Fix Webidl respec errors

[prushforth]

Closes #212

[Malvoz]

I believe we'll want to remove:

https://github.com/Maps4HTML/MapML/blob/9368ab7a95807bf64525941c695e95c12ea866ae/spec/index.html#L1886-L1891

There's also mention of legends in:
https://github.com/Maps4HTML/MapML/blob/9368ab7a95807bf64525941c695e95c12ea866ae/spec/index.html#L3341

[prushforth]

I believe we'll want to remove:

https://github.com/Maps4HTML/MapML/blob/9368ab7a95807bf64525941c695e95c12ea866ae/spec/index.html#L1886-L1891

I don't believe so, but there's certainly more work that we should do in improving how to define and link to, display and interact with legends.

Specifically, the legend link relation is used in the

<link rel="legend" href="https://cwfis.cfs.nrcan.gc.ca/geoserver/public/ows?service=WMS&request=GetLegendGraphic&format=image/png&width=20&height=20&layer=fdr_current"/>

construct, which is used to create an <a href="...">Layer label</a> in the layer control:

which was about the best I could do with the legend abstraction at the time, working as I was with the WMS specification.

There's also mention of legends in:
https://github.com/Maps4HTML/MapML/blob/9368ab7a95807bf64525941c695e95c12ea866ae/spec/index.html#L3341

I'm no longer convinced that legends are fully out of scope, since legend information could be important for map accessibility. I wrote that, I think I did, at least, while thinking about the relationship of maps to HTML documents at large, and not wanting to introduce more abstractions into this proposal than strictly necessary ("legend" being an abstraction). Do you think I should delete or amend the example of what's out of scope? That paragraph relates to non-interactive user agents such as printers, but of course legends play an important role in printed maps as well.

[Malvoz]

Okay, while I think legends being links (rel="legend") is somewhat problematic as described in #80 (comment) we can leave them as such for now.

[Malvoz]

There's a merge conflict, other than that LGTM!