I’m archiving old tweets to my site and realizing that #smooshgate really captured my attention in 2018
I vaguely remember seeing something about clarifying this goal in some tc39 notes but haven’t been able to find it.
@littledan @ljharb does #TC39 have any published guidelines on the meaning and motivation for “don’t break the web”? Then tc39 could point to that rather than repeatedly discussing on proposals like global & #smooshgate
@chiccychiccy The interesting part is why/how not to break old parts of the web. The dramatic part is bashing old code and determining what to name things.
@chiccychiccy Haha, it’s a complicated web developer drama with it’s own hashtag. #smooshgate medium.com/@sarunint/what…
I suspect that the vast majority of MooTools sites aren’t using flatten
@jankrems @_jayphelps So my understanding is that Array defers to any existing methods, but the non-enumerable spec method doesn’t get copied to MooTools’s Elements prototype.
I don’t remember seeing the moment.js global proposal conflict getting as much of a reaction as MooTools with Array.prototype.flatten. Global seems more significant to me. github.com/tc39/proposal-…
@ryanflorence Interesting to see that 2.6% of all website use MooTools. For the top 10,000 sites that drops to 0.6% w3techs.com/technologies/h… w3techs.com/technologies/c…
@ryanflorence Going forward, modifying the prototype of builtins pre-standardization should be understood as a risk by developers and shouldn’t hold up standardization. However, I can’t imagine past tools could be reasonably held to this expectation.
@ryanflorence I want to see backwards compatibility upheld but my impression is that this expectation has already been broken for MooTools sites. I could be wrong though.
@ryanflorence How many standards have broken MooTools? The 1.6.0 release mentions something about Array.from() compatibility problems.
Also look for posts tagged #smooshgate on other websites registered with octothorp.es.