[Pellet-users] owlSite and imports

[Schmitz, Jeffrey A]

 "I think by far the best suggestion if you want to use imports, and you
can't put the file up on the web, is what you said...just put it up on
localhost using apache and reference that url in your imports
statement."

If I understand, I think I've tried this, but can't seem to get even my
top level ontology loaded this way (much less its imports).  Doesn't the
IP address need to be accesible from your server?  Or is owlSight
running completely from within the browser?  If that's the case I'm
wondering why I'm getting a load error?  I can view my ontology's
rdf/xml directly in my browser by entering the same IP, so it should be
accessible at least locally.

Thanks,
Jeff 

-----Original Message-----
From: Michael Grove [mailto:mike at clarkparsia.com] 
Sent: Thursday, September 06, 2007 12:53 PM
To: Bijan Parsia
Cc: Schmitz, Jeffrey A; pellet-users at lists.owldl.com
Subject: Re: [Pellet-users] owlSite and imports


On Sep 6, 2007, at 1:41 PM, Bijan Parsia wrote:

> On Sep 6, 2007, at 5:44 PM, Schmitz, Jeffrey A wrote:
>
>> Hello,
>>    I was wondering how owlSite handles imported ontologies?
>
> It handles them in the "normal" way, i.e., how the OWL API handles it 
> when you don't do anything funky with it (i.e., programmatically set 
> up a "physical" URI distinct from the "logical" one).
>
>> Do the
>> imported ontologies need to reside at their URI location on the 
>> internet for owlsite to be able to pick them up and reason with them?
>
> Yes, at least at the moment. Oh, wait, they need to reside at the URI 
> which is the object of an imports statement. Their logical uri could 
> be different. And I imagine file:// uris would fail.
>
>>   Is there
>> another way in case that's not an option?
>
> If you can programmatically merge the ontologies, that would do the 
> trick. I presume you mean that the ontologies can't be made public at 
> all? I don't know if a browser can javascriptly punk through the 
> filesystem so a catalog work around might be tricky. You could always 
> run a webserver off your local machine and serve the files that way.
>
> So there are two challenges if you are unwilling to change the object 
> of your owl:imports statement for OWLSight:
> 	1) how to know where to get the ontologies (since the uri in the

> statement may not tell you where the ont is)
> 	2) how OWLSight would handle things like loading from the file
system 
> from within a browser

No, OWLSight won't play nicely with file:/// url's in owl:imports
statements.  The ontology is loaded on the server, so presumably it
would look on the server for those files and obviously not find them.
For us to search the users system, we'd have to somehow pre- parse the
ontology in the browser, chase file links and send them to the server
with the original url, and that's assuming we have a reasonable
interface to the file system via the browser, which i dont think we
have.  I think by far the best suggestion if you want to use imports,
and you can't put the file up on the web, is what you said...just put it
up on localhost using apache and reference that url in your imports
statement.

Cheers,

Mike

>
> The first is being tackled by an OWLED task force:
> 	<http://code.google.com/p/owl1-1/wiki/ImportsManagement>
>
> I'm not so clear about the latter. Mike?
>
> (The easiest work around at the moment is to change your ontology to 
> point to a localhost http server.)
>
> Hope this helps.
>
> Cheers,
> Bijan.
>
>
> _______________________________________________
> Pellet-users mailing list
> Pellet-users at lists.owldl.com
> http://lists.owldl.com/mailman/listinfo/pellet-users
> _______________________________________________
>
> Sponsored by Clark & Parsia, LLC http://clarkparsia.com/