> This is just a brief notice of some very preliminary rdfa-related work
> at http://nomisma.org/ .

> The overall idea for nomisma is to provide stable URIs for numismatic
> entities and to describe those entities using rdf/rdfa/linked data.

> As a test, we've made the text of Inventory of Greek Coin Hoards
> available. At this point there are URIs similar to
> http://nomisma.org/id/igch0156 for all 2387 hoards described in that
> work.

> If you click-through to igch0156, you'll see links at the bottom of
> the page. "Show Markup in Page" does just that. "rdf n3 triples" links
> to the w3 rdf extractor that makes use of the embedded rdfa.

> At this stage of the game, I am working to develop an xhtml/rdfa
> based convention for marking up hoards of ancient coins. I have two
> main goals: make the information visible to the rdfa extractor, and
> also keep the markup simple enough so that I can teach it to our
> not-very-technical future collaborators.

> I'm trying to keep myself to the following patterns for the rdfa:

> <a rel="nm:*" href="relative url"></a>
> <span property="nm:*" content=""></span>

> It seems clear, however, that I won't be able to express all the
> information that I want in ways that are visible to a single-pass of
> an rdf extractor.

> For example, the line "Contents: c. 260 AR" indicates that all the
> coins are silver. I don't think there is rdfa that will generate
> triples to that effect for both the Euboean and Histiaean coins. So
> for now I group the information in hierarchical instances of '<span
> class="nm:g"></span>' where nm:g is short for group. This anticipates
> my writing a style sheet to generate rdf describing 18 instances of
> silver drachmas from Histiaea. Perhaps I'll indicate this with GRDDL.

> And I want to work on linking my graphs to external graphs.

> For example, http://nomisma.org/id/phaselis has the markup '<a
> rel="owl:sameAs"
> href="http://pleiades.stoa.org/places/639051">pleiades</a>'. Is that
> actionable in both directions between nomisma and pleiades? Not yet
> but what are the conventions I should use to make it so?

> This is all extremely preliminary so I will appreciate any feedback.
> Though I do recognize that the above leaves much to be desired as an
> introduction.

> -Sebastian

> What would you think about sharing a findspot property with the IAph
> data?

> I see an immediate problem with asserting that your Phaselis is the
> same as http://pleiades.stoa.org/places/639051: the latter URI
> identifies not a place but a description of a place. We really need to
> address this soon on the Pleiades end. We could obtain non-information
> resources for free by inviting everybody to use URIs like http://pleiades.stoa.org/places/639051#foo
> , but I'd rather not.

> Cheers,
> Sean

> --
> Sean Gillies
> Software Engineer
> Institute for the Study of the Ancient World
> New York University

> On Jan 21, 2009, at 9:53 PM, Sebastian Heath wrote:

>> This is just a brief notice of some very preliminary rdfa-related work
>> at http://nomisma.org/ .

>> The overall idea for nomisma is to provide stable URIs for numismatic
>> entities and to describe those entities using rdf/rdfa/linked data.

>> As a test, we've made the text of Inventory of Greek Coin Hoards
>> available. At this point there are URIs similar to
>> http://nomisma.org/id/igch0156 for all 2387 hoards described in that
>> work.

>> If you click-through to igch0156, you'll see links at the bottom of
>> the page. "Show Markup in Page" does just that. "rdf n3 triples" links
>> to the w3 rdf extractor that makes use of the embedded rdfa.

>> At this stage of the game, I am working to develop an xhtml/rdfa
>> based convention for marking up hoards of ancient coins. I have two
>> main goals: make the information visible to the rdfa extractor, and
>> also keep the markup simple enough so that I can teach it to our
>> not-very-technical future collaborators.

>> I'm trying to keep myself to the following patterns for the rdfa:

>> <a rel="nm:*" href="relative url"></a>
>> <span property="nm:*" content=""></span>

>> It seems clear, however, that I won't be able to express all the
>> information that I want in ways that are visible to a single-pass of
>> an rdf extractor.

>> For example, the line "Contents: c. 260 AR" indicates that all the
>> coins are silver. I don't think there is rdfa that will generate
>> triples to that effect for both the Euboean and Histiaean coins. So
>> for now I group the information in hierarchical instances of '<span
>> class="nm:g"></span>' where nm:g is short for group. This anticipates
>> my writing a style sheet to generate rdf describing 18 instances of
>> silver drachmas from Histiaea. Perhaps I'll indicate this with GRDDL.

>> And I want to work on linking my graphs to external graphs.

>> For example, http://nomisma.org/id/phaselis has the markup '<a
>> rel="owl:sameAs"
>> href="http://pleiades.stoa.org/places/639051">pleiades</a>'. Is that
>> actionable in both directions between nomisma and pleiades? Not yet
>> but what are the conventions I should use to make it so?

>> This is all extremely preliminary so I will appreciate any feedback.
>> Though I do recognize that the above leaves much to be desired as an
>> introduction.

