2020-08-12T03:30:19 #kisslinux <muevoid> Is it safe to make a symbolic link between /lib/libc.so and /lib/libc.so.6?
2020-08-12T03:30:28 #kisslinux <muevoid> This application works if I do so
2020-08-12T03:30:37 #kisslinux <muevoid> But i'm not sure if it is safe to do so
2020-08-12T03:47:59 #kisslinux <muevoid> rrm
2020-08-12T04:30:13 #kisslinux <muevoid> Another question does the kiss package manager support making tars out of packages that are from git so they can be compiled offline
2020-08-12T16:25:13 #kisslinux <E5ten> dylanaraps: in k imo for the xsnprintf lines in the getline loop, for size you should pass sizeof(srcp) (or sizeof(desp) for the other one) instead of PATH_MAX, so if you change the size of those buffers you only need to change them in one place
2020-08-12T16:38:27 #kisslinux <mcpcpc_kirc> oh man... i think `kirc` is finally "usable", lol!
2020-08-12T16:39:17 #kisslinux <mcpcpc_kirc> and backspace works! apparently VERASE is not set by default?
2020-08-12T17:33:28 #kisslinux <aosync> What's kirc
2020-08-12T17:35:09 #kisslinux <mcpcpc_kirc> an irc client that i have been developing
2020-08-12T17:35:21 #kisslinux <mcpcpc_kirc> and E5ten has been fixing :P
2020-08-12T17:36:05 #kisslinux <mcpcpc_kirc> https://github.com/mcpcpc/kirc
2020-08-12T17:38:55 #kisslinux <mcpcpc[m]> still a ways to go before i can say it’s ready for “release”.
2020-08-12T18:49:33 #kisslinux <midfavila> Hey, everyone. I'm looking at switching to KISS from CRUX, and noticed it has an option to use the 9base coreutils as an alternative to busybox, which is pretty awesome; only problem is that the package manager seems to rely on busybox's implementation of diff in particular (or maybe GNU-like extensions in general?), as when replacing something like Busybox diff with 9base's diff, it fails to compare checksum
2020-08-12T18:49:33 #kisslinux <midfavila> n't stat -". Is this a known issue?
2020-08-12T18:53:37 #kisslinux <j`ey> is it that 9front doesnt support POSIX fully too?
2020-08-12T18:56:14 #kisslinux <midfavila> Potentially, but from what I understand it does. Admittedly part of why I wanted to use KISS is to experiment with alternative userspaces and the like, so I'm somewhat ignorant in that regard.
2020-08-12T18:59:42 #kisslinux <admicos> https://k1ss.org/guidestones says "All distribution tooling and shell code must be written in a portable way." with the only documented exception being "sed -i". The diff portability might be a bug
2020-08-12T18:59:59 #kisslinux <j`ey> surprising if diff doesnt support -
2020-08-12T19:00:20 #kisslinux <midfavila> *nod*
2020-08-12T19:00:27 #kisslinux <midfavila> I'll do a clean install just to be absolutely sure
2020-08-12T19:00:35 #kisslinux <j`ey> midfavila: doubt that's needed
2020-08-12T19:00:48 #kisslinux <j`ey> just try to use 9base's diff with those args
2020-08-12T19:01:00 #kisslinux <j`ey> something like https://github.com/kisslinux/kiss/blob/master/kiss#L474
2020-08-12T19:01:18 #kisslinux <midfavila> alright, one sec, gotta reinstall them
2020-08-12T19:02:27 #kisslinux <merakor> 9base isn't POSIX though, it is a replication of the plan9 tools. POSIX wasn't a thing back then
2020-08-12T19:03:32 #kisslinux <merakor> I am the maintainer of 9base, I didn't use its diff before, sorry about that
2020-08-12T19:05:24 #kisslinux <midfavila> Oh, it's not a problem at all! At the moment I'm just tinkering with KISS anyways, so this is good to learn about
2020-08-12T19:10:46 #kisslinux <midfavila> yeah, it definitely seems like it might be a 9base diff issue. I figured I'd pop a new install anyways, and after removing the busybox symlink and switching to 9base using `kiss a`, I'm getting the same error
2020-08-12T19:10:47 #kisslinux <midfavila> rpi
2020-08-12T19:10:51 #kisslinux <midfavila> rip, even.
