| --- Log opened Sat Jan 13 00:00:24 2018 |
| 00:55 | | celticminstrel [celticminst@Nightstar-m9434e.dsl.bell.ca] has joined #code |
| 00:55 | | mode/#code [+o celticminstrel] by ChanServ |
| 01:02 | | Derakon[AFK] is now known as Derakon |
| 01:42 | | Kindamoody is now known as Kindamoody[zZz] |
| 02:06 | | Degi [Degi@Nightstar-othrpl.dyn.telefonica.de] has quit [Connection reset by peer] |
| 03:12 | <&[R]> | Maybe busybox? |
| 04:01 | | * McMartin watches his Game Boy unit test fail, notices he was adapting the wrong function. |
| 04:01 | <&McMartin> | You know what spoils you real fast? being able to load more than one byte at a time |
| 04:02 | <&McMartin> | (A thing the Z80 can do and the LR35902 cannot) |
| 04:14 | | McMartin [mcmartin@Nightstar-rpcdbf.sntcca.sbcglobal.net] has quit [[NS] Quit: upgrades] |
| 04:27 | | McMartin [mcmartin@Nightstar-rpcdbf.sntcca.sbcglobal.net] has joined #code |
| 04:27 | | mode/#code [+ao McMartin McMartin] by ChanServ |
| 04:47 | < Mahal> | re "blaming MS when an OS upgrade breaks your software" is, AIUI, largely because you are 'required' to upgrade your OS for all kinds of reasons |
| 04:47 | < Mahal> | those reasons don't seem to be as important in non-MS contexts? |
| 04:48 | <&McMartin> | It seems like it's more "just recompile the universe so your stuff all works" |
| 04:48 | <&McMartin> | Which, uh, is a bad idea as anyone here who as attempted to run a Gentoo system can tell you |
| 04:49 | <&McMartin> | But which is a popular attitude to take~ |
| 04:49 | <&McMartin> | Though I suppose the hostility to systemd might be taken as an example of this |
| 05:01 | | Derakon is now known as Derakon[AFK] |
| 05:01 | | celticminstrel is now known as celmin|sleep |
| 05:01 | | Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 05:19 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 05:19 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 05:23 | <&[R]> | Because everything and their dogs depend on it now? |
| 05:23 | <&McMartin> | "Because when it was introduced, people blamed systemd, not software incompatible with systemd" |
| 05:25 | <&[R]> | To be fair, blaming alpha version software makes more sense than blaming established software. |
| 05:29 | | Kindamoody[zZz] [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection reset by peer] |
| 05:30 | | Kindamoody[zZz] [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code |
| 05:30 | | mode/#code [+o Kindamoody[zZz]] by ChanServ |
| 06:42 | | Vornlicious [Vorn@Nightstar-1iqvki.sub-174-211-0.myvzw.com] has joined #code |
| 06:43 | | macdjord is now known as macdjord|slep |
| 06:45 | | Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 08:55 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 09:01 | | Kindamoody[zZz] [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection reset by peer] |
| 09:02 | | Kindamoody|afk [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code |
| 09:02 | | mode/#code [+o Kindamoody|afk] by ChanServ |
| 10:43 | | Degi [Degi@Nightstar-othrpl.dyn.telefonica.de] has joined #code |
| 12:29 | <&ToxicFrog> | On windows, compatibility with older software is a lot more important because "just recompile the old software" is usually not an option even in theory |
| 12:29 | <&ToxicFrog> | So if it can't run as is, you're hosed. |
| 12:34 | | Degi [Degi@Nightstar-othrpl.dyn.telefonica.de] has quit [Connection closed] |
| 12:44 | < Vornlicious> | I still can't figure out Apple's strategy on this. |
| 13:14 | | ktemkin [sid249770@Nightstar-ks9.9ff.184.192.IP] has quit [Operation timed out] |
