LTSP 5 is in minimal maintenance mode
The new LTSP is hosted at https://ltsp.github.io

IRC chat logs for #ltsp on irc.freenode.net (webchat)


Channel log from 2 September 2019   (all times are UTC)

04:55
<alkisg>
rc.d/X95-run-x-session: XDG_SEAT=${XDG_SEAT:-seat0} XDG_VTNR=${SCREEN_NUM#0} su - ${LDM_USERNAME} -c "$CLIENT_ENV $MY_LANG DISPLAY=$DISPLAY ICEAUTHORITY=$ICEAUTHORITY XAUTHORITY=$XAUTHORITY $LDM_XSESSION $LDM_SESSION"
04:55
That's why I'm always complaining to quote things :D
04:55
su - "$LDM_USERNAME" should at least fail there
05:12
<vagrantc>
oh, that would be interesting
05:12
that's a pretty simple fix for that ... still would like to know *why* it's happening
05:19* alkisg could look into it but later, maybe in a couple of weeks...
05:20
<alkisg>
su - "" -c whoami
05:20
root
05:20
Bah
05:21
<vagrantc>
su - "" -c whoami
05:21
su: user does not exist
05:21
works as desired on buster...
05:22
<alkisg>
Maybe fixed on buster... we can't rely on that though :/
05:22
<vagrantc>
test -z "${LDM_USERNAME}"
05:22
test -z "${LDM_USERNAME}" || exit 1
05:22
er... test -n ...
05:22
should
05:22
solve the issue ... just a question of where to put it
05:23
unless the guest/autologin stuff relies on that being empty... but not at that point, surely.
05:24
i think it's in the C code ... making some assumption about the user's shell
05:24* alkisg would put it in ldm-script, xsession)
05:26
<vagrantc>
sounds feasible
05:26* vagrantc will test a bit more...
05:26
<vagrantc>
some other day...
05:26
<alkisg>
:)
05:26
shift change
05:26
<vagrantc>
heh
05:27
<alkisg>
vagrantc goes, alkisg comes, like the old days :D
05:27
<vagrantc>
indeed
05:27kjackal has joined IRC (kjackal!~quassel@2a02:587:3110:a100:a551:b0ce:a287:d7de)
05:36vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Ping timeout: 264 seconds)
05:42ricotz has joined IRC (ricotz!~ricotz@ubuntu/member/ricotz)
07:27kjackal has left IRC (kjackal!~quassel@2a02:587:3110:a100:a551:b0ce:a287:d7de, Remote host closed the connection)
07:30kjackal has joined IRC (kjackal!~quassel@2a02:587:3110:a100:75d7:fcb3:75b3:1718)
08:05woernie has joined IRC (woernie!~werner@p50867A86.dip0.t-ipconnect.de)
08:13statler has joined IRC (statler!~Georg@gwrz3.lohn24.de)
11:08alexxtasi[m] has left IRC (alexxtasi[m]!alexxtasim@gateway/shell/matrix.org/x-kenvmmvyqovmypoj, Write error: Connection reset by peer)
11:08uumas has left IRC (uumas!uumasmatri@gateway/shell/matrix.org/x-lrtvthotcangbmmh, Remote host closed the connection)
11:14alexxtasi[m] has joined IRC (alexxtasi[m]!alexxtasim@gateway/shell/matrix.org/x-oiqmennsoysvyeyr)
11:48uumas has joined IRC (uumas!uumasmatri@gateway/shell/matrix.org/x-iihxhmtbcopxmyum)
12:01Faith has joined IRC (Faith!~Paty_@unaffiliated/faith)
12:02section1 has joined IRC (section1!~section1@178.33.109.106)
12:11
<alkisg>
uumas, would you be willing for me to troubleshoot that ldm issue in your setup over vnc?
12:11* alkisg has migrated to ltsp19, and would take some time to setup ltsp5 again...
12:19
<uumas>
alkisg: Sure! Do you have time now?
12:20* uumas uses ansible, which makes spinning up a new vm take 15min of waiting and 1min of work.
