You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
<span> not specified as allowed descendant of <coordinates> #215
Span elements - as defined in HTML - can take phrasing content, so I guess MapML would have to define its own <span> that has a restricted content model for coordinate positions? That doesn't seem right, shouldn't something new be used instead of <span>?
<span> is semantically neutral, AFAICT, so it would not be ideal to have to add a map-specific version of it, although we will have to update our schema and spec to include it, which might get messy, tbd.
So <span> doesn't seem to be the same thing in HTML and MapML. SVG has <tspan>, why did they not (re)use and define <span>? Perhaps because its content model is different, and they wanted to avoid confusing authors/implementers?
I think this suggests we need a new <cspan> (coordinate span) element for use in <coordinate>.
Activity
Malvoz commentedon Sep 11, 2021
Going a bit off topic:
Span elements - as defined in HTML - can take phrasing content, so I guess MapML would have to define its own
<span>that has a restricted content model for coordinate positions? That doesn't seem right, shouldn't something new be used instead of<span>?prushforth commentedon Sep 12, 2021
<span>is semantically neutral, AFAICT, so it would not be ideal to have to add a map-specific version of it, although we will have to update our schema and spec to include it, which might get messy, tbd.Malvoz commentedon Sep 12, 2021
My concern is that we're reusing HTML elements although we expect them to behave or be used differently.
For example,
<iframe>is phrasing content, where:is valid HTML.
But:
is invalid MapML.
So
<span>doesn't seem to be the same thing in HTML and MapML. SVG has<tspan>, why did they not (re)use and define<span>? Perhaps because its content model is different, and they wanted to avoid confusing authors/implementers?I think this suggests we need a new
<cspan>(coordinate span) element for use in<coordinate>.