[00:03] <wgrant> hallyn: I normally use the 45W one, but I sometimes use my old bigger T400 one without trouble.
[00:09] <hallyn> wgrant: you've never had this charging trouble I'm talking about i take it?
[00:10] <hallyn> I really don't wanna be laptop-less for 2 weeks as i send this one in... :(
[00:10] <wgrant> hallyn: Nope :/
[00:10] <wgrant> hallyn: Will it turn back on immediately if you unplug the external battery and disable the internal one, then replug?
[00:11] <hallyn> how do i disable the internal one?
[00:11] <hallyn> but, the last two times it did after i unplugged the external, but the first time it did not
[00:11] <hallyn> so, want a few more experiments i guess
[00:11] <wgrant> hallyn: There's a button on the bottom.
[00:11] <hallyn> (the first time i was just really pissed off)
[00:11] <hallyn> oh?
[00:11] <wgrant> IIRC
[00:12] <wgrant> Hm, I thought there was.
[00:12] <wgrant> Oh, maybe it's not externally accessible, damn.
[00:12] <wgrant> I was remembering the internal one.
[00:13] <wgrant> But yeah, I've never had a problem like that.
[00:13] <wgrant> The biggest issue I have is that it doesn't detect a hot-swapped external battery without a suspend/resume or unplug/replug cycle, and I think that's a Linux issue.
[00:15] <hallyn> ok - thx
[03:09] <desrt|pdx> hallyn: you can't do subtree listens for bus names -- you have to register each one separately
[03:16] <hallyn> desrt|pdx: hm, did try that, didn't seem to get interface org.freedesktop.Scope method Abandon to register, but I'll keep going down that route then, thanks
[03:17] <desrt|pdx> oh
[03:17] <desrt|pdx> if you want new _interfaces_ then you need to make a new call to register_object
[03:18] <desrt|pdx> and ya -- in that case, subtree may make sense depending on what you're up to
[03:19] <hallyn> systemd-logind calls method Abandon on interface Scope to let us know we can remove the cgroup.  i just wanna listen to that so we can act on it
[03:20] <desrt|pdx> you'll probably need to use a subtree
[03:20] <desrt|pdx> because of the stateless thing
[03:23] <hallyn> ok.  actually i guess i just got that to actually create me some cgroups.
[04:24] <pitti> Good morning
[06:32] <dholbach> good morning
[06:36] <dholbach> seb128, does the desktop team still use the ~ubuntu-desktop branches?
[06:36] <seb128> dholbach, yes, why?
[06:37] <dholbach> seb128, ok, it was just a few I found of date
[06:37] <seb128> of date?
[06:38] <dholbach> out of date
[06:38] <dholbach> sorry
[06:39] <seb128> which ones?
[06:40] <seb128> things is that some packages don't always have an Ubuntu delta, and we don't commit updates when packages got e.g back in sync with Debian
[06:40] <dholbach> ah ok
[07:47] <LocutusOfBorg1> Hi, what is needed to fix lp 1310049?
[07:47] <LocutusOfBorg1> the patch is already there
[07:48] <seb128> LocutusOfBorg1, subscribe ubuntu-sponsors to the bug?
[07:58] <LocutusOfBorg1> yes, seb128 I had already done this, a few minutes ago, I was looking for the SRU documentation
[07:58] <seb128> LocutusOfBorg1, https://wiki.ubuntu.com/StableReleaseUpdates
[08:49] <LocutusOfBorg1> yes seb128  was looking at it ;)
[08:49] <seb128> LocutusOfBorg1, looking at what?
[08:49] <seb128> oh, the SRU documentation, I though you were asking for it
[09:02] <mvo_> mardy: hi, just a quick heads-up, I uploaded a new accounts-qml-module with a trivial multi-arch releated change, the control file says direct uploads are ok, but if you prefer a branch instead just say so and I will create one
[09:17] <xnox> desrt|pdx: hey, you are not totally hooked up for non-uploading DD status =) https://nm.debian.org/public/process/desrt
[09:17] <xnox> I'm sure seb128 would be very happy about that =)
[09:18] <seb128> desrt|pdx is going for dd?! ;-)
[09:19] <xnox> seb128: yeap! non-uploading though =)
[09:19] <xnox> seb128: although we might trick him into an uploading DD later on.
[09:19] <seb128> is the process less crazy than for DD with upload? ;-)
[09:20] <xnox> seb128: it's less crazy - one must pass the secret handshake test, licensing test, etc. But not e.g. how to use dput test.
