[tei-council] <styleDefDecl> and @scheme

[James Cummings]

On 15/03/13 12:30, Gabriel Bodard wrote:
> Agreed re new attribute, and putting @scheme in a class.
>
> I'm not sure I like the idea of "voting" as any kind of formal method of
> resolving tickets. The TEI Council works on more of a consensus model,
> not requiring unanimity by any means, but informally weighting people's
> objections so that if I say, "I'm a bit uncomfortable with that, but if
> everyone else goes for it I'll bow to the majority" a ticket can pass,
> but if Sebastian says, "No no no, that is abject insanity" a ticket will
> probably stay open until we've all discussed it some more.
>
> What would voting gain us?

I think there probably isn't a way to restrict it to developers 
as well. So, on feature requests for example, the community could 
be motivated to vote for a particular one that they feel is 
important. So If we did enable it, I think it would only be as a 
measure of popularity of a particular feature request... for 
seting priorities.

But for the reasons you mentioned I'm a tad reluctant to do so 
for any serious decision making process. there are cases where 
voting is appropriate. (Do you want solution A or solution B?) 
But not usually in the decision to implement or not...

-James

>
> G
>
> On 2013-03-15 12:15, James Cummings wrote:
>>
>> I'd vote in favour of the new attribute, and the class.
>>
>> Coincidentally, the new SF site does allow *voting* on tickets.
>> I don't think we currently have this turned on (and I've no idea
>> how it works or what constraints there are)...  Is this something
>> we'd like to consider?
>>
>> -James
>>
>> On 14/03/13 22:05, Martin Holmes wrote:
>>> On 13-03-14 02:35 PM, Syd Bauman wrote:
>>>> What are the pros and cons you see of adding a new @schemVersion
>>>> rather than adding "css2.1" and "css3" as values to the existing
>>>> @scheme?
>>>
>>> There are two cons of adding e.g. "css2.1":
>>>
>>> 1. We'd have to keep adjusting the allowed value list every time a new
>>> version of CSS begins the route to recommendation.
>>>
>>> 2. Other style languages may be added at some point, and they may have
>>> their own versions, so it makes sense to separate the versions from the
>>> family names.
>>>
>>> Cons for the new attribute are that it's a new attribute, and we have
>>> hundreds of them already. :-)
>>>
>>> Cheers,
>>> Martin
>>>
>>>>>> <styleDefDecl scheme="css"> is perhaps not granular enough, in that
>>>>>> it does not allow specification of the version of CSS in use. Would
>>>>>> it be a good idea to add @schemeVersion so that users could specify
>>>>>> e.g. CSS 2.1 or 3?
>>>>>
>>>>> +1
>>>>>
>>>>>> While looking at this, I noticed that<styleDefDecl> and<rendition>
>>>>>> both have @scheme attributes, locally defined, which have the same
>>>>>> legal values. Shouldn't this be an attribute class? (I have a vague
>>>>>> memory of a discussion about this before, but I can't find any
>>>>>> trace of it; forgive me if this has already been dealt with.)
>>>>>
>>>>> I don't recall this, but it seems appropriate to me.
>>>
>>
>>
>


-- 
Dr James Cummings, James.Cummings at it.ox.ac.uk
Academic IT Services, University of Oxford