* Ovid <[EMAIL PROTECTED]> [2008-08-18T06:50:00]
> JSON is fairly well implemented and new implementations are trivial. This is
> not true for YAML. Trying to define a minimum standard of YAML for extended
> TAP is a quagmire. With JSON, we can punt and just point to a fairly
> well-established JSON spec.
I would absolutely love to use JSON instead of YAML. There are no good,
well-documented implementations of YAML for Perl, period. YAML::Tiny has less
of a reliable spec than YAML.
I've been doing a lot of work lately with both YAML and JSON, and in every case
dealing with JSON is very easy and dealing with YAML is very frustrating. I
love the idea of YAML, but in practice, beyond writing simple config files, it
ends up really driving me nuts.
It is slightly less readable, and I think that bears noting:
YAML:
neck: 17
chest: 38
feet: { left: 9 1/2, right: 10 }
JSON
{
"neck": 17,
"chest": 38,
"feet": { "left": "9 1/2", right: "10" }
}
...but I don't care, because I trust that the presentation layer can do
something nice, if I come to hate just viewing the TAP stream. It's still
readable enough.
JSON makes numbers/strings clearly distinct, which is very nice when in
non-Perl lang. It also makes booleans clear. Seriously, I will gladly fight
this fight. I thought it was lost in Oslo, but if Ovid is willing to bring it
up again, bring it on.
> To be honest, this does not block the IETF proposal at all because we'd
> merely be describing the areas of agreement which need to be reached.
> However, we do need to discuss this and going from "I has an implementation"
> to "I has a spec" is rather like going from "I has Perl" to "I has PPI" if
> we stick to YAML.
This also really helps us keep the spec simple, as we can say "...emitted as
JSON data [RFC 4627]."
--
rjbs