[09:20] <seb128> k
[09:20] <seb128> no "rewrite your debian/rules in pure makefile without using the dh tools"? ;-)
[09:25] <mlankhorst> hah
[09:25] <mlankhorst> well DD is easy. :P
[09:25] <mlankhorst> even uploading
[10:46] <tjaalton> just takes time
[10:49] <mardy> mvo_: if you create a branch, you would save me some time -- if you do have time yourself, that is :-)
[11:22] <sil2100> dholbach: hey! I have a small request - could you move my this-month's patch pilot duty one week later? :)
[11:22] <sil2100> dholbach: and if you want, you can put my name 2 times in the month if anything
[11:30] <stgraber> apachelogger: is that on utopic?
[11:31] <tseliot> pitti: hey, do you know why/how requesting suspend from logind could result in a timeout?
[11:31] <tseliot> pitti: in 14.04
[11:34] <apachelogger> stgraber: trusty
[11:35] <stgraber> odd. I'm aware of some kernel issues with unprivileged overlayfs on utopic but not on trusty...
[11:36] <stgraber> apachelogger: is that just a straight unprivileged lxc-start-ephemeral or are you doing nesting?
[11:36] <apachelogger> stgraber: straight one, on AWS if that matters
[11:37] <stgraber> apachelogger: ok, just to confirm it's a generic issue and nothing to do with your jenkins setup or its directories, can you try a simple (still unprivileged): lxc-create -t download -n test -- -d ubuntu -r trusty -a amd64 && lxc-start-ephemeral -o test
[11:42] <mvo_> mardy: thanks,  I will make sure to create a branch for you
[11:48] <apachelogger> stgraber: http://paste.ubuntu.com/8214763/
[11:48] <apachelogger> stgraber: ah mind you, the container is utopic
[11:51] <apachelogger> pft
[11:51] <apachelogger> stgraber: I can't reproduce it anymore :S
[11:51] <apachelogger> ah, actually
[11:53] <apachelogger> stgraber: I since changed the entire home to the first subuid as I found no other way to make rw bind mounts work, so I guess that would have solved the permission problems
[11:57] <apachelogger> stgraber: fails for new user http://paste.ubuntu.com/8214821/
[12:00] <stgraber> apachelogger: does a simple "lxc-start -n test" work for that new user?
[12:04] <apachelogger> stgraber: yep http://paste.ubuntu.com/8214858/
[12:08] <pitti> tseliot: could be bug 1253456 or bug 1252121, I just never could reproduce it on a box that I could debug on
[12:10] <tseliot> pitti: I can't reproduce it here either but an engineer can here: https://bugs.launchpad.net/fwts/+bug/1360195
[12:10] <tseliot> I don't know if it's the same issue though
[12:38] <stgraber> apachelogger: hmm, sadly I'm in the middle of a move and my trusty dev box isn't quite back online yet...
[12:38] <stgraber> it's a bit puzzling why it works for one user and not for the other though
[12:40] <mvo_> xnox: could you please try http://paste.ubuntu.com/8215080/
[12:40] <mvo_> xnox: ?
[12:45] <mvo_> xnox: you can test using ./build/bin/apt-helper detect-proxy http://foo.bar
[12:46] <apachelogger> stgraber: it works on the one user because I manually cheated my way to a working setup by changing the uid entirely -> http://paste.ubuntu.com/8215121/
[12:47] <dholbach> sil2100, you can move it wherever you like
[12:47] <dholbach> sil2100, it's just there as a reminder
[12:47] <mlankhorst> stgraber: can someone from the release team ack mesa 10.3 or xorg-server 1.16?
[12:48] <aerocarbine> join #teen
[12:50] <apachelogger> stgraber: I am not sure how this is intended to work but from what I understand the ephemeral setup fails because it tries to create a folder (delta0) in a folder that is not owned by the mapped uid... e.g. in the jenkins example share/local/lxc/foo would be owned by uid=120 but the map only starts at 100000 which results in permission errors. I also observed the same thing with a snapshot clone and a bind mount (which is why I flipped the
[12:50] <apachelogger> uid of the jenkins user), as long as the to-be-bound-dir doesn't have write permissions for uid100000 it quite simply won't be writable
[12:50] <apachelogger> I do find that last bit weird though considering 100000 is a subuid of whatever uid the user has xD
[12:52] <stgraber> apachelogger: ah, you shouldn't be needing that kind of trick at all, what you do need however is execute access for the mapped uid all the way to .local/share/lxc
[12:52] <stgraber> apachelogger: which typically is as simple as "chmod +x /home/<user>"
[12:53] <apachelogger> stgraber: mhh, I did the chmod on both accounts though :/
[12:53] <stgraber> apachelogger: let me see if I can figure out how to start a canonistack instance so I can do a very quick test there
[12:59] <pitti> wgrant: oh, a langpack! https://translations.launchpad.net/ubuntu-rtm/14.09/+language-packs
[12:59] <pitti> wgrant: was this an one-off thing, or is that cron'ed now?