| 13:14 | | ktemkin [sid249770@Nightstar-h2b233.irccloud.com] has joined #code |
| 15:10 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 15:10 | | mode/#code [+qo Vornicus Vornicus] by ChanServ |
| 15:15 | | JustBob [justbob@ServerAdministrator.Nightstar.Net] has joined #code |
| 15:15 | | celticminstrel [celticminst@Nightstar-m9434e.dsl.bell.ca] has joined #code |
| 15:15 | | Netsplit over, joins: McMartin, Tamber, VirusJTG |
| 15:15 | | mode/#code [+o macdjord|slep] by ChanServ |
| 15:15 | | mode/#code [+o Tamber] by ChanServ |
| 15:15 | | Netsplit over, joins: @PinkFreud, ktemkin, macdjord|slep |
| 15:16 | | ktemkin is now known as NSGuest40900 |
| 15:16 | | JustBob is now known as NSGuest1617 |
| 15:16 | | Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code |
| 15:16 | | mode/#code [+o Kindamoody|autojoin] by ChanServ |
| 15:17 | | Orthia [quassel@Nightstar-ksqup0.co.uk] has joined #code |
| 15:17 | | mode/#code [+o Orthia] by ChanServ |
| 15:20 | | Alek [Alek@Nightstar-7or629.il.comcast.net] has joined #code |
| 15:20 | | mode/#code [+o Alek] by ChanServ |
| 15:20 | | gnolam [lenin@Nightstar-ego6cb.cust.bahnhof.se] has joined #code |
| 15:20 | | mode/#code [+o gnolam] by ChanServ |
| 15:31 | | Vornlicious [Vorn@Nightstar-1iqvki.sub-174-211-0.myvzw.com] has quit [[NS] Quit: Bye] |
| 15:31 | | Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 16:34 | | gnolam_ [lenin@Nightstar-ego6cb.cust.bahnhof.se] has joined #code |
| 16:37 | | gnolam [lenin@Nightstar-ego6cb.cust.bahnhof.se] has quit [Ping timeout: 121 seconds] |
| 17:08 | | Kindamoody|autojoin is now known as Kindamoody |
| 17:26 | | Kindamoody [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has quit [Connection closed] |
| 17:33 | | Kindamoody|autojoin [Kindamoody@Nightstar-eubaqc.tbcn.telia.com] has joined #code |
| 17:33 | | mode/#code [+o Kindamoody|autojoin] by ChanServ |
| 17:34 | | Kindamoody|autojoin is now known as Kindamoody |
| 18:09 | <~Vornicus> | so Uplink got a gui mod, uplinkOS, which tries to reduce some of the pain points; it does a decent job of that, but there's already a list of a decent number of improvements I can think of. |
| 18:09 | <~Vornicus> | Why does this game do this to me |
| 18:49 | <&[R]> | browser_with_arg=`echo "$browser" | sed 's %s '"$1" ` |
| 18:49 | <&[R]> | ... |
| 18:49 | <&[R]> | Gods, why didn't I realize it sooner? |
| 18:50 | <&[R]> | browser_with_arg=`printf "$browser" "$1"` |
| 18:57 | | Derakon[AFK] is now known as Derakon |
| 18:58 | <&[R]> | And after reading the manual: printf -v browser_with_arg "$browser" "$1" |
| 19:16 | <&[R]> | MISSFILE |
| 19:17 | <&[R]> | Ignore that |
| 19:18 | <@Alek> | that was a pretty fun webcomic. |
| 19:18 | <&[R]> | It's still going |
| 19:44 | <@Alek> | I know. But I haven't been keeping up with it. :( |
| 19:53 | | ErikMesoy [Bruker@Nightstar-hq72t5.customer.cdi.no] has quit [Ping timeout: 121 seconds] |
| 20:06 | | Vornlicious [Vorn@Nightstar-16082j.sub-174-210-3.myvzw.com] has joined #code |
| 20:09 | | Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 20:13 | | Degi [Degi@Nightstar-othrpl.dyn.telefonica.de] has joined #code |
| 20:18 | | Vornlicious [Vorn@Nightstar-16082j.sub-174-210-3.myvzw.com] has quit [Ping timeout: 121 seconds] |
| 20:21 | | Vorntastic [Vorn@Nightstar-sjo6le.sub-174-210-15.myvzw.com] has joined #code |
| 20:37 | | Vorntastic [Vorn@Nightstar-sjo6le.sub-174-210-15.myvzw.com] has quit [[NS] Quit: Bye] |
| 20:37 | | Vorntastic [Vorn@Nightstar-1l3nul.res.rr.com] has joined #code |
| 20:46 | | NSGuest1617 is now known as JustBob |
