Andy Allan

@petzlux there's a hack you can use, by writing XML inline within carto, e.g. ``` [...] text-size: 24; text-placement-type: list; text-name: "[name]"; [...] ``` See also: - https://github.com/mapbox/carto/issues/196#issuecomment-9192028 - https://github.com/mapbox/carto/blob/74fa914c8c96732648c9f7ab7f4f76c84a6f55ce/lib/carto/tree/definition.js#L143

I'm using .mvt tiles rather than geojson tiles - but otherwise the UX sounds about right.

Gah! Yes, I meant instances :-) I've updated the report to save further confusion.

@tmcw - sorry, it's not clear whether you're saying there's a workaround (by rearranging things), or that carto is behaving as expected. Do you agree there's a bug?

> I'm closing this as won't fix. That's a shame. > although it may not be obvious at first sight I think that's quite an understatement! The whole point of...

> a satisfactory solution without changing the way how instances work My suggestion would be to track the index of the instance rules, ignoring if the particular attribute declaration is...

I've run into the same issue. `git bisect` shows that it's 4d4abb2 that introduced the switching. Simple testcase that I've been using: ``` #countries { ::outline { line-color: #85c5d3; line-width:...

I'd prefer it to not throw, and to let you set the option twice. The cascading rules of CSS should be clear, that defining the same value for the second...

2b: also prevent flapping. I think since we're volunteers we'd trade some uptime for fewer headaches, rather than seeking the perfect 100%-uptime solutions. Is now the time to ask for...

> I know @gravitystorm has some thoughts on this Yes, I definitely have strong thoughts on this! And you are right, the first step in the long-term plan is to...