10:30:39 Any news regarding the rust-1.56 front? I still would like to see some sm release compiling with it. 10:36:20 JamminUnit: I think WG9s's builds are already built with rust 1.56 10:37:18 njsg: so where can I grab the patches for seamonkey-2.53.10b1? :D 10:44:33 444444444444444444444444444444444444444444444444444444444444\\\\\\\\\\Iiiii'; 10:44:35 U& 10:45:31 good morning Bellla! 11:03:36 https://www.neelc.org/posts/chase-how-not-to-design-a-website/ more user agent sniffing gone Real Wrong 11:03:44 and of course, it has to be a goddamned bank 11:06:39 JamminUnit http://logbot.buc.me/seamonkey/20211113#c7483 14:32:09 Howdy. Got a couple of bug reports, probably noticed before but I'd like to mention them... 14:33:08 1. Following a direct link to a youtube channel doesn't work. eg. https://www.youtube.com/c/NASASpaceflightVideos Using latest version of Seamonkey, note that Noscript is also in the mix. Can anyone verify this? 14:34:45 2. Wayback machine link on http://archive.org/ disappears behind other page elements, no way to enter a link. Verification also requested on this. 14:35:29 for 1) there have been reports on newsgroups and here of some youtube pages showing up blank, that might be related to user-agent sniffing; login state was also leading to different outcomes? 14:35:52 It's always possible some videos are missing because Scripp News took them down, of course 14:37:17 and noscript - how's noscript configured regarding youtube and google? youtube does require javascript, I think 14:37:35 for 2) I've been just assembling archive URLs directly, I don't think I've tried to enter these in the front page, let me see 14:37:56 "redirecting you to a lite version of archive.org", fancy 14:38:08 which is... JSON? 14:39:04 ah no, it shows a link to a json document, but that's not the actual redirect 14:41:28 okay, they've changed this recently, it's not caused by seamonkey changes, the home page is definitely different 14:41:34 njsg: I've disabled all blocking on noscript with the same result, though noscript's presence might be doing something. 14:42:22 Huh. Opening that channel page in a private window resulted in the page coming up, so it is login related. 14:43:50 And yes, it shows up as blank except in the top elements. 14:44:57 until there's a solution for archive.org, http://web.archive.org/address-goes-here should be enough if there's a stored snapshot 14:45:37 saving should work with /save/address but that hasn't been working properly for months (years?) now, sometimes works, sometimes works but fails to load the saved copy, sometimes fails 14:46:09 frg_Away: thank you. It worked! :D 14:48:50 If I sneak up to the channel page, searching for it and clicking the link in search, it works. Hit it directly, and blank. 14:49:45 And clicking "send feedback" to report the problem gives me a 403 error, lol. 16:00:02 JamminUnit: are you a Vob Marley fan? https://www.youtube.com/watch?v=oFRbZJXjWIA 16:02:42 WG9s: not really, but that music is very nice :) 16:39:48 seems archive.org uses now some adavanced css features whatever they are. 17:04:19 i suppose could be same failing clang-13 new warnings 17:04:31 oops worng channel 21:24:32 frg_away: I went one step forward with the rustc cross-compiling problem. I was confused by the fact that rustc lists the needed target i586-unknown-linux-gnu (for CPUs without SSE) when asked but a component for this target had to be installed with rustup. Having done this, I get different binaries when compiling a simple file like helloworld.rs. Transferred the my old hardware, only the... 21:24:35 ...binary build with rustc --target i586-unknown-linux-gnu works. The other shows "Ungültiger Maschinenbefehl". 21:26:13 and with LC_ALL=C before the command? :-) 21:27:58 Having found out this, I introduced ac_add_option --target=i586-unknown-linux-gnu in the mozconfig file and no my host is about to build. 