| 20:47 | | mode/#code [+o JustBob] by ChanServ |
| 21:07 | <&[R]> | Anyone know a way to copy files over ssh that doesn't involve tar | ssh tar? |
| 21:07 | <&[R]> | scp is absolutely braindead when it comes to symlinks |
| 21:09 | < celticminstrel> | I'm just wondering how you'd expect it to work for symlinks. |
| 21:10 | <@TheWatcher> | [R]: rsync? |
| 21:10 | < celticminstrel> | Symlinks refer to a path, right? Either it's a relative path and it still works when copied to another computer, or it's an absolute path and it breaks. |
| 21:11 | <@TheWatcher> | In general with pretty much any network file copy, rsync is your friend |
| 21:20 | <&[R]> | I can't guaranteed rsync being on both ends |
| 21:21 | <&[R]> | Half tempted to just make a wrapper script for tar |ssh tar |
| 21:25 | <@TheWatcher> | ... you just broke my brain |
| 21:34 | <&[R]> | I have a number of different hosts, not all of them have working package management systems in place |
| 21:35 | <&[R]> | I simply cannot get rsync onto all of the hosts, so a non-rsync solution is preferable, given rsync needs to be on both ends to work. |
| 21:37 | <@TheWatcher> | scp a static linked rsync binary first?~ |
| 21:40 | <&[R]> | I could do that for some of the hosts, but not all (due to archetectures primarily). I could do an inventory I guess, see which ones are capable of hosting rsync |
| 21:48 | | Vornicus [Vorn@Nightstar-1l3nul.res.rr.com] has quit [Ping timeout: 121 seconds] |
| 21:57 | < [> | why not use the tar method? |
| 21:58 | < [> | also annoying how every single question and answer i found on google doesn't say anything about scp in the answers, just "use rsync" or "use ssh+tar" |
| 22:01 | < McMartin> | And nobody says "use sftp" though apparently sftp is a hilarious abomination and if you *actually* wanted that you wanted ftps |
| 22:02 | <&[R]> | Oh right, sftp |
| 22:03 | < McMartin> | (I think the issue with sftp is that it's not actually ftp, while ftps actually is ftp, just over a properly encrypted connection? I think I've used each one once in like twenty years) |
| 22:03 | <&[R]> | sftp isn't too bad |
| 22:04 | <&[R]> | I think most FTP clients (the good ones) support both |
| 22:04 | <&[R]> | Filezilla definately does. I think lftp does as well. |
| 22:04 | <&[R]> | Don't think there are any others that matter. |
| 22:04 | < McMartin> | Definitely so for ftps |
| 22:04 | < McMartin> | Funny how "ftp" is not an acceptable ftp client~ |
| 22:05 | <&[R]> | It's fine if you only want to move one file at a time. |
| 22:05 | < [> | anyone know who scott hanselman is? he posted on his blog a couple years ago complaining about "file corruption" over ftp when he was messing w/ a raspberry pi for fun |
| 22:05 | < McMartin> | Not offhand |
| 22:06 | < McMartin> | I hope he didn't mix up ascii vs binary mode |
| 22:06 | < [> | should have been titled "ms dev learns about text conversion in ftp" |
| 22:06 | < [> | indeed |
| 22:06 | < McMartin> | ms devs should be the ones who know in advance |
| 22:06 | < McMartin> | Since they have to do text conversion in C and Python and UNIX guys don't and blame everyone else for being idiots when the UNIX code assumes there's no difference between fopen ("foo", "r") and fopen ("foo", "rb") |
| 22:07 | < [> | his title is funnier i guess, 13 hours, lol https://www.hanselman.com/blog/13HoursDebuggingASegmentationFaultInNETCoreOnRaspberryPiAndTheSolutionWas.aspx |
| 22:07 | <&[R]> | Last few ftpd's I looked at default to binary mode and will ignore the first few attempts to set text mode. |
