Re: [xmile] recent updates

[Bobby Powers <bobbypowers@gmail.com>]

Hi Will,

Thanks for the update.  I have replies inline, below.

> 1.  ARRAYS - non-apply-to-tall
>
> * We both like the syntax recently proposed on the list.  Karim is adding
> this to the word doc.

Great!

> 3. MODULES.  We had discussion about <of> and <to>.  Right now this is in
> the standard, but Karim is polling people on the committee (specifically
> Bob, who hasnt spoken up yet) if this might make sense to remove.

I think the best place for discussion about the specification is on
the mailing list rather than 1-on-1 conversations off-list, but I look
forward to hearing Bob's point of view one way or the other.

> 4. MODEL.URL vs INCLUDE.RESOURCE
>
> * We wish to leave a url type property for <model> and <module> as the
> include functionality does not completely replace this.  Specifically, the
> model.url property allows the inclusion of submodels without a name.

What is the use case for a non-named submodel?  I don't really
understand how that can usefully happen.

>
> * For consistency, we are changing the "url" property across the board to be
> "resource" instead of url.  The difference is that a resource can be
>    -- relative file path
>    -- absolute file path
>    -- URL
>
> For includes the URL is an unchanging resource which may be cached, but this
> is not necessarily try for URLs on other tags.
>
> * On a related note the <image> tag now takes an optional parameter
> "resource" which allows the image to load a file or url with the same rules.
> And the "sound" attribute is now a resource not a URL.

sounds good, I don't feel strongly about this one way or another.

> 6.  MODELS WITHOUT DIAGRAMS
>
> * I have added an explicit comment that models without a diagram should be
> simulatable.  (i.e. software which now requires a diagram -- like iThink --
> should be able to load a model without a diagram).

Great!

> 7.  DASHES vs UNDERSCORES
>
> * We looked at elements such as <style> which mostly have attributes with
> dashes (the CSS properties) but also have some attributes with underscores.
> I'd proposed making all attributes dashes for consistency.  After
> discussion, we want to now leave this as is, due to the last minute nature
> of this proposal.  (But it's possible we could reopen this issue in the
> Fall).

I strongly feel the time to change small, trivial things like _ vs -
is now, before many people have implemented the specification.  Anyone
wishing to have compatibility with 'v1' xmile models and some future,
better 'v2' will effectively have to support both dashes and
underscores.  If we think this change is a good idea (and I do, and
several other TC members do), we should really change this now.



yours,
Bobby