[13:01] <wgrant> pitti: Oh, sorry, forgot to poke you about that. We should coordinate the cron schedule.
[13:02] <wgrant> The import backlog completed over the weekend.
[13:02] <pitti> wow, still 199 MB? I had expected it to be a lot smaller
[13:03] <wgrant> It might make sense to disable langpack support for some of the templates.
[13:03] <wgrant> I haven't looked to see where the major size is.
[13:03] <pitti> wgrant: so, incidentally I just fixed langpack-o-matic to now fully get along with update tarballs (we need that for ubuntu anyway)
[13:03] <pitti> so the requirement for "always need a full export" can be dropped
[13:03] <wgrant> All templates for sources that we copied were copied.
[13:04] <stgraber> apachelogger: got a trusty instance up, though it's freakishly slow, trying to get lxc running in there now :)
[13:05] <pitti> wgrant: so from my side I'm not fussed about any particular time in teh week
[13:05] <pitti> wgrant: so you can run it at the time that is best
[13:07] <wgrant> pitti: Thursday for the export?
[13:07] <pitti> wgrant: sounds good! then Friday for the build/upload
[13:07] <wgrant> Yep
[13:22] <smoser> adam_g`, i do have access to such logs.
[13:29] <israel> Hi, I am wondering how to access files outside of a squashfs... I am building a Live CD without ubiquity.  Is there a way to link to files in the /live directory of the actual image to link to the LiveCD $HOME?
[13:30] <israel> in Debian it is ln -s /lib/live/mount/medium/live/ but this directory structure does not exist in the image I am building... what is the equivalent?
[13:31] <xnox> mvo_: building
[13:36] <shadeslayer> israel: how are you building the ISO?
[13:37] <israel> shadeslayer I made a script... I am using chroot, basically I followed https://help.ubuntu.com/community/LiveCDCustomizationFromScratch
[13:38] <israel> But I am customizing it using an alternate installer called OBI, rather than Ubiquity as I am targetting machines with 128MB RAM using JWM and no display manager... just xinit in $HOME
[13:38] <shadeslayer> uh, not entirely sure
[13:39] <shadeslayer> israel: why do you want to do this linking btw?
[13:39] <israel> I have a working image, however I need to access a tar.xz stored in image/live/tarballs
[13:39] <Saviq> sergiusens, do you have a bug for ciborioum and system hint?
[13:39] <israel> :) I was answering as you typed
[13:40] <shadeslayer> israel: doesn't answer my question :)
[13:40] <shadeslayer> why do you need to access a tar on the binary ISO
[13:40] <shadeslayer> instead of having the tar inside the squashfs
[13:40] <israel> OBI uses the tar to install an entire OS.  It is a tarball of an existing system
[13:41] <israel> the entire device
[13:41] <xnox> mvo_: git diff output doesn't tell me base commit id nor branch the diff is against =) hence, i was asking for "git format-patch" as that has metadata for me to find out what the diff is for. As is that fails to compile, failing to find proxy.h headers. Let me fudge it in to compile.
[13:41] <xnox> and/or undo refactoring.
[13:42] <shadeslayer> wait so
[13:42] <israel> https://help.ubuntu.com/community/OBI
[13:42] <shadeslayer> israel: your ISO will have squashfs of the system as well as a tar of the entire system
[13:42] <shadeslayer> so basically doubling the size?
[13:42] <israel> yes... but the image is around 600MB total with the tar
[13:43] <mvo_> xnox: hm, give me some minutes, I shall have something with a proper git format and test then
[13:44] <israel> shadeslayer is there a way to copy the squashfs to the HD as the OS? and run some post install scripts to change a few things? (like user and moving /root to /home/$USER
[13:45] <shadeslayer> israel: well, the media will be mounted in /media/user/something/ , so mayeb you can patch OBI to look at that path?