>> -Sebastian

> Hi Sebastian,

> What would you think about sharing a findspot property with the IAph
> data?

> I see an immediate problem with asserting that your Phaselis is the
> same as http://pleiades.stoa.org/places/639051: the latter URI
> identifies not a place but a description of a place. We really need to
> address this soon on the Pleiades end. We could obtain non-information
> resources for free by inviting everybody to use URIs like http://pleiades.stoa.org/places/639051#foo
> , but I'd rather not.

> Cheers,
> Sean

> --
> Sean Gillies
> Software Engineer
> Institute for the Study of the Ancient World
> New York University

> I'm reminded here about the ORE recommendation, and the use of <uri>
> to refer to the thing, and <uri>#description to refer to the
> description.  I know this isn't without problems (not least of which
> is the tendency to resolve the non-information URI's to the
> description), but it does allow the relationship between things to be
> expressed in a way that seems more natural to me (that is, it makes
> more sense to me to say "this object is related to this place" than
> "this object is related to a place that is is also described by this
> place description").  I may also be over-interpreting Sean's concerns,
> or misinterpreting Sebastian's rdfA.  Just sayin'.

> - Ben

> On Thu, Jan 22, 2009 at 6:26 PM, Sean Gillies  
> <sean.gill...@gmail.com> wrote:

>> On Jan 21, 2009, at 9:53 PM, Sebastian Heath wrote:

>>> This is just a brief notice of some very preliminary rdfa-related  
>>> work
>>> at http://nomisma.org/ .

>>> The overall idea for nomisma is to provide stable URIs for  
>>> numismatic
>>> entities and to describe those entities using rdf/rdfa/linked data.

>>> As a test, we've made the text of Inventory of Greek Coin Hoards
>>> available. At this point there are URIs similar to
>>> http://nomisma.org/id/igch0156 for all 2387 hoards described in that
>>> work.

>>> If you click-through to igch0156, you'll see links at the bottom of
>>> the page. "Show Markup in Page" does just that. "rdf n3 triples"  
>>> links
>>> to the w3 rdf extractor that makes use of the embedded rdfa.

>>> At this stage of the game, I am working to develop an xhtml/rdfa
>>> based convention for marking up hoards of ancient coins. I have two
>>> main goals: make the information visible to the rdfa extractor, and
>>> also keep the markup simple enough so that I can teach it to our
>>> not-very-technical future collaborators.

>>> I'm trying to keep myself to the following patterns for the rdfa:

>>> <a rel="nm:*" href="relative url"></a>
>>> <span property="nm:*" content=""></span>

>>> It seems clear, however, that I won't be able to express all the
>>> information that I want in ways that are visible to a single-pass of
>>> an rdf extractor.

>>> For example, the line "Contents: c. 260 AR" indicates that all the
>>> coins are silver. I don't think there is rdfa that will generate
>>> triples to that effect for both the Euboean and Histiaean coins. So
>>> for now I group the information in hierarchical instances of '<span
>>> class="nm:g"></span>' where nm:g is short for group. This  
>>> anticipates
>>> my writing a style sheet to generate rdf describing 18 instances of
>>> silver drachmas from Histiaea. Perhaps I'll indicate this with  
>>> GRDDL.

>>> And I want to work on linking my graphs to external graphs.

>>> For example, http://nomisma.org/id/phaselis has the markup '<a
>>> rel="owl:sameAs"
>>> href="http://pleiades.stoa.org/places/639051">pleiades</a>'. Is that
>>> actionable in both directions between nomisma and pleiades? Not yet
>>> but what are the conventions I should use to make it so?

>>> This is all extremely preliminary so I will appreciate any feedback.
>>> Though I do recognize that the above leaves much to be desired as an
>>> introduction.

>>> -Sebastian

>> Hi Sebastian,

>> What would you think about sharing a findspot property with the IAph
>> data?

>> I see an immediate problem with asserting that your Phaselis is the
>> same as http://pleiades.stoa.org/places/639051: the latter URI
>> identifies not a place but a description of a place. We really need  
>> to
>> address this soon on the Pleiades end. We could obtain non-
>> information
>> resources for free by inviting everybody to use URIs like http://pleiades.stoa.org/places/639051#foo
>> , but I'd rather not.

>> Cheers,
>> Sean

> owl:sameAs is there as somewhat of a stub/conversation starter so I
> agree that it overstates the case. Any current ideas on how I can
> indicate a rel with pleiades? I'll keep something in there so the
> triple exists but will update on the basis of your future work.

> So... this is all a draft with lots of stubs and opportunities to
> display my obvious missteps. Glad to have feedback as I go along.

> -Sebastian

> On Thu, Jan 22, 2009 at 6:26 PM, Sean Gillies  
> <sean.gill...@gmail.com> wrote:

>> Hi Sebastian,

>> What would you think about sharing a findspot property with the IAph
>> data?

