Re: Error recovery

On Sun, Nov 18, 2012 at 5:05 AM, Liam R E Quin <liam@w3.org> wrote:

> On Sat, 2012-11-17 at 13:27 -0700, Uche Ogbuji wrote:
>
> > > (a) <e a="x"/>, or
> > >
> > > (b) <e a="x/">
>
> > I prefer (a) since it's the likeliest interpretation of the author's
> > intention.
>
> <p><a href=/socks/>more interesting articles</a></p>
>

I see this example as the best argument for (b). Other minor considerations
in favour for (b)

- if the parser defaults to (b), the user can easily get the (a)
interpretation by inserting a space <e a=x />; but if the parser defaults
to (a), the user has to use quotes to get (b)

- it's slightly simpler: assuming you are allowing / in unquoted attribute
values, it's simplest to say that unquoted attribute values are terminated
by whitespace or >.

Since MicroXML is/was aimed at Web usage, I think (b) the better choice,
>

I agree that representing HTML-based markup languages is an important use
case for MicroXML. However, it seems a bit of a leap to argue for a
particular interpretation for all MicroXML documents based on one HTML
example. There are HTML examples that go the other way, such as:

<img src=socks.jpeg/>

*or* build-in to the parser a list of empty HTML elements and use (a)
> for those and (b) for the rest.


I don't like this at all. I think it's important for MicroXML to handle
empty element syntax in an XML-like way rather than an HTML-like way. That
means that <x></x> and <x/> should always equivalent for all x; empty
element syntax can be used for all element types not just for element types
that are always empty. If <img src=socks.jpg/> means <img
src="socks.jpg"/>, then <script src=socks.js/> should mean <script
src="socks.js"/>.

Overall, (a) seems the best choice to me.

James