> On 17 Jul 2014, at 6:48 pm, Shane Stephens <shans@google.com> wrote: > > As I've already noted in this thread, I made a mistake here. Please don't hold that against me, these transform things are tricky :) > > On a happier note, this is evidence to support both our arguments! A. Creating a simplification for an inherently complex thing increases the risk of confusion and misunderstanding. B. Transforms are hard and that even people who should understand what they are doing are prone to making mistakes. To be clear, I'd definitely like to simplify transforms (and, in general, I think CSS has to be a candidate for one of the most confusingly complicated things in this industry - congratulations us!). I would just like to see if we can do it in a way that won't lure people into thinking they have a complete understanding and then punish them as soon as they push against the boundaries. Every feature has an up-front learning cost. Reducing that cost is a noble goal, as long as we're sure the deferred cost isn't too much more than the saving. Drive-by comment: there isn't a good fallback behaviour for these suggested new properties :( Dean
Received on Thursday, 17 July 2014 11:24:11 UTC