2020-08-12T19:12:03 #kisslinux <j`ey> could maybe just not install 9base diff
2020-08-12T19:12:06 #kisslinux <j`ey> and keep busyboxs
2020-08-12T19:13:07 #kisslinux <midfavila> yeah, but that seems kind of redundant. if I have to I can just find a POSIX-compliant alternative or somethin'
2020-08-12T19:13:13 #kisslinux <midfavila> worst comes to worst I'll stick with bbox
2020-08-12T19:28:23 #kisslinux <micr0> i'm making a package to help bootstrap my kiss installation, called talyn-config. in the package directory, i have files/etc/{fstab,inittab}, files/home/micr0/{.gitignore,.profile}, etc. I am trying to install stuff with `install -D etc/fstab "$1/etc/fstab"` but getting "install: can't stat 'etc/fstab': No such file or directory". any suggestions?
2020-08-12T19:32:15 #kisslinux <micr0> Interesting, it seems kiss does not preserve the files/ layout. so you cannot have multiple files with the same name, or really any directory structure
2020-08-12T19:33:15 #kisslinux <micr0> in fact i wonder if i had files/dirA/file and files/dirB/file which one would get installed during build
2020-08-12T19:52:11 #kisslinux <dylanaraps> E5ten: Yeah. Will be doing that. Regardless... I did some benchmarks and the performance is identical unless /a lot/ of files need to be installed.
2020-08-12T19:52:31 #kisslinux <dylanaraps> (Python for example has 3800 files so a speedup is seen)
2020-08-12T19:52:54 #kisslinux <j`ey> how much of a speedup?
2020-08-12T19:53:16 #kisslinux <dylanaraps> From 3s down to 1s roughly
2020-08-12T19:54:08 #kisslinux <j`ey> was the goal to rewrite everything, or just -remove -install?
2020-08-12T19:54:21 #kisslinux <dylanaraps> Initially it was the whole thing.
2020-08-12T19:54:29 #kisslinux <dylanaraps> Then I decided to focus on install/remove.
2020-08-12T19:54:34 #kisslinux <dylanaraps> Now it's not /that/ worth it.
2020-08-12T19:54:38 #kisslinux <dylanaraps> I have other plans though.
2020-08-12T19:55:44 #kisslinux <dylanaraps> Installation/removal were my first candidates for a C rewrite as they are where things are typically slow.
2020-08-12T19:56:07 #kisslinux <dylanaraps> Conflict resolution (prior to install) is another place.
2020-08-12T19:56:59 #kisslinux <dylanaraps> Now... the package manager calls cp/rm/whatever _once per file_. If I were to bunch them together and call these commands less, we'd see near the same performance.
2020-08-12T19:57:04 #kisslinux <j`ey> 3s is still not bad, assuming that python is one of the bigger packages
2020-08-12T19:57:12 #kisslinux <j`ey> yah ^
2020-08-12T19:57:15 #kisslinux <dylanaraps> It's one of the bigger packages in terms of number of files.
2020-08-12T19:57:27 #kisslinux <j`ey> right, that's what I meant
2020-08-12T19:57:43 #kisslinux <dylanaraps> Speed is more based on disk speed here
2020-08-12T19:57:50 #kisslinux <j`ey> I have some rust code which built some basic kiss packages
2020-08-12T19:57:56 #kisslinux <dylanaraps> Shell is just slower because we call commands per-file
2020-08-12T19:58:56 #kisslinux <dylanaraps> You can see the function here: https://github.com/kisslinux/kiss/blob/master/kiss#L899-L947
2020-08-12T19:59:26 #kisslinux <j`ey> I can see the function, doesn't mean I will understand the shell code within :P
2020-08-12T19:59:37 #kisslinux <micr0> dylanaraps would this be considered a bug? https://gist.github.com/jedahan/7400afcd3d7a0ee44685de15e6b28455
2020-08-12T20:00:11 #kisslinux <dylanaraps> Nope
2020-08-12T20:00:16 #kisslinux <dylanaraps> Sources can overwrite sources
2020-08-12T20:00:41 #kisslinux <dylanaraps> This is a feature
2020-08-12T20:00:52 #kisslinux <micr0> is there a way to specify which file to install?
2020-08-12T20:01:24 #kisslinux <dylanaraps> Well. Each source is copied to the cache directory for build in the order they appear in the sources file.
