Re: [Yaml-core] Special Keys revisited

[Brian I. <in...@tt...>]

On 09/08/02 13:03 -0400, Clark C . Evans wrote:
> On Fri, Aug 09, 2002 at 09:22:30AM +0000, sh...@zi... wrote:
> | >   (#): Comment
> | >   (=): &0 Migration value
> | 
> | I don't think either of these belongs in YAML. 
> 
> I was considering taking it out of the spec altogether, but
> replacing this section with a pointer to a (stuff) registry
> where we control in an IANA like manner a set of special 
> values, like (true) and (false) in a uniform manner.
> 
> The question is, what does a parser do when it runs into
> a (thingy) that it doesn't recognize.  (on) => !bool 1
> and such, so (thingy) isn't a type family as much as its
> an escape hatch for stuff which doesn't fit into our basic
> recognition system.

If we are going to support special keys, then I think they belong in the
spec. Simply putting them in another document only muddies their purpose.

That said, I think we could probably just do away with comment keys. I just
don't think they offer much (except complexity) to a data serialization
language.

SATTP, Brian