| 22:07 | < McMartin> | That has definitely not been my experience with the ftpds I still occasionally have to contact |
| 22:08 | | * [R] was looking at FOSS ones at least |
| 22:10 | < McMartin> | One of my favorite things related to text conversion is that the PNG has an 0x0a in its magic number specifically so that you can instantly identify PNGs that have been corrupted by a textmode transmission link |
| 22:10 | < McMartin> | *the PNG file format |
| 22:11 | <&[R]> | lol |
| 22:15 | < [> | it's kinda disappointing that scp just don't have an option to control symlink behavior |
| 22:15 | < [> | doesn't* |
| 22:16 | <&[R]> | Yeah |
| 22:16 | < [> | it's simple enough that it must have been a concious decision to not include it |
| 22:16 | < [> | for whatever reason |
| 22:16 | <&[R]> | By default it'll treat them like whatever they point to... even if that's an infinite loop. |
| 22:17 | <&[R]> | Had a cheap-o backup fuck up because of that |
| 22:17 | < [> | :( |
| 22:17 | < [> | that's what you're fixing now? |
| 22:17 | <&[R]> | No |
| 22:17 | <&[R]> | That was several years ago, when Arch wasn't shit |
| 22:17 | < [> | lol |
| 22:18 | <&[R]> | So 2007-2009 |
| 22:18 | < McMartin> | I'm having trouble understanding that kernel-libc mutual dependency issue you described before |
| 22:18 | < [> | the distro environment seems to be losing steam |
| 22:18 | < McMartin> | It seems like it should either hit everybody or would hit nobody unless you bent over backwards to engineer it |
| 22:18 | < McMartin> | The former seems false, and the latter implies they had more energy than they clearly did~ |
| 22:18 | <&[R]> | I had it occur on multiple systems |
| 22:19 | <&[R]> | So I don't know |
| 22:19 | < McMartin> | Multiple arch systems or multiple distros? |
| 22:19 | < McMartin> | (e.g., is this a thing kernel packagers need to watch out for generally or is Pacman simply not up to the task) |
| 22:19 | <&[R]> | I was just entering my depression, so I didn't really have the time to properly debug it either. |
| 22:19 | <&[R]> | Multiple arch systems |
| 22:19 | < McMartin> | k |
| 22:20 | <&[R]> | I know it's related to a configure option when compiling glibc |
| 22:20 | <&[R]> | You tell it what's the minimum kernel version to support |
| 22:20 | < McMartin> | So kernels aren't statically linked? |
| 22:20 | < McMartin> | (modulo, uh, modules) |
| 22:20 | <&[R]> | Then for the kernel wanting a newer glibc, I forget the specific error, but I think I can dig it up from my logs, but ATM going back that far is a massive pain. |
| 22:21 | < McMartin> | Yeah, this is mostly me noticing that I know very little about how modern Linux systems are constructed at the file level |
| 22:21 | < McMartin> | Don't bother straining for answers |
| 22:21 | <&[R]> | So I couldn't update glibc, since my kernels were too old. Couldn't update kernels because glibc was too old. Everything else worked for a while until bash complained that glibc was too old... D: |
| 22:22 | <&[R]> | I also had 3-4 hosts die to hardware failures over the course of 5-6 months too during this. |
| 22:22 | <&[R]> | So basically massive shitpile of issues I had no time to fix properly |
| 22:23 | < McMartin> | Mmm. OK, actually, I think I know how dnf and apt would deal with that on the systems I've used. |
| 22:23 | <&[R]> | And a job I hated, that ate up a ton of time (IE: quite a bit of OT, and a long commute) |
| 22:23 | < McMartin> | (Lay down the kernel but keep the old runnable one, refuse to boot that one until the glibc update was also laid down) |
| 22:23 | <&[R]> | Other distros generally make it so you have multiple kernels installed at once. |