2020-08-12T20:01:25 #kisslinux <micr0> lets say i want files/a/file to be installed to /home/a/file, and files/b/file to be installed to /home/b/file?
2020-08-12T20:02:02 #kisslinux <dylanaraps> Wait
2020-08-12T20:02:04 #kisslinux <micr0> its kinda the same surprise i have that checksums flatten the path
2020-08-12T20:02:09 #kisslinux <dylanaraps> I may be misunderstood.
2020-08-12T20:02:10 #kisslinux <j`ey> dylanaraps: what happens to files in /etc, how are they installed?
2020-08-12T20:02:27 #kisslinux <dylanaraps> j`ey: 3-way handshake similar to Arch's pacman.
2020-08-12T20:02:34 #kisslinux <dylanaraps> micro0: Care to elaborate a little?
2020-08-12T20:02:39 #kisslinux <j`ey> oh yeah I remember
2020-08-12T20:02:44 #kisslinux <j`ey> I just need to find that code
2020-08-12T20:03:08 #kisslinux <dylanaraps> Honestly, I'm not happy with /etc/ in general.
2020-08-12T20:03:15 #kisslinux <j`ey> pkg_etc
2020-08-12T20:03:51 #kisslinux <dylanaraps> The package manager must install these files but /not/ keep track of them. What then happens to updates to these files, etc? It's quite convoluted.
2020-08-12T20:04:11 #kisslinux <j`ey> what's wrong with the 3 way handshake style?
2020-08-12T20:05:22 #kisslinux <claudia02> hello
2020-08-12T20:05:41 #kisslinux <dylanaraps> I'd personally prefer that all config files be tracked by the package manager and installed somewhere outside /etc/. It's then up to the user to move them to /etc/ if they'd like to use one. Just an idea
2020-08-12T20:05:58 #kisslinux <dylanaraps> Dunno.
2020-08-12T20:06:01 #kisslinux <dylanaraps> This has flaws too.
2020-08-12T20:06:25 #kisslinux <dylanaraps> I barely touch /etc/ anyway. Mine is rather clean so I'm OK with things as they are.
2020-08-12T20:06:31 #kisslinux <micr0> There are two things going on here dylanaraps. The first is I have a package with files/a/file and files/b/file, and want to install files/a/file to /home/a/file and files/b/file to /home/b/file.
2020-08-12T20:06:31 #kisslinux <dylanaraps> (personally)
2020-08-12T20:06:41 #kisslinux <dylanaraps> micr0: I get it now.
2020-08-12T20:06:47 #kisslinux <dylanaraps> The package manager flattens the path.
2020-08-12T20:06:51 #kisslinux <micr0> yep
2020-08-12T20:06:53 #kisslinux <dylanaraps> Checksums are only for one file.
2020-08-12T20:07:00 #kisslinux <dylanaraps> The last in the list overwrites the first.
2020-08-12T20:07:07 #kisslinux <dylanaraps> Yeah?
2020-08-12T20:07:18 #kisslinux <micr0> the first fix i thought of would be a major breaking change but maybe theres a way to transition to it
2020-08-12T20:07:43 #kisslinux <micr0> the second issue is more an inconsistency, that kiss c outputs two checksums, even though the package manager flattens the path
2020-08-12T20:08:04 #kisslinux <dylanaraps> If that is the case, it may be an easy fix.
2020-08-12T20:08:40 #kisslinux <dylanaraps> (Package manager turns files/a/file into a/file, creates a and copies file there)
2020-08-12T20:09:07 #kisslinux <micr0> that would be a good improvement
2020-08-12T20:09:09 #kisslinux <dylanaraps> I still need to dive into the code and think about this more but that sounds OK at first glance.
2020-08-12T20:09:19 #kisslinux <micr0> personally, I think what would be the least magic is: use full paths everywhere
2020-08-12T20:09:32 #kisslinux <micr0> checksums, and in install, keep the whole files/a/file path
2020-08-12T20:10:01 #kisslinux <micr0> but it would be a breaking change, unless you wrote a 'first check full path, then flattened' at least for a little while
2020-08-12T20:10:14 #kisslinux <dylanaraps> Yup.
2020-08-12T20:10:52 #kisslinux <dylanaraps> It'd be rather complex for a while.
