Re: Starting with Contexts and Schema.org

Hi Vikash,

Sounds like you've gained good insights – it's usually in these kinds of
sprints that you see the good patterns emerge.

Regarding contexts for various schema.org examples, I think the principle
is to start with "@vocab" set to "http://schema.org/". Then, to compact and
simplify uses of datatypes (dates and such), declare those terms in the
context with "@type" set to an appropriate datatype (in regular RDF, those
are normally the XSD datatypes). And for cases where you know there will be
more than one value, declare "@container": "@set". (Or, if those values
should be ordered, such as "track" for "MusicAlbum", use "@list".)

Note that, ideally, this might actually become unnecessary, since
schema.orgmay eventually publish a complete, full context to be linked
to instead.
But this work can reasonably serve as input to that, regarding how certain
details should look. There are some good examples of what that full context
might become, such as this one by Sandro Hawke:

    http://www.w3.org/People/Sandro/schema-org-context.jsonld

But in your task, I suggest making, small, concise contexts for the various
examples, and look at e.g. the above, full context when in doubt. I suspect
that most examples will need no more than a handful of terms specified to
enable syntactic brevity, with the rest being resolved against "@vocab".

(In certain advanced cases, it might be necessary to instead spell out all
the terms, and not use "@vocab" at all. Specifically if you intend to mix
schema.org with other vocabularies, and need to manage all the terms
explicitly to avoid conflicting names. Whether or not this is a good
pattern for these examples I cannot tell, though I'd recommend waiting with
it until needed. Others may advice differently.)

Keep up the good work!

Cheers,
Niklas



On Thu, Jul 18, 2013 at 8:01 PM, vikash agrawal <vikashagrawal1990@gmail.com
> wrote:

> hi Everyone,
>
> Finally some good news after loads of mistakes and errors. finally the
> website has been completed for a while (the indentation part is just being
> postponed for a while). Indeed there were a lot of half-baked things and I
> have tried to finish them as soon as I could.
>
> I believe I made some serious errors specially with opening multiple bugs
> of same nature and not closing it or using it in the right way.But yes, so
> far the learning has been great. One thing I learnt is sure, commit small,
> commit early :)
>
> I would love to extend my gratitude to everyone, who have patiently heard
> me and took out time to help me in all circumstances :-) Manu, Markus,
> Dave, David and everyone else! Thank-You so much!
>
> I am making some quick progress in achieving the targets and deadlines and
> the flame once started shall not go. Now can I get some more inputs over
> writng contexts apart from reading the spec :P
>
> Regards
> ~Vikash
>
>
> On Wed, Jul 17, 2013 at 6:11 PM, Markus Lanthaler <
> markus.lanthaler@gmx.net> wrote:
>
>> On Monday, July 15, 2013 8:55 PM, vikash agrawal wrote:
>> > hi Everyone,
>> >
>> > I am willing to start my learning and writing process of Contexts and
>> > Schema for the JSON to JSON-LD creator tool. But I am very new to the
>> > subject, so I am looking for some good resources and helping hand. A
>> > short and formal introduction and tutorial will be very beneficial
>> > too.
>> >
>> > For me, this subject is something very new, so I am looking to
>> > understand what is it in whole.
>>
>> The best thing is probably to read the spec :-P
>>
>>
>> > The end reasults of my agenda is to write contexts for Person, Place
>> > and Events and use it in the tool.
>>
>> While I understand that it is more exciting to switch between different
>> tasks, I would like to reiterate my request to first finish the website
>> re-design. The site is still buggy (and has been so for quite some time
>> now). This should be your top priority at the moment IMO.
>>
>> Cheers,
>> Markus
>>
>>
>> --
>> Markus Lanthaler
>> @markuslanthaler
>>
>>
>>
>>
>>
>