[13:45] <shadeslayer> israel: ubiquity :p
[13:46] <israel> Ubiquity does not run in less than 384 MB  it locks the computer up... hence the need for OBI
[13:46] <dholbach> didrocks, your blog is a bit busy, right? :)
[13:46] <didrocks> dholbach: it seems it went popular, I'm rebooting it in a few :p
[13:49] <shadeslayer> israel: right
[13:49]  * shadeslayer would just use the text installer
[13:50] <israel> hmmm... OBI is a one button installer... it is very easy for a novice... text installer is much more complicated than ubiquity, don't you think?
[13:50] <shadeslayer> but ... who would have computers with 128 MB's of RAM :S
[13:50] <shadeslayer> seems like a corner case
[13:50] <xnox> mvo_: hm, also local build system seems to link against system libapt-pkg (and headers) instead of the just built ones....
[13:51] <didrocks> dholbach: up again
[13:51] <israel> shadeslayer people in poor economies who cannot afford a new computer, and cannot afford to stay with XP since it is unsupported
[13:52] <dholbach> didrocks, thanks! :)
[13:52] <shadeslayer> israel: but I doubt you can run much with 128 MB's of RAM
[13:52] <shadeslayer> for LO is out of the picture, as is firefox, chrome
[13:53] <shadeslayer> midori maybe, not sure how much RAM it takes
[13:53] <mvo_> xnox: http://paste.ubuntu.com/8215517/ <- this should actually be much nicer, sorry that it took me so long
[13:53] <israel> shadeslayer... I have a nice old 128MB device.  It runs just fine :)  The OS runs in about 60MB ram, so you can do A LOT  This isn't Gnome or even LXDE... ya know?
[13:54] <Unit193> shadeslayer: w3m will work fine.  Also, yes, you can generally use the squashed system itself as the new system.  I did an install once without ubiquity, not bad.
[13:54] <mvo_> xnox: against the debian/sid branch, you can test with apt-helper
[13:54] <xnox> +++ b/apt-pkg/contrib/proxy.h
[13:54] <xnox> +#include <apt-pkg/proxy.h>
[13:54] <israel> Midori works fine, FF works albeit slowly and Abiword is always an alternative to LibreOffice
[13:54]  * xnox is not sure how that works.
[13:55] <shadeslayer> xnox: heh
[13:55] <israel> Unit193 agreed, I use JWM though.  and without a display manager it eats up less ram
[13:56] <shadeslayer> israel: so returning to my original question, why can't the tar.gz be included in the squashfs
[13:56] <shadeslayer> would make your life easier?
[13:56] <mvo_> xnox: jetlag
[13:57] <israel> shadeslayer, I do not think this works...
[13:57] <xnox> mvo_: =))))))))
[13:57] <shadeslayer> israel: oh ? :S
[13:57]  * xnox passes starbucks loyalty card to mvo_ 
[13:57] <shadeslayer> ah, probably because the ISO mounted as ramfs
[13:57] <shadeslayer> and that would be huge
[13:58] <xnox> unless that will compile....
[13:58] <mvo_> xnox: maybe I'm super jetlaged, but I don't see this include in http://paste.ubuntu.com/8215517/ - at least not in proxy.h :)
[13:58] <israel> shadeslayer  as a side note, I figured it out... :)
[13:58] <israel> ln -s /live/image/live/ /root/img
[13:58] <israel> ln -s /live/image/live/tarballs /
[13:59] <israel> Sometimes it just helps talking about it, eh?  Like partner programming :)
[13:59] <mvo_> xnox: what do you think about adding a config for apt in libproxy-tools directly? so that when you install libproxy-tools apt will automatically use the "proxy" command? or do you think that is too suprising?
[13:59] <xnox> mvo_: hm =) i just nuked the whole git repo tree with hard reset and clean, and I no longer have it either =)
[13:59] <mvo_> xnox: *puhh* :) glad to hear
[13:59] <shadeslayer> israel: cool :)
[13:59] <mvo_> xnox: so the tea helped at least a bit
[13:59]  * xnox is also jetlagged a bit =)
[14:00] <israel> Thanks for being an ear to bounce ideas off of :)
[14:01] <xnox> mvo_: depends for which distro. In Debian, i think i need manual configuration variable for the PAC file url, on ubuntu it all needs to happen automagically if proxy-webkit is available and e.g. ubuntu services are used to set global proxy and pac file settings.
[14:02]  * xnox has debs \o/
[14:02] <xnox> testing
[14:05] <xnox> mvo_: magic =) all works like a charm!
