Re: [Crm-sig] CRM documentation with Parrot

[Georg Hohmann]

Dear Vladimir,

> (Which bright mind decided to add a space before the number for the new classes and properties? Please remove that space!
> E 91 Co Reference Assignment, ... P 153 assigned co reference to ...
> Please also, add the dash in Co-Reference and co-reference)

I don't understand what you mean.
In the ECRM are no spaces before the number for the new classes and
properties, and there are no missing dashes either.
For instance, the class and property you mention are defined like this:
<owl:Class rdf:about="http://erlangen-crm.org/140220/E91_Co-Reference_Assignment">...</
owl:Class>
<owl:ObjectProperty
rdf:about="http://erlangen-crm.org/140220/P153_assigned_co-reference_to">...</owl:ObjectProperty>

> Then I checked what's up at http://erlangen-crm.org/documentation
> - Please remove the Ontology Browser links, these are long dead

Good idea. In opposite to what they have stated here ...
http://owl.cs.manchester.ac.uk/web-services-update/
... it seems that the service will not be online again.

> - IMHO the OwlDoc http://erlangen-crm.org/docs/ecrm/current/index.html looks bland compared to Parrot

In my opinion both html representations have their drawbacks. For
instance the large clusters of classes and properties at the beginning
of parrots html are not very well-arranged and useful. I guess that is
a matter of taste and nothing one could fix with some lines of css
:-).

Best regards

Georg

[Vladimir Alexiev]

> > Please remove that space!
> > E 91 Co Reference Assignment, ... P 153 assigned co reference to ...
> > Please also, add the dash in Co-Reference and co-reference)

> In the ECRM are no spaces before the number for the new classes and
> properties, and there are no missing dashes either.

I mean rdfs:label, sorry that I was unclear.
Cheers!

[Georg Hohmann]

Dear Vladimir, dear List,

> - Please remove the Ontology Browser links, these are long dead

As suggested the documentation of the current version is now generated
by Parrot.
http://erlangen-crm.org/documentation

> If you decide to switch to Parrot, the best is to host the HTML file locally
> - You also need to host the icons locally (ask me for a folder "report" with subfolders for css, js, images).
> - I do this in batch, and remove some cruft at the bottom, else you get breakage in the bottom toolbar:
> curl -X POST -F datafile=@ontology.rdf -F mimetypeFile=application/rdf+xml -o ontology.html http://ontorule-project.eu/parrot/parrot
> perl -n -e "print unless m{<!-- Feedback -->} .. m{<!-- standalone -->}" -i ontology.html

I simply use "wget" to get a local and fully woking copy of the
documentation: e.g.
wget -p http://ontorule-project.eu/parrot/parrot?documentUri=http%3A%2F%2Ferlangen-crm.org%2Fcurrent%2F

Best regards
Georg

[Vladimir Alexiev]

> http://ontorule-project.eu/parrot/parrot?documentUri=http%3A%2F%2Ferlangen-crm.org%2Fcurrent%2F

Great!
Suggestion: remove the space after "E" in thes labels:
E 91 Co Reference Assignment , E 92 Spacetime Volume , E 93 Spacetime Snapshot

Also, these classes are missing from the Glossary for some reason:
http://ontorule-project.eu/parrot/parrot?documentUri=http%3A%2F%2Ferlangen-crm.org%2Fcurrent%2F#glossary

What do you think about inverse properties, should the label include "i": e.g. "P87i identifies"?
I see that my script removes "i" on purpose, since it writes skos:notation="P87i":
https://github.com/erlangen-crm/ecrm/commit/763b71fd4901b3f008ac703ca29cc1d8ff6e54ab#diff-2eeab9e3e93b6694d8031a245af11213R14
But now I have doubts... :-)

[Georg Hohmann]

Dear Vladimir,

> Suggestion: remove the space after "E" in thes labels:
> E 91 Co Reference Assignment , E 92 Spacetime Volume , E 93 Spacetime Snapshot

That was a bit strange. These classes did not have any labels at all,
and Parrot somehow added a space between the "E" and the number. To
fix this, we now added rdfs:label and skos:notation to those classes
and properties where they were missing.

> What do you think about inverse properties, should the label include "i": e.g. "P87i identifies"?

No, I think it's okay the way it is.

Best regards
Georg