- From: François REMY <francois.remy.dev@outlook.com>
- Date: Sat, 15 Mar 2014 18:23:37 +0100
- To: "Dirk Schulze" <dschulze@adobe.com>, "L. David Baron" <dbaron@dbaron.org>
- Cc: "Chris Eppstein" <chris@eppsteins.net>, "Tab Atkins Jr." <jackalmage@gmail.com>, "www-style list" <www-style@w3.org>
> I think Tab’s argumentation to use underscores > because customized media queries might or might > not end up with underscored names is very weak. Consider this another way: with "var-" we have no good opportunity to create an omnipresent concept of user-defined identifier in CSS because a lot of usages of those identifier will not be variables - something we didn't consider when [css-variables] started because custom properties was the only source of user-definenes envisaged at the time. With an underscore, we can: so why wouldn't we go in that direction? > -var- was a compromise and I do believe we should > not change this at this point but publish CR ASAP > if not done yet. As far as I know, to the notable exception of Tab, no one ever expressed positive agreement to -var- while other proposals did receive endorsments; so I wouldn't call this a compromise but rather a choice made at a time a choice had to be made. But now that the only person that ever defended this choice publicly backtracked, and made a (new) counter-proposal that received a surprisingly good welcome, I wouldn't mind revisiting this -- especially given Tab has the green light from the currently active implementor. Almost all the other things that have been "taken out of the pandora box" in this thread are, I believe, things that will not happen in L1 anyway (but may be added to a future revision of the spec without breaking compatibility; think: not forcing use of "var(...)" by allowing just "..."; allow to reference the specified value of any property and not just custom ones; ...)). This thread of course had mentions to them, because it's important to make a choice that leaves as many doors as possible open; that absolutely doesn't mean those things will be included in L1 (they will not) nor does it garantee they will ever enter in any working draft; it just means that some people wanted them to be considered in the choice. I think it's a sane approach towards making a good choice.
Received on Saturday, 15 March 2014 17:24:03 UTC