[whatwg] Cryptographically strong random numbers

Brendan Eich brendan at mozilla.com
Wed Feb 16 12:00:37 PST 2011


On Feb 16, 2011, at 11:31 AM, Mark S. Miller wrote:

> This issue seems to be the only significant remaining controversy here, so more words settling it more decisively would be welcome. Thanks.


Not to drag this out, but the "randomness quality" issue (or non-issue, now) is not the only remaining one. Allen raised the question of whether typed arrays should be used at all when Array might do. This was in reply to your suggestion that the "sooner not later" simple and usable API depend on the forthcoming binary_data strawman.

I'm not sure what the browser vendors who don't yet support typed arrays think. Purely out of expediency and "I'm all right, Jack" Firefox boosting, typed arrays are fine and have some advantages over plain old Array, as Adam has laid out.

But ideally, we should hash this out with Microsoft people weighing in here on es-discuss (I'm told they're not allowed to participate on whatwg lists).

I miss the Opera presence on TC39 but I know a few still read and post. Would be great to hear from them soon too.

/be


More information about the es-discuss mailing list