21:29:14 njsg: "Illegal instruction" is the evil formula you wanted to read, didn't you? :-) 21:29:16 21:44:32 mlh85386: I comile using i686-pc-linux-gnu 21:52:27 dont you lov it when someone askes for help and then disconects! 21:53:48 Oh you nare back I do my 32-bit builds by using the target i686-pc-linux-gnu 21:55:59 mlh85386: for what kind of processor was this? 21:56:31 njsg: For an AMD Duron 21:58:19 wg9s: It's one of the mysteries of rust, when to use i686-pc-linux-gnu and when i686-unknown-linux-gnu. We will see whether this leads to new problems 21:58:21 my processor is amd also 21:59:10 of course you might need to do a rustup target install i686-pc-linux-gnu 22:00:30 perhbaps you should do a rustup target list|to see what targets are avalable for your rust version 22:00:47 wg9s: the i686-pc-linux-gnu was provided by default, (no cross-compiling) 22:01:11 so I think that is what you shuld try 22:01:17 it works for me 22:02:25 WG9s much older CPU. Won't work. 22:02:39 i686-pc-linux-gnu is probably the default on a GNU/Linux system on IA-32 22:02:40 wg9s: with i686-pc-linux-gnu, the SSE2 set of instructions will be used and the binary does not work 22:02:40 i do my 32-bit builds whidh seem to work for everyone who has tried on 64-bit os with 64 bit tools so specify that as the target and all seems well 22:03:50 rust documentation says i586-unknown-linux-gnu is for CPUs without SSE2 22:03:56 OIC your processor does not have sse2 instructions? 22:04:41 wg9s: This forces me to build Seamonkey 22:06:23 * njsg would expect i586 to be "for pentium or newer" :-P 22:06:25 so you are corect i586-unknown-linux-gnu should be corret for that. hmm 22:06:40 wonder if this is a SeaMonkey issue or a Mozilla issue 22:06:53 and then... why not i586-pc-linux-gnu? 22:08:15 njsg: perhaps because that target does not exist? 22:08:19 mlh85386: and yes, thanks for the "Illegal instruction". I need to learn German. 22:09:14 these are the 32-bit targets that exist 22:09:16 i586-pc-windows-msvc 22:09:17 i586-unknown-linux-gnu 22:09:19 i586-unknown-linux-musl 22:09:21 i686-linux-android 22:09:22 i686-pc-windows-gnu 22:09:24 i686-pc-windows-msvc 22:09:25 i686-unknown-freebsd 22:09:27 i686-unknown-linux-gnu (installed) 22:09:28 i686-unknown-linux-musl 22:10:39 wg9s: It is a mozilla issue. But the mozilla documentation is poor. Fortunatly linuxfromscratch made me hope being able to compile Seamonkey it on my own. They delivered as well a mozconfig file to start with 22:12:01 njsg: I only teach bavarian german :-) 22:13:44 mlh85386: so when i ask is this a mozilla i mean a mozilla issue versus a SeaMonkey issue. can you build the latest firefox and get it to run? 22:14:18 if not then file a mozilla bugzilla bug on the issue 22:15:34 wg9s: I did not try. The firefox from Debian Buster worked out of the box. But crashes too often. 22:19:47 well then try filing a bug agianst the Debian Buster version about the crashing issue would be my ingestion. rather than trying to roll your own build. 22:23:20 wg9s: I do not like working with Firefox. It serves just as substitute until a Seamonkey-Version works with the current Debian release. 22:28:15 It's getting late. The build will take several hours. Good night. 22:29:19 WG9s SSE2 is mandatory since around 53 or so. 22:32:00 mlh85386: so seems frg_Away is saying we no longer support non SSE2 PUas 22:32:04 cpus 22:32:37 beacuse moazilla/Firfos no longer does 22:32:44 Yes. 100% unsupported in 2.53.x and already for Windows in 2.49.x 22:33:18 mlh85386: so for SeaMonkey might need ot stay on 2.49 22:36:37 wg9s: I know. 2.49 was the last version that worked. 2.53.1 was the first I built on my own last year. I only hope that it's possible to compile 2.53.9.1 as well now. 22:39:08 I will see tomorrow and I'm going to inform you.