[TagCommons-WG] mechanisms for sharing tag data

[Marja Koivunen]

Nitin Borwankar wrote:
> Marja Koivunen wrote:
>
>> Nitin Borwankar wrote:
>>
>>> Marja Koivunen wrote:
>>>
>>>> I vote for keepings things simple for users. That's why we have 
>>>> tags anyway. At least that's why I created tags in Annotea. And 
>>>> that's probably why so many people in our field did not understand 
>>>> it at first. But we are doing tools for users who are not so exact, 
>>>> who are not in our field, and who do not necessarily know what 
>>>> ontology means and even if they do they need to learn it.
>>>>
>>>> Annotea has a basic schema already. Why cannot we use that? It can 
>>>> be easily extended when needed. And what best , other available 
>>>> metadata can be combined with that.
>>>>
>>>>  
>>>>
>>>
>>> Tom,
>>>
>>> Could we have a blog post inviting candidate schemas/datamodels and 
>>> people
>>> can provide links in the comments if they want to submit something 
>>> to be considered.
>>> Please keep in mind that we will need to consider schemas that are 
>>> not focused on one particular type such as bookmarks.
>>
>> Type of what? I'm confused here. What are the other particular types 
>> except a bookmark? Is there a link that explains your thinking here?
>
>
> Hi Marja,
>
> Externally facing web applications are accessible via URL's to which 
> you can attach tags.
> But tags can be attached to data in applications that are not web 
> applications and that don't use URL's for addressing.  Tag usage is 
> not resticted to web applications.
>
> A simple example of that would be a rich client desktop application 
> that is used for managing email and allows tags to be attached to 
> email messages in the inbox.
Yes, and then you need the messages to have some (hopefully unambiguous)
ids, which could be anything but as well they could be URIs.
> Here we need to be able to discuss tags without using a bookmark 
> metaphor - one of the reasons to actively be aware of and avoid 
> putting implicit assumptions in the model.
I still don't know why I could not bookmark my e-mail messages, the
bookmark metaphor does not prevent that in any ways. This is actually
one thing early on I wanted to do when I suggested an Annotea demo
annotating/bookmarking spam messages at W3C.

Short time later the idea was warmed up and realized:
1) id's were added to the e-mail messages, and
2) a client (not integrated to e-mail application like in my vision but
a good starting point) was created for making the annotations.

Today this is in use to filter spam from W3C discussion lists. And I
have ideas for other applications as well.

Marja
>
> So in your model the implict assumptions would include addressability 
> via URL's which restricts discussion to web applications only.
>
> Hope that clarifies.
>
> Nitin
>