[tei-council] "global" @source [was Re: Fwd: RE: @resp]

Thu May 30 13:12:47 EDT 2013

On 13-05-30 09:58 AM, Gabriel Bodard wrote:
> So when you say you have this in hand, you mean you intend to come up
> with a full proposal for which new elements/classes you would like @resp
> to be available on?

I was tasked with adding some use-cases from a variety of contexts 
supporting my contention that @resp ought to be available in lots of 
places. Personally, I still think it's simpler and cleaner to add it to 
att.global, but there is substantial resistance to that, so I guess 
we're going to end up with a list of classes or elements that meet the 
use-cases. I think att.sourced is a similar case; at the moment, it's 
only available on quote q writing egXML, but I can see arguments that it 
should be much more widely available too. Your final comment on the 
ticket suggests that it's @source that you need, rather than @resp. I 
think the arguments tend to go in parallel for these two attributes; for 
instance:

<pron resp="#fred"> (Fred is responsible for attesting to this 
pronunciation)

<pron source="#lass"> (The biblio item "lass" is the source of this 
pronunciation)

So I think we should broaden the ticket and handle both together. We're 
talking about _attribution_ of two different kinds; perhaps there's an 
argument for att.attribution or some such thing.

Cheers,
Martin

>
> Should we discuss @source separately, or in parallel?
>
> G
>
> On 30/05/2013 12:15, Martin Holmes wrote:
>> I have an open ticket on @resp:
>>
>> <http://sourceforge.net/p/tei/feature-requests/443/>
>>
>> which I haven't had time to proceed with. I do have it in hand, though.
>>
>> Cheers,
>> Martin
>>
>> On 13-05-30 07:42 AM, James Cummings wrote:
>>> On 30/05/13 15:31, Gabriel Bodard wrote:
>>>> I don't think there's a ticket for this yet (please correct me if I've
>>>> missed it) but this question (of making @source (a) available more
>>>> widely than just on quote, egXML, etc., and (b) expanding its semantics
>>>> to the source of a piece of information, datum, translation, encoding
>>>> rather than just a quotation) is intimately tied up with
>>>> https://sourceforge.net/p/tei/feature-requests/443/ (on making @resp
>>>> more widely available).
>>>>
>>>> As I understand and remember it, we had gotten as far as agreeing that
>>>> both @resp and @source would usefully be more widely available than they
>>>> currently are. (We would probably also agree that neither of these
>>>> should be technically global attributes.) But how do we go forward to a
>>>> decision somewhere in between the two extremes?
>>>
>>> Yes, I seem to remember specifically quite some resistance to the
>>> idea that either of these be truly made global.
>>>
>>>> I think the way forward is to collect (in a ticket? in a wiki page? in
>>>> emails to the list/to me?) specific and documented use-cases of elements
>>>> which we need to be able to attribute to a particular encoder, or whose
>>>> content we need to attribute to a bibliographical source somewhere. From
>>>> these examples, we should try to come up with a coherent proposal for
>>>> the extension of both of these attributes.
>>>
>>> I've not personally got any use-cases to hand where I've felt I
>>> needed this, but agree with some of the arguments for extending
>>> it more widely. I'm really not sure how to implement that
>>> sensibly or where to draw the line in this case.
>>>
>>> -James
>>>
>>>>
>>>> Does this seem reasonable? Does anyone else want to collate this
>>>> information? Any preference as to where/how we do this? (Do we need a
>>>> new ticket alongside FR 443, or should we have both conversations
>>>> together in there?)
>>>>
>>>> Gabby
>>>>
>>>> On 02/02/2013 14:39, Gabriel Bodard wrote:
>>>>> Surely the definition of @source only contains the word "quotation"
>>>>> because it was designed for <q>, <quote>, etc. If we consider it a