12:20
<alkisg>
uumas, I have time in 30 mins, is it ok?
12:21
<uumas>
Should be
12:21kjackal_v2 has joined IRC (kjackal_v2!~quassel@2a02:587:3110:a100:7c2c:406a:3a98:c136)
12:24kjackal has left IRC (kjackal!~quassel@2a02:587:3110:a100:75d7:fcb3:75b3:1718, Ping timeout: 252 seconds)
12:43
<alkisg>
uumas, ready, do either one of those:
12:43
!vnc-edide
12:43
<ltsp>
vnc-edide: To share your screen with me, open Epoptes → Help menu → Remote support → Host: srv1-dide.ioa.sch.gr, and click the Connect button
12:43
<alkisg>
!vnc-dide
12:43
<ltsp>
vnc-dide: To share your screen with me, run this: sudo apt-get --yes install x11vnc; x11vnc -connect srv1-dide.ioa.sch.gr - this is a reverse connection, it doesn't need port forwarding etc.
12:44
<uumas>
Ah, I'll need to find my way into a gui
12:44
<alkisg>
uumas, doesn't the ltsp server have ui?
12:44
(or even the host outside it...)
12:45
<uumas>
Yes, it does. Just not used to accessing it :D
12:45
<alkisg>
Hehe
12:45
The quickest troubleshooting is with reverse vnc connections and then via epoptes to reach the clients
12:46
<uumas>
Here we go
12:47
<alkisg>
VNC is trying to receive a connection, but I think it fails
12:47
What does it say at your end?
12:48
<uumas>
Nothing signaling an error, but could be the firewall we're put behind. I'll connect to a vpn to bypass that.
12:55alkisg_web has joined IRC (alkisg_web!2ef6c856@46.246.200.86.dsl.dyn.forthnet.gr)
12:55alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Read error: Connection reset by peer)
12:56
<alkisg_web>
uumas, thunderstorm => my office went down, could you try: x11vnc -connect 46.246.200.86
12:56
<uumas>
Oh wow xD
12:57
That ip doesn't even respond to ping
12:57
<alkisg_web>
OK office is back up
12:57
Sorry for this, let's try this again:
12:58
!vnc-dide
12:58
<ltsp>
vnc-dide: To share your screen with me, run this: sudo apt-get --yes install x11vnc; x11vnc -connect srv1-dide.ioa.sch.gr - this is a reverse connection, it doesn't need port forwarding etc.
12:58alkisg has joined IRC (alkisg!~alkisg@ubuntu/member/alkisg)
13:09
<alkisg>
uumas, why doesn't fish allow normal shell commands like $(id -u)?
13:14woernie has left IRC (woernie!~werner@p50867A86.dip0.t-ipconnect.de, Remote host closed the connection)
13:14
<uumas>
alkisg: It does it using (id -u)
13:15
<alkisg>
uumas, but then it's not a normal posix shell
13:15
So we can't use it in ltsp to run commands
13:15
OK so we have 3-4 things to address:
13:16
1) Your cupsd is running at 100%; I'm wondering how you're able to work in your clients like this, with 1 cpu lost :D
13:16
(this doens't happen to other users though, might be something with your firewall etc; it's for you to find out)
13:16
<uumas>
Doesn't almost every shell have a different syntax for that? Some shells have `id -u` which I believe also works on bash
13:17
<alkisg>
`id -u` is supported in posix shells, but it's been deprecated by $()
13:17
2) So, we need to find out how to run commands over ssh using sh, not the user shell
13:18
<uumas>
ssh sh -c <command> ?
