0.8 Features ?

[uoccou]

Hi,

Congrats on getting 0.8 out, I look forward to trying it out soon.

1) Do you have plans to provide a CORS [1] compatible output at any
point ?
2) Are there plans for internationalisation ? Will there ever be non-
english vocabularies or equivalents ?



Ultan O'Carroll


[1]http://en.wikipedia.org/wiki/Cross-Origin_Resource_Sharing

[Mike Bergman]

Hi Ultan,

Thanks; you are the first to congratulate. I saw this, and
thought it was cute:

http://uoccou.wordpress.com/2010/10/28/pancakes/

On 11/15/2010 10:11 AM, uoccou wrote:
> Hi,
>
> Congrats on getting 0.8 out, I look forward to trying it out soon.
>
> 1) Do you have plans to provide a CORS [1] compatible output at any
> point ?

I have seen some about CORS recently, and we generally support
JSON in our standard import/export services. Our current plans
for 1.00 have not identified this enhancement, but we'd welcome
working with you (others?) on such an implementation.

> 2) Are there plans for internationalisation ? Will there ever be non-
> english vocabularies or equivalents ?

If you check out the full specs, you will see the
internationalization approach, which has been true for UMBEL
since the first version.

Yes, we'd welcome multiple language versions. One of the
motivations for version 1.00 with its mapping to Wikipedia is the
relative ease this will bring in linking alternative languages.

The idea of reference concepts is to avoid using labels as the
linking basis; rather, the idea or definition of the concept that
is the referent is the important point.

This means that multiple language potential is baked into the
UMBEL cake. We just need to complete the next Wikipedia steps,
and the pathway for multiple languages should be made much, much
easier.

BTW, these questions (plus some received offline) will cause us
to add a FAQ section to the new UMBEL web site.

Thanks, Mike

>
>
>
> Ultan O'Carroll
>
>
> [1]http://en.wikipedia.org/wiki/Cross-Origin_Resource_Sharing
>

--
__________________________________________

Michael K. Bergman
CEO Structured Dynamics LLC
319.621.5225
skype:michaelkbergman
http://structureddynamics.com
http://mkbergman.com
http://www.linkedin.com/in/mkbergman
__________________________________________

[uoc]

On 15/11/2010 17:14, Mike Bergman wrote:

Hi Ultan,

Thanks; you are the first to congratulate. I saw this, and thought it was cute:

http://uoccou.wordpress.com/2010/10/28/pancakes/

Y - I'll be coming up with more recipes soon ! Brownies next I think.

I was messing about trying to find a way of getting RDFa into wordpress.com but couldnt find a way (theres a plugin for custom installations). Maybe when I move to Drupal7...

BTW - Another place I'd love to get structured tagging based on Umbel would be Firefox  - I've got too many tags that mean the same thing, I use different terms for the same thing and so on - Im just a simple human. It would be a great way to show reduce that redundancy and duplication, and then on the retrieval side show off the possibilities for "related tags", disambiguation and sharing with others of course.

Not enough time in the day of course.

On 11/15/2010 10:11 AM, uoccou wrote:

Hi,

Congrats on getting 0.8 out, I look forward to trying it out soon.

1) Do you have plans to provide a CORS [1] compatible output at any
point ?

I have seen some about CORS recently, and we generally support JSON in our standard import/export services. Our current plans for 1.00 have not identified this enhancement, but we'd welcome working with you (others?) on such an implementation.

In my tech-demo application, I reverted to creating a server side proxy to make the Umbel call - accessing from in-page JS on another domain isnt possible - but would be very powerful for lots of page-based-apps, gadgets,  browser plugins etc. I've used JSONP successfully elsewhere, but its becoming more and more objected-to. CORS is gathering support, and is easier than doing a full blown OAuth style thing, even if there is some debate about it too. Should be easy to impl - Im a java head, so I'd probably create a servlet filter to do it, but have only glanced at the spec so far, so it would prob need some more thought. Its not much more than a whitelist, or a default "*" response to accept all requests. I'll be blogging it if I ever get around to it.