2020-08-12T20:11:09 #kisslinux <dylanaraps> Have you tried this:
2020-08-12T20:11:13 #kisslinux <dylanaraps> files/a/b a
2020-08-12T20:11:17 #kisslinux <dylanaraps> oops
2020-08-12T20:11:20 #kisslinux <dylanaraps> files/a/file a
2020-08-12T20:11:22 #kisslinux <dylanaraps> files/b/file b
2020-08-12T20:11:23 #kisslinux <dylanaraps> ?
2020-08-12T20:11:50 #kisslinux <dylanaraps> Actually. I'll stop being lazy and try it here. :P
2020-08-12T20:11:52 #kisslinux <j`ey> dylanaraps: I was just thinking about using some KISS_HOOKS to overwrite any changes I had to etc anyway
2020-08-12T20:12:28 #kisslinux <micr0> its funny, github gist has the same restriction as the current kiss package manager - i cannot have two files with the same name, and they are not allowed to have / in the name xD
2020-08-12T20:12:38 #kisslinux <dylanaraps> lol...
2020-08-12T20:14:00 #kisslinux <dylanaraps> Yeah. My suggestion does as described.
2020-08-12T20:15:02 #kisslinux <dylanaraps> The intention with sources is to deal with files. The second field in the format deals with destination directories.
2020-08-12T20:16:44 #kisslinux <micr0> wait you are able to specify two different sources with the same basename?
2020-08-12T20:16:49 #kisslinux <micr0> in build?
2020-08-12T20:16:50 #kisslinux <dylanaraps> Yes
2020-08-12T20:17:00 #kisslinux <micr0> hmm i cannot reproduce that
2020-08-12T20:17:09 #kisslinux <dylanaraps> What do you mean by build?
2020-08-12T20:17:32 #kisslinux <dylanaraps> <source> <dest> is the syntax with <dest> being created relative from the initial build directory.
2020-08-12T20:17:54 #kisslinux <dylanaraps> So 'files/a/file a' will copy the file to a directory called 'a' in the build directory.
2020-08-12T20:18:11 #kisslinux <dylanaraps> Otherwise the file itself is just copied to the initial build directory.
2020-08-12T20:18:38 #kisslinux <dylanaraps> By using two destination directories (in this case identical to the source location), you can have both files.
2020-08-12T20:18:47 #kisslinux <micr0> ohhh
2020-08-12T20:18:53 #kisslinux <micr0> lemme try
2020-08-12T20:19:02 #kisslinux <dylanaraps> I have it working here
2020-08-12T20:19:59 #kisslinux <dylanaraps> https://termbin.com/qom1
2020-08-12T20:20:47 #kisslinux <micr0> Ok so if my sources file looks like this, it works: http://ix.io/2u0C
2020-08-12T20:21:16 #kisslinux <micr0> I guess i would imagine the default be to just copy over the sources directory verbatim, but I can just update all my packages to do that manually
2020-08-12T20:22:07 #kisslinux <micr0> I also still think its a bit weird that checksums flatten things
2020-08-12T20:22:13 #kisslinux <micr0> but thats a separate thought
2020-08-12T20:22:19 #kisslinux <dylanaraps> Should just need to be 'a', 'b'.
2020-08-12T20:23:14 #kisslinux <dylanaraps> It doesn't /really/ matter. Ordering will always match ordering in sources, otherwise a checksum mismatch will occur.
2020-08-12T20:23:30 #kisslinux <dylanaraps> Has no effect on checksum verification
2020-08-12T20:26:03 #kisslinux <dylanaraps> Here's my full package: https://termbin.com/1q5p
2020-08-12T20:26:28 #kisslinux <micr0> Yeah I mean, it all works, I just need to think about it more to shake off why it feels either magical or inconsistent
2020-08-12T20:26:30 #kisslinux <dylanaraps> micr0: We could drop the file name from the checksums entirely.
2020-08-12T20:26:48 #kisslinux <dylanaraps> It's not actually needed.
2020-08-12T20:27:01 #kisslinux <dylanaraps> Handy for someone looking at it, sure!
2020-08-12T20:27:10 #kisslinux <dylanaraps> But then, it'd be handier to have the full path.
