[tei-council] wait there's (a few) more

[Laurent Romary]

Alles klar. So it's a no-brainer.
Laurent

Le 14 sept. 10 à 16:02, James Cummings a écrit :

>>> 3058674 Add model.graphicLike to content of glyph  ...
>> Can you update the ticket so that I understand the discussion  
>> between  you and James?
>
> I think the original request is complete clear (but maybe I'm  
> biased!)... it says:
>
> ===
> We replaced model.graphicLike in char and glyph with figure, and  
> then I later made a good argument on another ticket for adding it  
> back to char, which was accepted and duly done. However, char and  
> glyph work in very similar ways and if it is justifiable for char to  
> have model.graphicLike then glyph should have it as well.
>
> Recommendation: restore model.graphicLike to glyph, keeping figure  
> as well, as is the case in char.
> ===
>
> To be extra clear:  Once upon a time <char> and <glyph> both allowed  
> <graphic> as a child.  Council decided that it should also have  
> <figure>, in implementing this we carelessly removed  
> model.graphicLike from their content model.  I noticed this and said  
> "Hey, why can't my <char> have any <graphic/> elements in it any  
> more", put in a bug report, and model.graphicLike was added back to  
> the content model of <char>.  I neglected to complain about <glyph>  
> because I don't really use it that much.  But I was about to answer  
> "Why not use a <graphic/> inside your <glyph>?" to a TEI-L posting  
> when, upon checking, I found that I could not do so. When I asked to  
> restore model.graphicLike to <char> I should have also said "and  
> <glyph> as well". Mea culpa.  In all honesty I think this is just a  
> corrigible bug rather than something council should be bothered  
> with, because to argue that it shouldn't be added back to <glyph>  
> you have to make an argument for why <glyph>s can't have <graphic>  
> and <char>s can, or you have to argue against it being present in  
> <char>'s content model and say that it needs to be removed from both.
>
> On the ticket I also briefly mention that lots of the constituent  
> members of the content model of <char> and <glyph> are identical and  
> wonder why they are not provided by a class. However, that is really  
> a separate bug.
>
> Does this help?
>
> -James
>
> -- 
> Dr James Cummings, Research Technologies Service
> OUCS, University of Oxford

Laurent Romary
INRIA & HUB-IDSL
laurent.romary at inria.fr