>>>>> suitable mechanism for indicating the bibliographic source of a set of
>>>>> dimensions, for example, then that definition would have to change.
>>>>>
>>>>> [Aside: when marking up apparatus criticus, I use @resp to point to
>>>>> bibiographical references for readings and conjectures, not to persons.
>>>>> Does this mean @source would be more appropriate for this use?]
>>>>>
>>>>> G
>>>>>
>>>>> On 02/02/2013 19:14, James Cummings wrote:
>>>>>> Hi all,
>>>>>>
>>>>>> I can see att.responsibility being made available more generally
>>>>>> (I'm very reluctant to say globally until I've really sat down
>>>>>> and thought about the implications of that...) I understand it
>>>>>> may have once(?) been intended for editorial intrusions into a
>>>>>> transcription or edition but believe we've generalised out such
>>>>>> indications to refer to any markup or encoding.  Maybe it was
>>>>>> always intended as such.
>>>>>>
>>>>>> I can see arguments for @source (which is where this started
>>>>>> right?) on more things and that att.editLike should get it from
>>>>>> the att.source class.  But, I think we have to be careful that it
>>>>>> is available only on things which can be classified as containing
>>>>>> a 'quotation or citation' in some way since it "provides a
>>>>>> pointer to the bibliographical source from which a quotation or
>>>>>> citation is drawn."  Either that, or this definition would have
>>>>>> to be changed.
>>>>>>
>>>>>> Why does egXML get @source and not eg incidentally?
>>>>>>
>>>>>> -James
>>>>>>
>>>>>>
>>>>>>
>>>>>> On 02/02/13 18:27, Gabriel Bodard wrote:
>>>>>>> Seconded.
>>>>>>>
>>>>>>> I (as I pointed out in another venue recently) am regularly surprised to
>>>>>>> re-learn that @resp isn't global already. I can't imagine any element
>>>>>>> that I would not want to be able to say either who is responsible for
>>>>>>> the decisions it represents, or from what publication the information so
>>>>>>> tagged comes. (Certainly everything in msDesc, as well as editLike, at
>>>>>>> the very least.)
>>>>>>>
>>>>>>> G
>>>>>>>
>>>>>>> On 02/02/2013 17:07, Martin Holmes wrote:
>>>>>>>> On 13-02-02 02:47 AM, Lou Burnard wrote:
>>>>>>>>> Tomaz has a good point here. Presumably att.editLike should inherit the
>>>>>>>>> @source attribute from att.sourced ?
>>>>>>>>>
>>>>>>>>> I am waiting for someone to want @source to be added to att.global...
>>>>>>>>
>>>>>>>> And that would be me. I can imagine a use-case for virtually any
>>>>>>>> element. I'd also like @resp to be global, incidentally -- same
>>>>>>>> argument. I need to assign responsibility for <pron>, <seg>, <def> and
>>>>>>>> all sorts of other bits and pieces in a dictionary project I'm working on.
>>>>>>>>
>>>>>>>> Cheers,
>>>>>>>> Martin
>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> -------- Original Message --------
>>>>>>>>> Subject: RE: @resp
>>>>>>>>> Date: Sat, 2 Feb 2013 10:09:19 +0100
>>>>>>>>> From: Tomaz Erjavec <tomaz.erjavec at ijs.si>
>>>>>>>>> To: 'Lou Burnard' <lou.burnard at retired.ox.ac.uk>
>>>>>>>>> CC: <TEI-L at LISTSERV.BROWN.EDU>
>>>>>>>>>
>>>>>>>>> ...
>>>>>>>>>
>>>>>>>>> @source is the only attribute defined in the att.source class (with
>>>>>>>>> quote q writing egXML as members).
>>>>>>>>> But @source is also defined (directly, not via class) as an attribute of
>>>>>>>>> att.editLike, so in fact quite a lot of other elements already have it.
>>>>>>>>> Is there any particular reason that it is defined in two different
>>>>>>>>> places? I'd say it only confuses things.
>>>>>>>>> And, yes, it would probably be a good idea to have source on even more
>>>>>>>>> elements, e.g. person and all its descendants.
>>>>>>>>> Best,
>>>>>>>>> Tomaž
>>>>>>>>>
>>>>>>>>> -
>>>>>>>>>
>>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>
>>>
>>>
>>
>

-- 
Martin Holmes
University of Victoria Humanities Computing and Media Centre
(mholmes at uvic.ca)