Scott and all

Thanks for this very helpful refocusing on verification rather than 
storage. I agree with all the points made on assertions and verification. 
Certainly verification is what really matters from a practical, user point 
of view.

But I still think that the facts of storage are significant to a number of 
the people we may want to serve, for several reasons.

Firstly, responsibility. I want to know who to refer to about the use (and 
abuse) of information about me. Clearly this is a different scenario from 
ones where the information in question is free for anyone to view on the 
web. In this context, what matters is probably which data controller or 
data processor we are dealing with, not which of the physical machines 
under their authority actually holds the bits.

Secondly, comfort. While people are catching up with these leading ideas 
about how to conceptualise the holding of personal information, it may well 
be reassuring to know where information is held, as that provides the start 
of a possible investigatory trail if things were to go wrong. Much the same 
applies to my wanting to know the physical address of any supplier, or the 
issuer of a cheque. Also, topically, it is physical places that suffer from 
damage from unexpected explosions, as well as the virtual places that 
suffer from electronic intrusion. And from the cognitive point of view, 
having locations in mind may help people conceptualise what is happening to 
their information. Probably better if the locations people use for helping 
their conceptualisation relate to reality.

Thirdly, several groups of people have an interest in how and where the 
information is stored. Data controllers and data processors, as above, for 
obvious reasons. Providers of electronic storage and the associated systems 
and security, for resource / payment reasons.

When I wrote "perhaps one can qualify Stephen's original: it might be best 
only to have one location that acts as the identifier of the individual, 
but information relating to that individual can easily be distributed 
across several locations, depending on the various authorities for that 
information", the "locations" I have in mind are like the L of URL, rather 
than sector, track, disk, machine, room, building numbers. Perhaps it would 
be helpful when discussing this to view both storage and verification with 
this "URL" sense of "location" rather than the physical sense. Reading the 
extract including "Stephen's first rule of metadata" quoted by Mike, it 
seems clear to me that Stephen is using the concepts of "place" and 
"location" in this sense, not the physical sense.

Simon

At 07:23 2005-07-18, Scott Wilson wrote:
>Hi Simon,
>
>I think this reads better if you replace "stored" with "verified". Storage 
>isn't really very relevant, as bits can exist all over the place, and I'm 
>not sure "managed" is possible either. Various companies have all kinds of 
>opinions about how to spell my address, for example, and I'm not able to 
>manage that as I don't have access to the storage medium they each use - 
>but I can verify my own address if asked, so I'm still an authority.
>
>I think its easier to conceptualize this problem in assertions - agent x 
>asserts property y about subject z. Clearly when I assert that my name is 
>Scott (Scott, name="scott", Scott), I can be counted as the source of 
>verification about that statement. On the other hand, other agents may 
>make similar statements (UWB personnel dept, name="scott", Scott), but in 
>this case the source of verification may be the agency making the 
>assertion, or it may be the subject of the assertion; the verifier may 
>have to make a choice of who to ask depending on the nature of the assertion.
>
>Clearly the best agent to judge personal opinions and general biographical 
>facts is usually the subject; but when it comes to academic awards a more 
>'independent' agent may be preferred. Ultimately, its up to the receiving 
>agent to decide which agents can be trusted as authorities for various 
>types of assertions about various entities.
>
>("Back in the day" I worked on CRM projects, and one particular problem 
>was the validity of customer data; how to ensure records were kept 
>accurately up to date. The solution chosen was to enable role-cetric 
>update permissions; basically, if you're role relied upon the accuracy of 
>a data point, then you had the right to change it. So, for example, sales 
>got to update phone numbers, but billing got to update addresses.)
>
>I remember a conversation along similar lines several years ago, after the 
>announcement of Edutella; there was some disquiet about the idea of 
>collating assertions from multiple agents about a resource, as this 
>contradicts the idea of "objective metadata". Personally I've always felt 
>that all metadata was subjective - even the old RDBMS standby of creating 
>automatic timestamps via triggers are still assertions by the RDBMS based 
>on its internal clock, rather than objective truth!
>
>Going back to the aggregation example, its simple enough to package a 
>collection of assertions from various agents and pass it onto another 
>party; that party can then make judgements about what assertions they are 
>willing to trust without verification, and what verification measures to 
>perform on the rest.
>
>-S
>
>
>On 13 Jul 2005, at 19:52, Simon Grant wrote:
>
>>Looking back at the first suggested rule...
>>
>>We have considered the multiple authority question very much in the field 
>>of e-portfolio systems. Information about contact details would be best 
>>stored in one place under the control of the individual. Information 
>>about qualifications, grades, marks would perhaps better be stored 
>>authoritatively by the institution awarding them. It partly goes back to 
>>the question of what one counts as metadata about what. There may be a 
>>reasonable intepretation of "metadata" under which Stephen's first rule 
>>makes sense. Metadata about qualifications could best be seen in their 
>>own right and not as metadata about people. And opinions are their own 
>>entities. The authority for an opinion is the person holding the opinion, 
>>not the object of the opinion, so perhaps that's not a divisive issue.
>>
>>If one adopts the URI-as-identifier view, then it still makes sense that 
>>there should be one "way in" to the metadata about a person, even if the 
>>information presented (rather than held) there is actually aggregated 
>>from a number of different authoritative sources. Perhaps one could allow 
>>for common practice by saying that the details of a student held by an 
>>educational institution are not actually metadata about that student, but 
>>metadata about the institution's "customer", who may or may not be 
>>identical to (or identified with) a personally-managed record.
>>
>>URI as ULN, now there's a provocative thought (probably not original...)
>>
>>So, perhaps one can qualify Stephen's original: it might be best only to 
>>have one location that acts as the identifier of the individual (and 
>>perhaps the "source"), but information relating to that individual can 
>>easily be distributed across several locations, depending on the various 
>>authorities for that information.
>>
>>Simon
>>
>>At 16:13 2005-07-12, Mike Collett wrote:
>>>Stephen's first rule"
>>>  "metadata for a given entity should never be stored in more than one 
>>> place.
>>>If we have, for example, metadata about a given person (say, me), then it
>>>should be stored in one and only one location. (That does not mean that it
>>>cannot be aggregated or mirrored, but it does mean that there is one and
>>>only one location that would constitute the source of information about this
>>>person, and that aggregators and mirrors would update on a regular basis
>>>from this source."
>>>
>>>I strongly oppose this.
>>>In addition to the generic danger  of "eggs in baskets", learning records
>>>about a person are complex and managed variously by the person, the
>>>state(s), by learning organisations and by awarding bodies etc. These
>>>authorities have various responsibilities to maintain information and
>>>protect levels of privacy. To centralise this process is not a good idea.
>>>
>>>However: to suggest that a particular piece of information has a single
>>>authority makes more sense. As long as there can be many authorities for
>>>information about an entity.
>>>
>>>If the information is objective (definitive) then having a single source has
>>>advantages as pointed out by Stephen. If the information is subjective
>>>(non-definitive) then there should be the possibility for many sources for
>>>information, for example regarding curriculum use, expressing views in other
>>>languages etc.
>>>
>>>In other words to argue that it is possible to have, or even that there
>>>should ideally be, a single source (one giant metadata aggregation) for an
>>>entity is both naive and un-democratic.
>>>(I do not think that is Stephen's real intention but he did write " there is
>>>one and only one location that would constitute the source of information
>>>about this person")
>>
>>--
>>Simon Grant  http://www.simongrant.org/home.html
>>Information Systems Strategist http://www.inst.co.uk/
>>Please continue to use my established e-mail address
>>a (just by itself) (at) simongrant.org
>
>