The reasons why I like the term Object in the name is because
Object = Data + Behavior
The data aspect was already covered by JSON, but LDOM is adding
executable semantics, esp if the group agrees to include the rule
aspect. There is also the aspect of encapsulation, inheritance etc in
LDOM. From this point of view, Object is a more general term than Data,
and thus Data Object is a bit redundant. However, it is useful to have
both in the name, if only as a strong signal to the JSON and general OO
crowd.
But Karen may be right, that the term Object does not survive the
discussions in the group. I'll not give up yet though ;)
Holger
On 1/26/15, 2:09 AM, Karen Coyle wrote:
> Tom, the problem I see here is the potential conflict with OWL's
> terminology, where "DataProperty" and "ObjectProperty" are clearly
> distinct. Thus "data object" is a muddle. In addition, as Holger
> states, he comes from the OO world, where object has yet another meaning.
>
> In the W3C group we've gone around on terms, as I'm sure you can
> imagine, and "object" is one of those terms that an extra amount of
> ambiguity. Although Holger initiated his proposal as LDOM, I'd be
> surprised if the "O" survives a discussion in that group.
>
> kc
>
> On 1/25/15 7:39 AM, Thomas Baker wrote:
>> On Sun, Jan 25, 2015 at 11:05:15AM +1000, Holger Knublauch wrote:
>>> I think the direction is good, but it might be better to have a
>>> vowel in the name somewhere, so that it sounds better.
>>
>> I hesitate to put this forward, but I find the notion of a 'Data
>> Object', on its own, less objectionable than 'Linked Data Object', which
>> is ambiguous. (Does it mean "(Linked Data) Object" or "Linked (Data
>> Object)"?) After all, might any data be seen as an object, at any rate
>> by programmers?
>>
>> Hence a new proposal:
>>
>> Data Object Pattern Language for Linked Data or
>> Data-Object Pattern Language for Linked Data
>>
>> This is longer than 'Pattern Language for Linked Data', and it uses
>> 'Data' twice. However, that is perhaps the price one has to pay for
>> adding 'object'. I do not immediately see how referring to 'data' as
>> 'data object' would necessarily confuse matters conceptually.
>> 'Data-object' could be hyphenated, reinforcing the use of 'Data-Object'
>> as an adjective modifying Pattern Language. Either way, the acronym
>> would be:
>>
>> DOPL-LD
>>
>> While not a word, DOPL is at least pronounceable. 'Dopple' (or 'dopple
>> elDEE') is pleasant to the ear, with echoes of 'dapple', 'dabble',
>> 'double', and 'google', not to mention 'JSON-LD'.
>>
>> The acronym seems unencumbered with negative or confusing associations.
>> The top hits in Google have to do with the Division of Occupational and
>> Professional Licensing of the Utah Department of Commerce.
>>
>> I do not see any ambiguity in the sequence of words, which parse as
>> "(Data Object) (Pattern Language)" -- or does anyone read it as "((Data
>> Object) Pattern) Language"??
>>
>> I still prefer the shorter and straightforward 'Pattern Language for
>> Linked Data' but do not immediately see how the use of 'object' in
>> DOPL-LD would actually confuse or mislead.
>>
>> Tom
>>
>
|