[14:05] <mvo_> xnox: \o/
[14:06] <xnox> mvo_: this makes my life just a little bit more wonderful, when using "enterprise" network =)
[14:07] <mvo_> xnox: great, I will push into git - so libproxy-tools should ship a config or nt :) ?
[14:07] <mvo_> not
[14:08] <xnox> mvo_: not for now. In neither ubuntu or debian.
[14:08] <mvo_> xnox: *sigh*, ok :) I guess we need to put it into the documentation or something then as it seems to be super useful
[14:08] <xnox> mvo_: since in neither of those it would honor things as root. E.g. "apt-get download" would honor user session proxy settings from gnome, but "sudo apt-get download" would not.
[14:10] <xnox> mvo_: i will probably create a mini package which uses e.g. APT::PAC-URL thing to specify and fetch proxies, for debian (and/or as a patch for libproxy-tools, or apt with appropriate either compiled binary or python script)
[14:10] <xnox> and on ubuntu, i'll add support in ubuntu-service to store global PAC url for root user, and make libproxy plugin to talk to ubuntu-service to fetch all those root settings.
[14:10] <xnox> and then life will be complete.
[14:12] <mvo_> xnox: great
[14:14] <xnox> *****-
[14:23] <sergiusens> Saviq: no, did you create one?
[14:23] <Saviq> sergiusens, did not, doing now
[14:23] <Saviq> sergiusens, ciborium or lxc-android-config? both?
[14:31] <sergiusens> Saviq: both
[14:31] <sergiusens> I'll link if not
[14:37] <Saviq> sergiusens, bug #1364434
[14:37] <sergiusens> Saviq: thanks
[14:53] <tedg> jodh, hallyn, so where are we with fixing upstart cgroups?
[14:57] <jodh> tedg: I believe hallyn is on the case wrt the required systemd-shim changes - we can't really be convinced everything works as expected until we have that piece.
[14:58] <tedg> jodh, Okay, cool.
[14:58] <tedg> Hopefully we can get those pieces moving together.
[15:04] <hallyn> tedg: I'm working with $*&%(*$&% g_dbus.  I will have StopUnit - which kills all tasks on logout - working today.  I don't know how to get Abandon - which would just mark cgroups remove-on-empty on logout - working, but post what i have so hopefully someone who knows this stuff can tell me how to fix it
[15:04] <tedg> hallyn, Ah, cool.
[15:45] <barry> mvo_: ping
[15:45] <mvo_> barry: pong - want to play again?
[15:45] <barry> mvo_: still recovering from our last tourny :)
[16:00] <smoser> pitti, do you have any idea why ca-certificates wouldn't work ?
[16:01] <pitti> smoser: I haven't dug into that yet, but I suppose we sign the images and catalogue with a cert that isn't in ca-certificates?
[16:01] <smoser> this "works for me".
[16:01] <pitti> or sign it with a cert whose CA isn't there, I mean
[16:01] <smoser> $ wget https://cloud-images.ubuntu.com/query/released.latest.txt
[16:01] <pitti> smoser: that's on utopic?
[16:01] <smoser> yeah.
[16:02] <smoser> (note, all i did was that single command, but as i understood the bug you were asserting that fails)
[16:02] <pitti> infinity, slangasek, stgraber: TB meeting now
[16:02] <smoser> $ dpkg-query --show wget ca-certificates
[16:02] <smoser> ca-certificates	20140325
[16:02] <smoser> wget	1.15-1ubuntu1
[16:02] <pitti> same here
[16:02]  * pitti tries again in a schroot, but I thought I already did that yesterday
[16:02] <pitti> apt install wget
[16:03] <smoser> http://paste.ubuntu.com/8216407/
[16:03] <pitti> fun, it works there
[16:03] <smoser> you're sure you *had* ca-certificates in the root ?
[16:04] <pitti> Setting up ca-certificates (20140325) ...
[16:04] <pitti> yes
[16:04] <smoser> yeah, and cloud-iamge-utils does Depend on it
[16:04] <pitti> wget does, t oo
[16:04] <pitti> too
[16:05] <pitti> smoser: ah fun -- purging ca-certificates and reinstalling helps
[16:05] <pitti> so that smells like some weird upgrade issue
[16:05] <smoser> oh. that is fun.