>> I see an immediate problem with asserting that your Phaselis is the
>> same as http://pleiades.stoa.org/places/639051: the latter URI
>> identifies not a place but a description of a place. We really need  
>> to
>> address this soon on the Pleiades end. We could obtain non-
>> information
>> resources for free by inviting everybody to use URIs like http://pleiades.stoa.org/places/639051#foo
>> , but I'd rather not.

>> Cheers,
>> Sean

>> --
>> Sean Gillies
>> Software Engineer
>> Institute for the Study of the Ancient World
>> New York University

> ORE recommends that? That seems to turn the Hash URI approach shown at http://www.w3.org/TR/cooluris/#hashuri
>  inside out.

> Sebastian's RDFa does the right thing in using @about to specify that
> a non-information resource is the object of his triples. Pleiades just
> doesn't yet provide all the corresponding non-information resources
> for him.

> Sean

> On Jan 23, 2009, at 10:31 AM, Benjamin Armintor wrote:

>> I'm reminded here about the ORE recommendation, and the use of <uri>
>> to refer to the thing, and <uri>#description to refer to the
>> description.  I know this isn't without problems (not least of which
>> is the tendency to resolve the non-information URI's to the
>> description), but it does allow the relationship between things to be
>> expressed in a way that seems more natural to me (that is, it makes
>> more sense to me to say "this object is related to this place" than
>> "this object is related to a place that is is also described by this
>> place description").  I may also be over-interpreting Sean's concerns,
>> or misinterpreting Sebastian's rdfA.  Just sayin'.

>> - Ben

>> On Thu, Jan 22, 2009 at 6:26 PM, Sean Gillies
>> <sean.gill...@gmail.com> wrote:

>>> On Jan 21, 2009, at 9:53 PM, Sebastian Heath wrote:

>>>> This is just a brief notice of some very preliminary rdfa-related
>>>> work
>>>> at http://nomisma.org/ .

>>>> The overall idea for nomisma is to provide stable URIs for
>>>> numismatic
>>>> entities and to describe those entities using rdf/rdfa/linked data.

>>>> As a test, we've made the text of Inventory of Greek Coin Hoards
>>>> available. At this point there are URIs similar to
>>>> http://nomisma.org/id/igch0156 for all 2387 hoards described in that
>>>> work.

>>>> If you click-through to igch0156, you'll see links at the bottom of
>>>> the page. "Show Markup in Page" does just that. "rdf n3 triples"
>>>> links
>>>> to the w3 rdf extractor that makes use of the embedded rdfa.

>>>> At this stage of the game, I am working to develop an xhtml/rdfa
>>>> based convention for marking up hoards of ancient coins. I have two
>>>> main goals: make the information visible to the rdfa extractor, and
>>>> also keep the markup simple enough so that I can teach it to our
>>>> not-very-technical future collaborators.

>>>> I'm trying to keep myself to the following patterns for the rdfa:

>>>> <a rel="nm:*" href="relative url"></a>
>>>> <span property="nm:*" content=""></span>

>>>> It seems clear, however, that I won't be able to express all the
>>>> information that I want in ways that are visible to a single-pass of
>>>> an rdf extractor.

>>>> For example, the line "Contents: c. 260 AR" indicates that all the
>>>> coins are silver. I don't think there is rdfa that will generate
>>>> triples to that effect for both the Euboean and Histiaean coins. So
>>>> for now I group the information in hierarchical instances of '<span
>>>> class="nm:g"></span>' where nm:g is short for group. This
>>>> anticipates
>>>> my writing a style sheet to generate rdf describing 18 instances of
>>>> silver drachmas from Histiaea. Perhaps I'll indicate this with
>>>> GRDDL.

>>>> And I want to work on linking my graphs to external graphs.

>>>> For example, http://nomisma.org/id/phaselis has the markup '<a
>>>> rel="owl:sameAs"
>>>> href="http://pleiades.stoa.org/places/639051">pleiades</a>'. Is that
>>>> actionable in both directions between nomisma and pleiades? Not yet
>>>> but what are the conventions I should use to make it so?

>>>> This is all extremely preliminary so I will appreciate any feedback.
>>>> Though I do recognize that the above leaves much to be desired as an
>>>> introduction.

>>>> -Sebastian

>>> Hi Sebastian,

>>> What would you think about sharing a findspot property with the IAph
>>> data?

>>> I see an immediate problem with asserting that your Phaselis is the
>>> same as http://pleiades.stoa.org/places/639051: the latter URI
>>> identifies not a place but a description of a place. We really need
>>> to
>>> address this soon on the Pleiades end. We could obtain non-
>>> information
>>> resources for free by inviting everybody to use URIs like http://pleiades.stoa.org/places/639051#foo
>>> , but I'd rather not.

>>> Cheers,
>>> Sean

> --
> Sean Gillies
> Software Engineer
> Institute for the Study of the Ancient World
> New York University