[03:07] <lotuspsychje_> good morning
[03:53] <oerheks> on this day in 2008, Sun Microsystems announced it was acquiring MySQL for $1B
[03:53] <oerheks> 15 months later ..
[08:50] <lotuspsychje> troll gang hour
[08:50] <lotuspsychje> lets ask random unsolvable stuff for 5min
[08:50] <lotuspsychje> funny
[13:43] <leftyfb> what is this all about? "<Hanumaan> [Kiran Kumar Telukunta, IndiaYouth] hurt_locker,"
[13:44] <pragmaticenigma> dunno.. starting to get annoying
[14:21] <pragmaticenigma> great move... let's support 16.04 longer with someone that's goign to find more broken packages to complain about
[14:22] <leftyfb> pragmaticenigma: 16.04 is till supported and for another year. We should still help them
[14:22] <lotuspsychje> !eol
[14:23] <lotuspsychje> april 2021
[14:23] <leftyfb> that's over a year away
[14:23] <leftyfb> I still run 16.04 on my servers and support thousands of robots running it as well
[14:24] <pragmaticenigma> Yeah... let's help someone limp along a broken install that they just did. When all this could be fixed and we won't hear about the next thing they find not to their liking because the version is 6 years old
[14:24] <lotuspsychje> maybe we will get lucky and when xenial goes eol, support will grow again
[14:25] <leftyfb> pragmaticenigma: you're only basing that on a segfault from 1 package. Nothing else has shown they have a broken install
[14:26] <pragmaticenigma> leftyfb: no... I'm basing it off of "I just installed 16.04" which I don't understand why anyone would do unless they had a specific purpose
[14:26] <leftyfb> oh. .. hm, I missed that
[14:26] <leftyfb> well that's silly
[14:26] <pragmaticenigma> server, I get it, developer targetting specific development software... I get it... installing it watch movies... huh?
[14:27] <leftyfb> pragmaticenigma: you're right then
[14:35] <lotuspsychje> tomreyn: xenial will also have 3y ESM extend right on your !eol edit
[14:35] <leftyfb> hey TJ- , any idea how to fix /dev/random taking a long time to get going within initramfs?
[14:36] <TJ-> leftyfb: generate more entropy :)
[14:36] <TJ-> or use urandom
[14:36] <leftyfb> TJ-: how? Without touching the keyboard? (no mouse)
[14:36] <leftyfb> TJ-: I tried telling curl to use urandom but it didn't seem to help
[14:39] <pragmaticenigma> Doesn't some of the entropy get built out of the network traffic?
[14:39] <leftyfb> you would think
[14:40] <leftyfb> I tried writing a couple gig to /dev/null but it didn't really help
[14:41] <pragmaticenigma> I found the perfect article... stuck behind a paywall... thanks redhat
[14:41] <TJ-> what is trying to read /dev/random ?
[14:42] <TJ-> if you mean curl itself is hanging because it is trying to derive random data for negotiating a TLS ephermal key, then use "curl --random-file /dev/unrandom ..."
[14:42] <pragmaticenigma> this article recommends haveged to help stir the pot: https://www.techrepublic.com/article/how-to-add-more-entropy-to-improve-cryptographic-randomness-on-linux/
[14:43] <leftyfb> TJ-: As soon as I boot, I need to read from an https source, so SSL, which needs entropy
[14:43] <TJ-> or even "/dev/urandom" :)
[14:43] <TJ-> see "man curl" and "--random-file"
[14:43] <leftyfb> TJ-: I already tried that, didn't make a difference
[14:43] <TJ-> leftyfb: hmmm what is curl linked to? maybe libssl is the cause?
[14:43] <leftyfb> pragmaticenigma: already tried that as well
[14:44] <pragmaticenigma> suppose it doesn't start early enough in the boot process?
[14:44] <leftyfb> TJ-: I have curl and openssl pulled in via hooks
[14:45] <leftyfb> pragmaticenigma: I'm running all this in init-bottom , which I think is as close as I can get
[14:45] <TJ-> leftyfb: hmmm... I *think* there is an entropy stirring tool but can't recall what it is
[14:45] <leftyfb> I tried haveged, but also didn't make much of a difference
[14:45] <leftyfb> and rngd
[14:46] <TJ-> what arch is this on?
[14:46] <leftyfb> ubuntu 18.04 amd64
[14:47] <TJ-> leftyfb: have you seen 5.1.5 and lins from that? https://www.debian.org/releases/stable/amd64/release-notes/ch-information.html
[14:48] <leftyfb> there's no systemd in initramfs
[14:48] <TJ-> I was referring to "If you read this after upgrading a remote system to buster, ping the system on the network continuously as this adds entropy to the randomness pool and the system will eventually be reachable by ssh again. "
[14:49] <TJ->  and the lin to wiki etc
[14:49] <TJ-> grrr, s/lin/link/
[15:09] <ice9> how much performance loss between using the discrete card as primary GPU and using PRIME?
[15:09] <lotuspsychje> ice9: wich graphics card are we talking about exactly?
[15:17] <ice9> lotuspsychje, nvidia
[15:17] <ice9> and prop driver
[15:18] <lotuspsychje> thats the brand yes, do you know the exact chipset of your card?
[15:20] <ice9> lotuspsychje, 940mx
[15:20] <ice9> geforce
[15:22] <lotuspsychje> ice9: so your card is optimus capable and can be switched between powersaving mode and performance mode, what do you want to know exactly?
[15:23] <ice9> lotuspsychje, what do you mean by powersaving mode, (using the igpu)?
[15:23] <lotuspsychje> ice9: powersaving uses intel and performance mode nvidia
[15:24] <ice9> lotuspsychje, great, that's works fine on windows, but on linux you have to either use nvidia the primary GPU for ALL output or use PRIME to switch for it when needed
[15:24] <ice9> my question is that, is there any performance loss when using switching with prime?
[15:25] <lotuspsychje> ice9: switching what?
[15:27] <lotuspsychje> ice9: how about you run nvidia-settings and tell wich mode you have active?
[15:34] <lotuspsychje> ice9: 'if' your question means switching from powersaving to performance and vice versa, obviously its a big gain or loss