Hi Patrik,
Patrik Holmer wrote:
> Ryan, again, thank's for a quick response.
>
> It's obvious I'm missing the apparent, but this is what it looks like at my
> end of the stick:
>
> Initially following the instructions available at
> http://simile.mit.edu/semantic-bank/install.html or more specifically the
> last paragraph:
>
> ---
> To include 'starting points' and descriptive labels in your bank, or any
> other RDF on your filesystem, run your installation with the directory as
> another argument after the properties file (make sure to do this in the
> correct order, the ontologies must be loaded first):
>
> ./semantic-bank.sh run semantic-bank.properties src/rdf/ontologies
> # stop after it starts properly
Perhaps we should be more detailed and mention that 'properly' means the
Semantic Bank should be visible when browsed to, and that 'stopping' is
something along the lines of ctrl-C.
> ./semantic-bank.sh run semantic-bank.properties src/rdf/configurations
> src/rdf/facades
> ---
>
> With no errors echoed, I expected to see a default Semantic Bank, as
> depicted on the page http://simile.mit.edu/semantic-bank/index.html. What I
> got, running the ending session with the parameters "src/rdf/configurations
> src/rdf/facades", was a basic Semantic Bank. No starting point and no
> descriptive labels, with the warning "None of the predefined starting points
> contains data".
Have you added anything to your bank? The SB screenshot on the home
page is of one that's been used heavily. A new, empty bank has no tags,
no starting points (because there's no data), and not particularly
interesting typed data (classes, properties, etc.). What do you mean by
'no descriptive labels?' To me, that sounds like the ontologies haven't
been loaded, but I'm not sure what indicates to you that no descriptive
labels exist.
Is your instance of SB running in a public location? Can we see it?
> Now, with no further references provided - that I know of - on this site and
> with no further process feedback, than an execution echo or a flurry of
> event responses when pressing CTRL+C or Break in DOS or signalling
> STOP-TERMINATE-KILL in Linux to stop the first session, I'm past rasing only
> one swedish eyebrow and have resorted to trying every trick in the book -
> i.e. "...just close the console".
>
> I assume I'll have to drop this subject, load the project into Eclipse and
> reach my own findings - when time given.
I feel like you're trying to chase down a problem that doesn't exist.
Forgive me if I've seriously misunderstood, but nothing you've mentioned
seems out of the ordinary to me.
--
Ryan Lee ryanlee_at_w3.org
W3C Research Engineer +1.617.253.5327
http://simile.mit.edu/
Received on Thu Oct 20 2005 - 23:26:06 EDT