2020-08-12T20:27:15 #kisslinux <micr0> exactly
2020-08-12T20:27:28 #kisslinux <micr0> if its for humans, i'd prefer the full path to match the one in sources
2020-08-12T20:27:41 #kisslinux <dylanaraps> Though... this is the first talk of duplicates a year and a half into the distribution's life. lol
2020-08-12T20:27:41 #kisslinux <micr0> if its for machines, i'd prefer dropping it to really indicate 'this is for machines'
2020-08-12T20:27:47 #kisslinux <dylanaraps> Yes
2020-08-12T20:27:54 #kisslinux <dylanaraps> I feel that's better.
2020-08-12T20:28:26 #kisslinux <micr0> a real wrench in the works would be supporting directories in sources xD
2020-08-12T20:29:05 #kisslinux <micr0> so I could just have 'files/' as my only sources line but lets table that request for a few weeks
2020-08-12T20:29:09 #kisslinux <dylanaraps> I don't see why that's needed when the current format supports that.
2020-08-12T20:30:33 #kisslinux <micr0> sorry, i mean <source> being a directory
2020-08-12T20:30:41 #kisslinux <dylanaraps> Ah
2020-08-12T20:30:53 #kisslinux <micr0> I think that would add a lot of implementation complexity
2020-08-12T20:31:03 #kisslinux <micr0> depending on how things are implemented right now
2020-08-12T20:31:08 #kisslinux <dylanaraps> OK. It'd copy /contents/ to the build directory.
2020-08-12T20:31:22 #kisslinux <dylanaraps> Second field can be used to specify a destination.
2020-08-12T20:31:26 #kisslinux <micr0> it would also break the 1 line in sources = 1 line in checksums though if checksums is only for machines, whatevs
2020-08-12T20:31:48 #kisslinux <dylanaraps> checksums would have to be skipped.
2020-08-12T20:31:55 #kisslinux <dylanaraps> OR. It could tar the directory.
2020-08-12T20:31:59 #kisslinux <dylanaraps> Checksum that.
2020-08-12T20:32:05 #kisslinux <dylanaraps> Then extract to copy.
2020-08-12T20:32:17 #kisslinux <dylanaraps> Existing tar code could then handle it.
2020-08-12T20:32:37 #kisslinux <dylanaraps> That's disgusting lol
2020-08-12T20:33:00 #kisslinux <micr0> yknow i was thinking of doing that manually but then what popped into my head was "well but its nice to see each file in the checksums file, and would be a pain to have to re-tar when editing"
2020-08-12T20:33:24 #kisslinux <dylanaraps> lol
2020-08-12T20:33:50 #kisslinux <dylanaraps> Easy solution to transition over to checksums files without file names...
2020-08-12T20:33:56 #kisslinux <dylanaraps> Just ignore second field when comparing them.
2020-08-12T20:34:05 #kisslinux <dylanaraps> That way it supports both at once.
2020-08-12T20:34:24 #kisslinux <dylanaraps> This has no effect on verification whatsoever.
2020-08-12T20:34:40 #kisslinux <micr0> um starting to lose you a bit
2020-08-12T20:34:51 #kisslinux <micr0> so right now, we dont look at second field of checksums
2020-08-12T20:35:00 #kisslinux <dylanaraps> Package manager just does a diff
2020-08-12T20:35:14 #kisslinux <dylanaraps> It checks that the files match.
2020-08-12T20:35:25 #kisslinux <dylanaraps> If they don't in /any/ way, verification fails.
2020-08-12T20:35:45 #kisslinux <dylanaraps> By making the package manager only diff the first column (the checksums themselves), we'd support both formats at once.
2020-08-12T20:36:12 #kisslinux <dylanaraps> This is the only part which must /always/ match 1:1.
2020-08-12T20:37:09 #kisslinux <micr0> if we wanted to support sources/<sources> to be a directory, we could a) skip checksumming the directory, b) tar and checksum magic, c) drop the second column in checksums, or d) change the second column in checksums to be the full path
2020-08-12T20:37:43 #kisslinux <dylanaraps> Also note: We haven't yet made a breaking change to the package format since it was defined.
2020-08-12T20:38:12 #kisslinux <micr0> d only works if whatever recursive file listing function kiss uses is consistent
2020-08-12T20:38:57 #kisslinux <dylanaraps> It'd either be the file command (no fixed ordering) or a glob. Could even be a glob + file for the purposes of fixing the order.
