[00:13:17] <Krinkle>	 phedenskog: woah, very nice article
[00:13:23] <Krinkle>	 I like the writing style
[00:13:34] <Krinkle>	 and reference to https://en.wikipedia.org/wiki/Nagle's_algorithm
[00:14:13] <Krinkle>	 Nagle's algo is meant to improve performance, but it's now often known as the "Nagle delay" (or at least Aaron calls it that)
[00:20:47] <Krinkle>	 Chrome and Firefox disable Nagle on HTTP(S) conns it seems, (using TCP_NODELAY) - ref https://bugzilla.mozilla.org/show_bug.cgi?id=542401
[00:22:21] <Krinkle>	 WebKit I'm not sure, I don't find any reference on their bug tracker in the past 10 years that does anything with NODELAY or Nagle in either direction (except to disable it in WebRTC, but not more generally)
[00:22:40] <Krinkle>	 maybe it was never on, but probably not since chromium forking from webkit since disable dit
[00:22:46] <Krinkle>	 or... actually, they forked after they fixed it
[00:22:52] <Krinkle>	 so I guess webkit got it at the time
[00:22:53] <Krinkle>	 yay :)
[07:31:13] <phedenskog>	 Just browsed it yesterday, gonna read it today. 
[07:43:51] <phedenskog>	 Ok seems Moto G4/G5 is stileld valid. Wonder though if we should change to 4g instead of 3g testing. Today we mostly use 3g.
[10:56:39] <phedenskog>	 "Enabled Paint Timing" https://webkit.org/blog/11577/release-notes-for-safari-technology-preview-122/ :)
[20:51:58] <Krinkle>	 AaronSchulz: tentatively assinging this one as interrupt, seems a regression https://phabricator.wikimedia.org/T277056
[20:52:05] <Krinkle>	 ( gilles  )
[22:04:05] <AaronSchulz>	 Krinkle: I already have a patch for that
[22:05:47] <AaronSchulz>	 Krinkle: I dumb fix is to just remove the type hint though
[23:22:13] <AaronSchulz>	 *a dumb fix