| 22:23 | <&[R]> | Which would've made is /massively/ simpler to fix. |
| 22:24 | < McMartin> | Yep |
| 22:24 | <&[R]> | Also one paticularity about Arch's kernel package... the post-install script generated the initrd. |
| 22:25 | | macdjord|slep [macdjord@Nightstar-a1fj2k.mc.videotron.ca] has quit [[NS] Quit: Wenn ist das Nunstück git und Slotermeyer? Ja! Beiherhund das Oder die Flipperwaldt gersput!] |
| 22:28 | <&[R]> | Years later, I found a news post on the Arch site that I think was meant to solve the issue. Steps were: force install kernel telling pacman to /not/ bug you about glibc (default pacman.conf would block if glibc had an update, ask you to update glibc first and then only update glibc) |
| 22:28 | <&[R]> | Then update glibc. |
| 22:28 | <&[R]> | Then do a full system update. |
| 22:30 | <&[R]> | Funny thing is, I actually have some Arch installs from that era that still work (on portable drives). Not sure why it didn't occur to me to use those to try and fix things. |
| 22:41 | <@himi> | McMartin the kernel is statically linked, but the kernel ABI has changed over time in ways that require the compiler/libc to be aware of |
| 22:41 | <@himi> | So the kernel doesn't care about glibc at all, but glibc cares about the kernel version potentially quite a lot |
| 22:42 | <&[R]> | Yeah, like I said, it's a configure flag |
| 22:42 | | macdjord [macdjord@Nightstar-a1fj2k.mc.videotron.ca] has joined #code |
| 22:42 | | mode/#code [+o macdjord] by ChanServ |
| 22:42 | <&[R]> | http://www.linuxfromscratch.org/lfs/view/stable/chapter05/glibc.html <-- look for --enable-kernel |
| 22:43 | <@himi> | I thought it had become much less of an issue in recent decades, but I may be thinking of the libc ABI |
| 22:44 | <&[R]> | At the time, I had read something that led me to believe they were making it so that each glibc would require the absolute latest kernel at the time of the glibc compile (which pissed me off, and I suspect is wrong now) |
| 22:44 | <@himi> | I've never run Arch, so I'm not familiar with it outside it's rather awesome wiki |
| 22:45 | <&[R]> | It was really great back then. Now they're doing weird shit like using systemd and having weird sticky permissions on /usr and /var. |
| 22:46 | <&[R]> | From the PKGBUILD page, it looks like they went back on the "everything belongs in /usr" plan. |
| 22:46 | <@himi> | Using systemd stopped being weird some time ago - hate it as much as you like, it's solidly heading towards being the default |
| 22:48 | <&[R]> | Thankfully there are a number of saner alternatives (and the infection is /slow/) |
| 23:28 | | macdjord is now known as macdjord|dance |
| 23:36 | < Degi> | What's wrong with systemd |
| 23:40 | <&[R]> | 1) It's written by the bad Pottering (that's honestly my number 1 reason) 2) It needlessly wants to change /everything/ 3) The devs close any legitimate security issue with WONTFIX 4) The devs are responsible for the udevd 30 second wait issue a few years back 5) There are many, much simpler solutions that work better 6) It makes certain things non-deterministic (like unmounting) 7) The only thing good people have to say about it is "haven't had any issues here" |
| 23:42 | <&[R]> | 8) There was a point where there were weekly major security issues |
| 23:56 | <&[R]> | And my issues with Pottering are due to my issues with his other software NetworkManager and PulseAudio |
| 23:58 | <&[R]> | https://www.agwa.name/blog/post/how_to_crash_systemd_in_one_tweet <-- stuff like that |
| --- Log closed Sun Jan 14 00:00:26 2018 |