2020-08-12T20:38:58 #kisslinux <micr0> yeah this is why its engineering - balancing simplicity, features, speed, stability
2020-08-12T20:39:05 #kisslinux <dylanaraps> :P
2020-08-12T20:39:45 #kisslinux <micr0> nice part about changing checksums is that the file is meant to be output by kiss checksum and committed so that isnt bad at all
2020-08-12T20:40:26 #kisslinux <micr0> supporting <source> directories i think is also a non-breaking change, in that it is an added, optional feature (though maybe you want enforcing a trailing / to differentiate)
2020-08-12T20:41:02 #kisslinux <dylanaraps> Yes
2020-08-12T20:41:17 #kisslinux <dylanaraps> I exclude extensions to the format as these have had no effect on existing syntax
2020-08-12T20:41:43 #kisslinux <micr0> I don't know enough about file or POSIX globbing rules (*shudder*) to trust anything that isn't as heavy-handed as just like running `sort`
2020-08-12T20:41:53 #kisslinux <dylanaraps> That too
2020-08-12T20:41:57 #kisslinux <dylanaraps> How could I forget sort
2020-08-12T20:42:42 #kisslinux <micr0> sort on filename for humans, sort on checksum for robots. both would work.
2020-08-12T20:46:25 #kisslinux <dylanaraps> Checking only first column can be down with: awk 'NR==FNR{a[$1];next} !(($1) in a) {exit 1}' file1 file2
2020-08-12T20:46:44 #kisslinux <dylanaraps> Returns 1 for mismatch, 0 if match.
2020-08-12T21:36:37 #kisslinux <dilynm> Chromium just wants me dead, and by extension qt5-webengine
2020-08-12T21:37:03 #kisslinux <dilynm> What's up with devs ASSUMING libstdc++ for Linux and libc++ for osx, as if Linux users just can't have llvm...
2020-08-12T21:39:35 #kisslinux <dylanaraps> micr0: [200~https://github.com/kisslinux/kiss/commit/fecf4995f8b580714a2517d9dab6521876c27bd2~
2020-08-12T22:09:48 #kisslinux <micr0> woah, thanks dylanaraps!
2020-08-12T22:10:03 #kisslinux <micr0> also the code changes seem easier to follow
2020-08-12T22:11:26 #kisslinux <j`ey> that awk line is a bit of an eye sore tho
2020-08-12T22:11:38 #kisslinux <fehawen> herro o/
2020-08-12T22:16:29 #kisslinux <dylanaraps> micr0: https://github.com/kisslinux/kiss/commit/99f61fc3403ba393da1ee35d39ec5d302f5ef5e4
2020-08-12T22:17:25 #kisslinux <dylanaraps> Absolute paths should /maybe/ use KISS_ROOT. Not sure yet.
2020-08-12T22:17:51 #kisslinux <dylanaraps> Also need to see if 'path/.' is POSIX or not.
2020-08-12T22:18:04 #kisslinux <dylanaraps> (This makes cp copy contents instead of the directory itself)
2020-08-12T22:18:07 #kisslinux <j`ey> do you have an idea of what absolutee paths could be used for?
2020-08-12T22:18:13 #kisslinux <dylanaraps> No idea
2020-08-12T22:18:25 #kisslinux <dylanaraps> This change may not stick
2020-08-12T22:18:32 #kisslinux <dylanaraps> (Goes for any change really)
2020-08-12T22:19:20 #kisslinux <dylanaraps> I prefer getting a commit done regardless as it logs the work in Git.
2020-08-12T22:20:08 #kisslinux <j`ey> and no packages will use that feature until it's released anyway
2020-08-12T22:20:37 #kisslinux <dylanaraps> No packages in the official repositories will ever use it. Nor Community.
2020-08-12T22:20:56 #kisslinux <dylanaraps> (relative directories, if needed, yes. absolute ones, no)
2020-08-12T22:21:10 #kisslinux <dylanaraps> (same for files, as is already the case)
2020-08-12T22:21:34 #kisslinux <j`ey> oh
2020-08-12T22:22:27 #kisslinux <dylanaraps> It's not something /good/ per se but I don't see why the package manager shouldn't support it.
2020-08-12T22:27:07 #kisslinux <dylanaraps> Anyway
2020-08-12T22:27:09 #kisslinux <dylanaraps> I need to sleep
2020-08-12T22:27:11 #kisslinux <dylanaraps> o/