|
Chris Lilley
@svgeesus
|
3. pro |
|
Yes. Though I also wish implementers read the spec instead of just making all the existing tests pass. And wish more tests linked to specific assertions.
|
||
|
|
||
|
Boris
@bz_moz
|
3. pro |
|
Oh, for sure. In my ideal world, a spec writer writes tests, then implementors write an implementation based on the spec. Any tests they fail, that's probably a spec issue in the sense of the spec being unclear or wrong, if the implementor is competent.
|
||
|
|
||
|
Marcos Càceres
@marcosc
|
3. pro |
|
Why we should enforce a “no test? no merge!” rule across all specs via policy. If it was a W3C spec, let me know and I’ll add the template to prevent that in whatever WG is culpable.
|
||
|
|
||
|
Boris
@bz_moz
|
3. pro |
|
It was a W3C spec, yes. Rookie editor, which is part of what's going on. I'll message you the details offline.
|
||
|
|
||