Re: users vs topics

On 7/27/12 4:33 AM, Michiel de Jong wrote:
> Hi Kingsley,
>
> i prefer to start with modeling the stuff we need first, so for users
> just avatar, a free-text full name, maybe city and gender, and then a
> list of tools (read, subscribe, comment, message).

And as I said, you have a model for everything via the 
entity-attribute-value model. If you want to get started, with 
structured data schemas/vocabularies/dictionaries for the specific items 
above, then take a look at SIOC, Schema.org, and other sources for 
existing work that you can use or build upon etc..
> That way we can
> chart a big part of the federated social web already i think, except
> for things like mailing lists and chat channels.

As per comment above, a lot of work in this area has already been done, 
especially via the SIOC project.

>   So that's what i
> wanted to bring up in this thread. Maybe i didn't phrase my question
> clearly enough, sorry.

Links:

1. http://sioc-project.org/ -- note, this is a generic model for 
personal or enterprise oriented data spaces which my be hosted or unhosted
2. http://schema.org/docs/schemas.html -- schema.org effort
3. http://schema.rdfs.org/mappings.html -- some schema.org mappings .

Kingsley
>
> On Thu, Jul 26, 2012 at 5:52 PM, Kingsley Idehen <kidehen@openlinksw.com> wrote:
>> You need to model everything.
> Everything? although i doubt you really mean what you say there, i'm
> afraid that might be an uncomputable problem ;)
>
> When you write an application you should take care not to try to write
> an operation system instead. An application has a specific function,
> which is limited and is not 'it should do everything' (that is what
> operating systems and programming languages are for).
>
> if next year it becomes a big thing on the web to be friends with an
> elephant, then we might want to add a 'species' field to the
> useraddress.net search results, but for now that seems like overkill.
>
> as i thought a bit more about the concept of 'topics', i felt that
> maybe they are specific types of 'groups', because a chat session is
> defined ultimately by the people participating in it, and only
> secondarily by the topic that these people agreed on. it is also
> possible to have a group or group chat without a preset topic. So i
> think i might use 'group' instead of 'topic'.
>
> Anyway, it's just a proof-of-concept, so we can always change it later
> if we get new insights.
>
>
>


-- 

Regards,

Kingsley Idehen	
Founder & CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen