[Accessforall] Minutes of AccessForAll Meeting on 2012-01-31

Liddy Nevile liddy at sunriseresearch.org
Thu Feb 9 22:46:13 UTC 2012


when it comes to a registry, I suggest we ask Mitsuharu Nagamori to  
give us his - he has built the registry for the Dublin Core Initiative  
and other things and I think as a trusted and ready-to-roll registry,  
we could not do better. In fact, I am not sure we even need our own  
one - we could simply add our schemas etc to the existing registry, I  
think, to our benefit.

Liddy

On 10/02/2012, at 9:12 AM, Clark, Colin wrote:

> Hi All,
>
> And from my end, I had a minor epiphany while on the subway this  
> morning that helps me join the violent agreement. Here's a summary:
>
> * Gottfried's proposal fully makes sense to me now, and I agree with  
> it. The registry will contain a flat structure, and URLs are fine  
> namespaces.
> * Unlike the UIC example we've been bouncing around, our website and  
> registry should be more user friendly. As Liddy says, URL are for  
> machines, words are for people. We'll just need to design a better  
> interface for the registry.
> * The specific mechanism for representing the namespace is an  
> implementation-specific detail. In XML, they'll be represented using  
> colons and the xmlns attribute, and in other formats they may be  
> represented differently.
> * As long as nothing in the standard prevents implementations from  
> representing ISO 24751 using hierarchical (or any other) data  
> structures, I'm fine with whatever structure we like within the  
> specification and registry. Flat is a-okay.
>
> Gregg and I chatted further on the phone about his tags idea, and he  
> clarified that the tags were only designed for the registry system  
> itself, not part of the preferences profile. In other words, they're  
> a tool on our website to enable users to sort, search, and view the  
> registry's content. Sounds good to me.
>
> Colin
>
>
> On 2012-02-09, at 4:23 PM, Liddy Nevile wrote:
>
>> Gregg
>>
>> I think we are in 'violent agreement' - excellent!!!
>>
>> Liddy
>>
>>
>> On 10/02/2012, at 8:19 AM, Gregg Vanderheiden wrote:
>>
>>>
>>>
>>>
>>> On Feb 9, 2012, at 3:10 PM, Liddy Nevile wrote:
>>>
>>>> Colin
>>>>
>>>> I am sure that we will have a problem because some of us see this  
>>>> as a metadata problem and some don't. I don't know another word  
>>>> or way of describing what we are doing as I think we are trying  
>>>> to use descriptions of resources and services to adapt what is  
>>>> delivered to the user to their needs and preferences.
>>>>
>>>> Anyway, I am wondering why this group wants to reinvent how to do  
>>>> metadata? URIs are for machines - words are for humans. If you  
>>>> look at the ISO standard for metadata (N19788) you will see that  
>>>> there are words for humans and URIs for machines ????
>>>
>>> What are you trying to say?    The resource descriptions are for  
>>> Machines to use to match resources to user needs.    So are you  
>>> arguing FOR urls?
>>>
>>>>
>>>> As for what Gregg is calling tags - I am really confused. I don't  
>>>> know what they are.
>>>
>>>> I think what he wants (what you want, Gregg) is an easy way of  
>>>> looking at the property-value pairs once we have them - so a  
>>>> person can have a collection that is relevant to them and forget  
>>>> the rest?
>>> close.   it is so that they can more easily FIND one when they  
>>> don't know what it is called so can't search by its name.
>>>
>>>> or so they have the ones that are useful to them,even though in  
>>>> another context they might be useful to someone with a very  
>>>> different set of needs?
>>>
>>> not this.    TAGS are for searching the listing more easily.  They  
>>> have nothing to do with use.
>>>
>>>
>>>> For example, if my problem is difficulty with hearing but I have  
>>>> good eyesight, I want a collection of things to do with sound to  
>>>> work on but I won't need special things for sight.
>>> This is not what this is about.   There will be tools that make it  
