On Wednesday 25 August 2004 08:06 pm, Brian Ingerson wrote:
> Second let me damn it. SSYN could be renamed to SSAY. "SSYM Simply Ain't
> YAML". I'm not sure the level to which folks grok the differences
> between YAML and SSYM, but they are not at all the same. People need to
> know this before they get too excited about the shiny new toy.
> For starters the information models are just different. I can see
> compatibility between any two of XML, YAML and SSYM as being in the same
> boat. None is a subset or superset of the others. Compatibility is a mess.
> I had a chat with Tim Parkin about this on #yaml. My thoughts came down to:
> "Yes, SSYN would be easier to implement. No, it isn't something I'll find
> particularly useful after it is implemented."
> A close and highly respected friend remarked: "12:48 < xxxx > it's
> fascinating watching people reinventing s-expressions over and
> again. :)"
> In the end, you may get SSYM quicker, but you won't have YAML. Just be
> aware of it. On the other hand, if you never get YAML, SSYM is infinitely
> more valuable.
Okay, that's seems a little "rich". I suppose that's the proper perspective of
someone in the belly of the Yaml, but... What I mean to say is, that I think,
you've limited yourself by predicating a need for compatible information
models, there are other forms of cross-polleneation.
> Let me state a few things for the record.
> Firstly, Jason has been around this list since week 2. He's no stranger
> to YAML. He did one of the very first YAML implementations in C# in
> 2001. (Doesn't that almost predate C#? ;) Jason, please feel free to
> discuss your ideas here. You are always welcome.
> Oren Clark and I think that YAML's information model, "A typed graph of
> mappings, sequences and scalars", is it's most important feature. The
> syntax is nice too, but it is less important.
> This is mostly important because YAML is primarily intended to be a
> cross programming language *serialization* language for Perl, Python and
> Ruby. (and PHP, Java and others). The YAML info model is a very closely
> matched superset of those three languages' internal models. SSYN is not.
> Yes YAML is complex, but in my mind it is necessary complexity. What
> would we remove? Everything was put there for a reason. Everything was
> discussed in length and bounced off of every use case in a public forum.
> For over 3 years!
So I wonder... I don't know, I'm going out on limb here, but I wonder why the
spec isn't 1.0 final yet. --I'm sure there are reasons, but maybe too many?
Anyway, I don't think it hurts to talk about stuff even if the bottom line is
TAY! TAY! --That Ain't Yaml! THAT AIN'T YAML! And I don't think you have to
be a member of this list more than three minutes to do so. Know what I mean?
> So about YAML implementations...
Between BabelTest and Syck looks like implementation is well in hand! Its just
a matter of time now.... (Ha Ha! The troops are massing in the in the
> To close, I just want to thank everyone in the YAML community for making
> something as cool as YAML.
I dropped XML like a lead brick when I discoved YAML, and I use YAML all over
the place! So I want to personally thank you, too. Thanks Brian!!!!!!!!!!!!