Twitter | Search | |
Boris
1,821
Tweets
0
Following
869
Followers
Tweets
Boris Jan 28
Yep, I saw that PR, and of course the issue it's filed based on. Haven't had a chance to look at the PR carefully yet, and probably will not until next week, unfortunately.
Reply Retweet Like
Boris Jan 27
There are some issues with JSON serialization, etc, but those might be solvable with toJSON methods, maybe. Anyway, as an interim step to a future in which there's a real Array here in some way (but one that prevents "illegal mutations" somehow) this seems totally doable.
Reply Retweet Like
Boris Jan 27
If we _really_ cared we could even put Array.prototype on the proto chain (and in fact Gecko did that for some things for a while), but that depends on how we want instanceof to behave.
Reply Retweet Like
Boris Jan 27
`push` should Just Work, I think, as long as the indexed property setter steps for the type allow index == length. This general thing has been proposed in the past but shot down for reasons that are eluding me right now, probably because it's not "real array" enough...
Reply Retweet Like
Boris Jan 27
That is quite doable. IDL already specs that certain Array.prototype methods are copied to the proto of things with indexed props. We could copy more. There are some issue with `concat`, but I think @@isConcatSpreadable lets us solve those.
Reply Retweet Like
Boris Jan 27
Of course that's not actually an issue in the semantics Chrome implements for FrozenArray, which don't match the IDL spec for FrozenArray. No one else implements FrozenArray at all. And saying "the API shape webidl recommends" while implementing something else is odd too. ;)
Reply Retweet Like
Boris Jan 27
And yes, the constructable stylesheets spec author _also_ misused it (see "hard to properly use"). I filed on this a month and a half ago; crickets so far...
Reply Retweet Like
Boris Jan 27
We really do just need to solve the issues we've failed to solve these last 10+ years, and FrozenArray is not that solutions.
Reply Retweet Like
Boris Jan 27
Fwiw, FrozenArray as currently specced in Web IDL has all _sorts_ of issues. It's hard to propely use as a spec author, no one actually implements the currently specced semantics (because they are moderately insane), and it's not great for API consumers.
Reply Retweet Like
Boris Jan 14
Replying to @KWierso
8 char caps? Who are these people???? I mean, highly successful trading companies, but still....
Reply Retweet Like
Boris Jan 14
Replying to @mike_conley
Yes.... I thought that was a _non_-goal with passwords? ;)
Reply Retweet Like
Boris Jan 14
Replying to @KWierso
That presupposes they put my password, in raw form, in an SQL database. Which I _really_ hope they are not doing, but given everything else, who knows...
Reply Retweet Like
Boris Jan 14
Password: (1) Must contain only letters and numbers (2) Must be 8-15 characters long. Why do banks have these dumb rules? :( And yes, this one doesn't even allow '-', much less spaces.
Reply Retweet Like
Boris Jan 13
Replying to @TedMielczarek
That was my reaction too!
Reply Retweet Like
Boris Jan 11
It's still there, constant-false. Sites depend on it for browser-sniffing and if we remove it start relying on Blink/WebKit bugs... :(
Reply Retweet Like
Boris Jan 9
Replying to @andreasgal
It's been a long road. ;)
Reply Retweet Like
Boris Jan 9
Replying to @shaver
That was sort of my reaction too, and I knew people were actively working on it!
Reply Retweet Like
Boris Jan 9
Subject: [Bug 1448967] Remove enablePrivilege ... What |Removed |Added ---------------------------------------------------------------------------- Resolution|--- |FIXED Paging old Gecko hands for celebrations.
Reply Retweet Like
Boris Dec 23
Replying to @aryxnet
That's fine, because I never type that hostname. I search via my browser's search bar, and that's set to a different search engine anyway. ;)
Reply Retweet Like
Boris Dec 23
Turns it's a PITA to use modern distributed version control on a keyboard with a 'g' key that doesn't register presses half the time. So close to `alias it="git"` and `alias h="hg"`... Also very glad my next laptop will not have a butterfly keyboard.
Reply Retweet Like