2) Are there plans for internationalisation  ? Will there ever be non-
english vocabularies or equivalents ?

If you check out the full specs, you will see the internationalization approach, which has been true for UMBEL since the first version.

Yes, we'd welcome multiple language versions. One of the motivations for version 1.00 with its mapping to Wikipedia is the relative ease this will bring in linking alternative languages.

The idea of reference concepts is to avoid using labels as the linking basis; rather, the idea or definition of the concept that is the referent is the important point.

This means that multiple language potential is baked into the UMBEL cake. We just need to complete the next Wikipedia steps, and the pathway for multiple languages should be made much, much easier.

So I could do a subject-concept lookup of "chien" (FR) and get back the same referent as looking up "dog" ?

BTW, these questions (plus some received offline) will cause us to add a FAQ section to the new UMBEL web site.

Thanks, Mike

Ultan O'Carroll


[1]http://en.wikipedia.org/wiki/Cross-Origin_Resource_Sharing

--
Ultan O'Carroll
--------------------
Archive Link

email@

Lewist TDI, SkyTwenty, Blog
Open Data Ireland

[Frederick Giasson]

Hi Ultan!

I was messing about trying to find a way of getting RDFa into wordpress.com but couldnt find a way (theres a plugin for custom installations). Maybe when I move to Drupal7...

I think Drupal7 is the best way to go for RDFa stuff in the future :)

BTW - Another place I'd love to get structured tagging based on Umbel would be Firefox  - I've got too many tags that mean the same thing, I use different terms for the same thing and so on - Im just a simple human. It would be a great way to show reduce that redundancy and duplication, and then on the retrieval side show off the possibilities for "related tags", disambiguation and sharing with others of course.

Exact

Not enough time in the day of course.

That is the problem many people are facing :)

But, with the proper planning stuff can get done, "eventually". Just check UMBEL for example, we made it after more than a year... but we made it :) It is more a question of resources than anything else I guess. Having a new partner in UMBEL's development (OntoText) is what enabled us to put some concentrated time into its development, and what will lead to a version 1 by the end of this year.

On 11/15/2010 10:11 AM, uoccou wrote:

Hi,

Congrats on getting 0.8 out, I look forward to trying it out soon.

1) Do you have plans to provide a CORS [1] compatible output at any
point ?

I have seen some about CORS recently, and we generally support JSON in our standard import/export services. Our current plans for 1.00 have not identified this enhancement, but we'd welcome working with you (others?) on such an implementation.

In my tech-demo application, I reverted to creating a server side proxy to make the Umbel call - accessing from in-page JS on another domain

Yeah, the proxy is needed in that setup.

isnt possible - but would be very powerful for lots of page-based-apps, gadgets,  browser plugins etc. I've used JSONP successfully elsewhere, but its becoming more and more objected-to. CORS is gathering support, and is easier than doing a full blown OAuth style thing, even if there is some debate about it too. Should be easy to impl - Im a java head, so I'd probably create a servlet filter to do it, but have only glanced at the spec so far, so it would prob need some more thought. Its not much more than a whitelist, or a default "*" response to accept all requests. I'll be blogging it if I ever get around to it.

Ok great! Yeah, I will have to check that too (maybe even in a structWSF context). Anyway, as I said, these web services endpoints will be replaced (enhanced) by a new set of structWSF endpoints probably by the first half of 2011.

But this demo application certainly shows one direct usage to endusers.

This means that multiple language potential is baked into the UMBEL cake. We just need to complete the next Wikipedia steps, and the pathway for multiple languages should be made much, much easier.

So I could do a subject-concept lookup of "chien" (FR) and get back the same referent as looking up "dog" ?

If a French UMBEL would exists yes. But it doesn't right now.


Take care,

Fred