://stl.mie.utoronto.ca/colore/generic/time/lp-ordering.clif
       cl-imports http://stl.mie.utoronto.ca/colore/generic/time/lp-ordering.clif

3)  Moving all clif files to this googlecode repository.

4)  Moving the html files (index pages for each directory, etc.) for COLORE to 
this googlecode repository.

5)  Collecting issues regarding name-spacing for CL relations, etc.
-  Ex. For translation definitions between two modules that use the relation 
name "meets" with axiomatizations that are not equivalent, how do we 
distinguish between the two different relations if they have the same name?