[16:05] <smoser> pitti, yeah, very weird in that i dont have it here on my utopic
[16:05] <pitti> smoser: ok, then sorry for the noise! I'll set it to incomplete and see whether I can reprorduce that on an upgrade
[16:05] <smoser> which is upgraded pretty much daily-ish
[16:05] <pitti> I suspect some leftover conffiles from older versions, or missing upgrade cleanup or so
[16:09] <pitti> smoser: thanks for checking on your system!
[16:10] <smoser> pitti, if you see stuff like that, please feel free to msg/bump me in irc.
[16:10] <pitti> smoser: yeah, I did two days ago, but apparently it got lost in the noise
[16:10] <smoser> i miss a lot of bugmail, sadly i'm not as on top of things as you are  :)
[16:11] <pitti> so I decided to go for a bug report instead, and keep track of debugging info
[16:11] <pitti> wasn't all that trivial to get from "juju deploy" failure down to wget, so it at lesat helped wiht the notekeeping :)
[16:11] <pitti> ah, the beloved bug mail firehose :/
[16:18] <pitti> smoser: ok, upgrade worked as well; daily utopic dist-upgrade must have gotten that wrong at some point, but *shrug*, fixed itself :0
[16:18] <pitti> :) even
[16:18] <pitti> so only one bug left without a workaround
[16:54] <hallyn> pitti: desrt: ok, if i have a GVariant (returned by a dbus call), g_variant_get_type_string(variant) says "(as)" and g_variant_iter_n_children(iter) says 1, should I not be able to simply g_variant_iter_loop (&iter, "&s", &child) to get the content of that single string into child?
[16:55] <hallyn> (I asssume not -b c it's not working - but i've tried tons of variations like "(s)" "{s}" as well as using g_variant_iter_get_next)
[16:56] <hallyn> maybe i should just walk g_variant_get_child with a count
[16:56] <hallyn> it's not glib-y, but should work
[16:58] <pitti> hallyn: you mean g_variant_iter_next_value ()?
[16:58] <hallyn> yeah
[16:58] <pitti> or g_variant_iter_next () (more convenient in C)
[16:58] <pitti> hallyn: ought to work, yes (with just "s")
[16:58] <pitti> not &s
[16:59] <pitti> hallyn: but g_variant_get_child () seems altogether eaiser
[16:59] <pitti> hallyn: if g_variant_iter_next () with "s" doesn't work, then I think desrt will be better to answer that (I don't do much with variants, I have little experience with that API)
[17:05]  * pitti waves good night
[17:06] <hallyn> pitti: thanks, gnight.
[17:27] <hallyn> yay, got the string finally
[18:01] <hallyn> desrt: hi, if you have a few mins at some point, github.com/hallyn/systemd-shim #stopunit.2 has working org.freedesktop.systemd.Manager StopUnit support;  but I cannot get it to respond to org.freedesktop.systemd.Scope Abandon.  Any hints for obvoius reasons why would be greatly appreciated
[18:01] <hallyn> (could
[18:01] <hallyn> just be that i didn't set the method prototypes right, i suppose, but probably something even dumber)
[20:22] <JFSTWO> !ops
[20:30] <Unit193> henrix: Sorry to see you go, but thanks for your efforts!
[22:27] <cousteau> any chance of getting the font OCR-B on the repositories?  There's already OCR-A (fonts-ocr-a).  OCR-B is used often in Europe for ID cards and similar.
[22:27] <cousteau> (and I don't know if it can be freely distributed, but I found it for free on a quick "ocr-b ttf" google search)
[22:27] <cousteau> maybe I should ask debian?
[22:29] <cousteau> (it also seems to be available from Arch's AUR, so maybe this means it's legal to distribute, dunno)
[22:35] <Unit193> "my own contributions are public domain, and others have licensed their copyright claims."  Good luck, sir.
[22:43] <hallyn> desrt: hm, since I guess the error is "Failed to abandon scope session-6.scope: No such interface 'org.freedesktop.systemd1.Scope' on object at path /org/freedesktop/systemd1/unit/session_2d6_2escope"  I assume that measn I'm supposed to actually set up a new listener on that path when the unit is started?  meaning systemd-shim would have to persist for the duration of the session?  (or a child of its)
[22:44] <hallyn> i'd kinda prefer to hack systemd to call Manager.Abandon, but that wouldn't fly upstream :)
[22:49] <cousteau> Unit193, anyway that's not the font I found; the one I found has all iso8859-1 characters
[22:50] <cousteau> (I also found another one for 30-something dollars, so I'm not going to trust the legality of the first font I found)
[22:56] <hallyn> in wihch case maybe i didn't need to swicth to subtree after all