| | 1 | Attendee: ericc, hober, dhyatt, Zoltan, asterns, abucur |
| | 2 | |
| | 3 | == Status of Spec == |
| | 4 | |
| | 5 | * Firefox implements the full specification because CSS WG has stripped down the parts Firefox didn't implement. |
| | 6 | * New syntax uses "--" prefix to define a new CSS variable and uses "var()" function to reference a variable. - There was a very long thread on CSS WG. |
| | 7 | * Custom properties are all inheritable - this may pose an implementation challenge; we probably need a chain of custom property bundles. |
| | 8 | * Custom properties aren't exposed in computed style. |
| | 9 | |
| | 10 | |
| | 11 | == Historical Implementations == |
| | 12 | There are two historical implementations: the one implemented by hyatt and someone at Google. |
| | 13 | The Google implementation had many architectural problems. |
| | 14 | |
| | 15 | The hyatt implementation was taken out of the trunk because CSS WG could not agree on the exact syntax. |
| | 16 | It also supported changing the value of a CSS variable in scripts. |
| | 17 | var() function parser could be pulled in from the hyatt's implementation but the patch may not apply well due to it being so old. |
| | 18 | |
| | 19 | |
| | 20 | == Concerns == |
| | 21 | |
| | 22 | * Websites that don't use variables shouldn't slow down. |
| | 23 | * Breaking cycle - appears to be spec'ed but invalidating the value with cycles might be challenging in the implementation. |
| | 24 | |
| | 25 | |
| | 26 | == Who should implement? == |
| | 27 | * Hyatt says this would be fun to implement. |
| | 28 | * Alan says he might be able to find people who can help. |