>>> easy to determine what helps you.  no human should have to go  
>>> digging through the huge preference file and trying to make sense  
>>> of it.  Users should never see the preference registry.
>>>
>>>> People who work with metadata create sets like this by having  
>>>> what they call faceted views - or they have application profiles  
>>>> - maybe a set for mildly deaf users, and a set for severely deaf  
>>>> users??
>>> What is proposed allows all this and more.
>>>>
>>>> I think that fortunately many of the problems being raised here  
>>>> have been worked on for many years now by metadata experts and  
>>>> the solutions are not as far away as they seem.
>>> Not sure what you mean.    Who said the solutions were far  
>>> away?    With all the work that has been done - I think we are  
>>> very close to what we need.
>>>>
>>>> Liddy
>>>>
>>>>
>>>>
>>>> On 10/02/2012, at 3:59 AM, Clark, Colin wrote:
>>>>
>>>>> Hi Liddy,
>>>>>
>>>>> On 2012-02-09, at 1:39 AM, Liddy Nevile wrote:
>>>>>> I think we should remember that from the perspective of JTC1, a  
>>>>>> lot of work has been done to make a standard for metadata that  
>>>>>> is the simplest and most used way of providing and implementing  
>>>>>> metadata. I think at least for the purposes of JTC1, this  
>>>>>> expert advice should be noted. Beyond that, as most major  
>>>>>> entities in Europe use the sort of metadata advocated now by  
>>>>>> JTC1, I see no reason for trying to invent a better way of  
>>>>>> doing things?
>>>>>
>>>>> We're actually 100% in agreement here. I perhaps should have  
>>>>> been more explicit, but I see ISO 24751 as exactly an example of  
>>>>> that "judicious use of containment" that I was referring to in  
>>>>> my previous email.
>>>>>
>>>>> What we do need is to come up with a reasonable extension  
>>>>> mechanism, which is representation-independent (i.e. it can't  
>>>>> depend on, say, XML namespaces), so that communities can  
>>>>> contribute their own properties easily and in a way that won't  
>>>>> collide with others.
>>>>>
>>>>>> Eric Miller has pointed out that the magic of URIs, as Tim BL  
>>>>>> says, is that they can always be disambiguated and they make it  
>>>>>> easy for people to add their own. I am not an expert in  
>>>>>> optimization but I do know that the use of URIs does not mean  
>>>>>> that people get lost because everything is URIs instead of  
>>>>>> words. The usual implementations have clear instructions and  
>>>>>> wizards for humans that 'speak' about the definitions of the  
>>>>>> terms represented by the URIs in ways that are appropriate to  
>>>>>> the audience using the wizards - and these vary enormously  
>>>>>> without any effect on the resulting metadata.
>>>>>
>>>>> I get the magic of URIs. My point is about human users of the  
>>>>> specification. Compare Gottfried's example (http://myurc.org/TR/res-prop-vocab1.0/ 
>>>>> ) to the ISO 24751 spec. Which one is easier to understand, at a  
>>>>> glance?
>>>>>
>>>>> The optimization question is straightforward: we just need to  
>>>>> come up with a non-redundant technique for solving the very  
>>>>> important issue that Gottfried raises with his proposal:  
>>>>> namespacing.
>>>>>
>>>>> Hopefully this clarifies,
>>>>>
>>>>> Colin
>>>>>
>>>>> ---
>>>>> Colin Clark
>>>>> Lead Software Architect,
>>>>> Inclusive Design Research Centre, OCAD University
>>>>> http://inclusivedesign.ca
>>>>>
>>>>
>>>> _______________________________________________
>>>> Accessforall mailing list
>>>> Accessforall at fluidproject.org
>>>> http://lists.idrc.ocad.ca/cgi-bin/mailman/listinfo/accessforall
>>>
>>
>> _______________________________________________
>> Accessforall mailing list
>> Accessforall at fluidproject.org
>> http://lists.idrc.ocad.ca/cgi-bin/mailman/listinfo/accessforall
>
> ---
> Colin Clark
> Lead Software Architect,
> Inclusive Design Research Centre, OCAD University
> http://inclusivedesign.ca
>



More information about the Accessforall mailing list