01:17:17 maybe some day such websites will autodetect a slow connection properly and we can use 10base2 as a workaround :-P 01:18:02 Gmail got more annoying. It used to remember the "Simple HTML version" option, nowadays it doesn't, so you have to be quick enough to click on the link... 08:12:26 o/ 13:31:48 sup 13:32:32 how is everyone 15:08:22 https://worldofmatthew.com/img/88x31/anythingbut.webp 15:08:33 how ironic: an Anti-Chrome button... on a Google formagt 15:08:35 --format 15:13:27 As bad as Google dictating "standards" is, at least they're actually open royalty-free codecs. 15:13:39 It's better than Apple trying to stuff the Web full of HEIC and HEVC. 15:14:07 it's still a Google "standard", tho 15:14:44 the only problem with HEIC/HEVC are the MPEG-LA patent cartel, which is irrelevant outside 'murica and/or the EU 15:15:53 (the MPEG-LA *still* believes they have rights over any media compression/encoding format ever made, even those NOT created by its members!) 15:18:57 remind me, is MPEG-LA part of SCO? 15:21:19 dunno 15:21:28 I guess their lawyers play golf on the same courses 15:26:40 I am ok with png but webp is something no one needs. 15:28:17 I still find a quite ridiculous idea to repurpose video codecs and containers for delivering still pictures 15:29:08 I don't mind borrowing specific algorithms and ideas, but lifting codecs wholesale to deliver still images is not only overkill, it's also a waste of resources 15:29:45 what worries me is how some of these recently introduced formats require much more processing power, at least for encoding, but I think some also for decoding 15:30:09 the idea here is "you can always bundle silicon for that on your device" 15:30:24 AKA Apple's Sekret Sauce™ 15:30:51 which makes sense for new designs, except on cost reasons 15:31:06 and of course you end giving the finger to anyone using "old" computers and phones 15:31:29 and of course you're also banking on the idea the format is now the killer format and that it won't be replaced 15:31:58 [insert xkcd here] 15:32:28 https://xkcd.com/927/ if you insist~ 15:32:49 but yeah, the push against not-new-enough hardware seems to be more pervasive these days 15:33:21 * njsg reads the title text and wonders how long until nano-USB shows up 15:33:30 maybe they should have called USB C "nano USB"? 15:33:47 Part of the reason the ARMacs are speedy is not due to its highly efficient ARM cores, but because Apple has filled up their silicon with dedicated coprocessors, DSPs and HW codecs for a lot of stuff 15:34:04 otherwise USB is losing to SIM card and its form factors! 15:34:10 which again makes sense for portables and cellphones, not for ordinary PCs 15:34:28 njsg: didn't you got the memo? The future is "virtual USB", of course" 15:34:33 app-defined USB ports! 15:34:48 physical SIMs are for boomers! 15:35:02 the newest iPhone is CDMA, er, eSIM-only in USA 15:35:11 tomman they are speedy because ram is on die. Bad move. 15:36:03 frg_Away: actually it's on package, but that's a move that make sense for cellphones and ultraportables (and hey, cellphones have been doing PoP since... well, forever?) 15:36:18 for full-fat laptops and desktops? Not so much 15:36:53 of course people justify shaving a few nanoseconds on latency as the reason of giving up memory expansions of any kind 15:37:12 Socketed CPUs are already a thing of the past on laptops, sadly 15:37:24 (Intel last did those on Haswell, IIRC) 15:38:51 and intel processors and chipsets had also been showing up a lot in memory-limited laptops, I think 15:39:32 although it was fun to find one which actually had the memory CS line wired, even if all existing documentation said it was limited to the lower amount of memory... 15:40:19 "chipset" on Intel land these days basically means "USB, SATA, HD Audio and LPC interfaces" 15:40:43 GMCH (memory and graphics) have been on-package since Nehalem, and on-die since Sandy Bridge 15:41:23 and dunno if Atoms fully went the SoC way 15:41:32 ah, ok, this is why I was thinking it was actually the processor, ok 15:42:11 the newer one then had the limitation involving the processor, meanwhile I do have an older one which has a chipset limitation 15:42:28 on AMD is basically the same, if not earlier (memory controllers have been on-die since Hammer/K8) 15:48:30 hello 15:48:39 I have an Error with page loading 15:48:48 name and shame! 15:48:52 what site is today? 15:48:56 it's about Web Host manager script 15:49:05 WHM 15:49:23 it show only few options....script dosn't work properly 15:49:24 it's one of those web hosting control panels? 15:49:28 yes 15:49:40 what errors do give out the web console? 15:52:24 a lot of error 15:52:29 post them 15:52:49 customElements are not defined 15:53:05 oh joy, Google WebComponents™ 15:53:52 for that one you may be able to workaround that with https://github.com/martok/palefill 15:54:34 but you will need to edit the addon to add a rule for each affected site (which involves unpacking the XPI and editing the ruleset) 15:55:05 any other error? (Most of them should be fallout from the customElements one, but there might be more Chromeisms present) 15:55:42 https://github.com/martok/palefill/issues/48 yay, it got fixed~! 17:39:50 Palefill dev just pushed version 1.22 of his addon yesterday, now it works again out of the box on SM, and brings yet more fixes for the latest Chromeisms at Giggityhub 18:26:07 FF's "Library" window is resizable, "drag able", but look what happens if you do it, heh. 18:26:09 https://postimg.cc/xNPYdYkb 19:15:32 impossible, that's not pure white!! 19:16:23 now honestly, that should work, shouldn't it? what happened, one more element between the library UI and the window, some resize code failed? 19:20:46 i'm thinking the UI is one of them dumb boxes - which is the entirely of FF these days, like any of the about: "dialogs", simply overlaid ontop of a window. so the window itself resizes, but the dialog is fixed. 19:21:06 cya 19:22:04 oh, & the FF window itself /just happened/ to be behind the Library window (in my screenshot) 19:45:56 tomman you are right. I thought I saw a photo with an explanation when the M1 pro was released with indicate on die ram but seems only the memory controllers are on die. But the tight integration is what makes then so fast. Memory bandwith is incredible compared to standard setups. But also means they will be all dead in a few years when ram cells or one of the components develops a fault. 19:48:37 therube they should have kept xul for chrome and extensions needing ui access but doing a "me too is Chrome" is all what was wanted.