13:18
<alkisg>
The problem starts with this line : IFS=':' read -r LDM_USERNAME dummy pw_uid pw_gid pw_gecos LDM_HOME pw_shell <<EOF
13:18
ssh runs the command with the user shell, so if fish blocks $(), it won't allow sh -c to run with them,
13:18
and ssh doesn't allow remote quoting either
13:18
So it would be best to send a shell script and execute it or something
13:19
or use ssh sh <<EOF, things like that
13:19
The failing line is at: /usr/share/ldm/rc.d/X01-localapps
13:19
But even when I fixed that, fish again made the client hang elsewhere
13:19
So, 3) troubleshoot more things that go wrong with fish, even after that line, and
13:20
4) even if fish fails, it shouldn't login as root; which would be a test in X95-run-x-session
13:20
I'm not sure vagrantc will have time to work on all these, I certainly don't want to invest all that time to ltsp5, when ltsp19 is already more stable
13:21
So maybe (4) is enough, to fail but at least not allow root login
13:21
(btw I committed the fix for the 5.2 kernel this morning)
13:21
<uumas>
I'd say the more important thing is to not login as root. Having some little used shells not work isn't that big of a problem
13:22
<alkisg>
Indeed
13:22
<uumas>
Yeah, it worked on 5.2 now
13:22
<alkisg>
I'll write those to the bug report and let vagrant decide what to do with it
13:22
ty for the remote desktop
13:22
<uumas>
No problem :D
13:24
<alkisg>
ah
13:24
uumas, I just saw that I did some test wrong
13:25
Can we reconnect?
13:27
<uumas>
Sorry, need to go for a while. Can probably try in 1-2 hours.
13:27
<alkisg>
np
13:27
I'll comment that "maybe this will work" in the bug
13:28
Done
13:33alkisg_web has left IRC (alkisg_web!2ef6c856@46.246.200.86.dsl.dyn.forthnet.gr, Remote host closed the connection)
15:32vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc)
15:45
<alkisg>
vagrantc, I commented on the bug report about the ldm bug
15:45
It's in X01-localapps; but we can ignore that part and just put the "test -n "$LDM_USERNAME" || die" thing
15:46
<vagrantc>
saw it
15:47
should probably put it all over the place...
15:47
i mean, check anywhere it actually uses it
15:47
<alkisg>
Maybe once after the "read -r" would be enough
15:47
<vagrantc>
could pop up an error with ldm-dialog
15:49
alkisg: well, the same thing could happen somewhere else with another plugin
15:49
<alkisg>
LDM_USERNAME is properly set; but X01-localapps destroys it if read fails
15:49
I don't think it can be empty elsewhere...
15:50
<vagrantc>
in ltsp/ldm, maybe it doesn't, but someone else could implement the same type of error in some other hook
15:51
better to double-check in our code rather than assume it is set...
15:52
at least test -n "$LDM_USERNAME" wherever we call "su"
15:53
alkisg: thanks for finding the cause!
16:03
<alkisg>
👍
16:09
<uumas>
alkisg: Do you still need to do testing? I'll be home in ~40min
16:10
<alkisg>
uumas, if you want to use fish, i have a workaround that might work; if you don't care, I don't either :D
16:10
We will avoid the root thing, so np there
16:11
<uumas>
Yeah, I'm switching to ltsp19 soon anyway and can just have `exec fish` in bashrc for now, so that won't be a problem.
16:12
<alkisg>
ok
16:20
<vagrantc>
does ltsp19 also have similar issues with non-bournish shells?
16:21
<alkisg>
AFAIK no as I don't exec anything
16:21
I run `ssh exit` ):
16:27
<vagrantc>
oh, i was just wondering why uumas would need to exec fish from bashrc
16:28
<alkisg>
I think he means in ltsp5
16:30
<vagrantc>
ah, "for now [with ltsp5]"
16:31
fish is a really nice shell
16:32
<alkisg>
I didn't understand what they had against $() though, why they only support ()
16:33
vagrantc: did you say it fails with some other shell too?
16:33
Are there other shells that don't like $() ?
16:35
<vagrantc>
according to wikipedia $ is only used for variable expansion in fish
16:35
alkisg: the csh varians
16:35
although i'm not sure that's *why* it fails
16:37
we should fork ltsp5 to only support fish.
16:39adrianorg has joined IRC (adrianorg!~adrianorg@179.177.215.228.dynamic.adsl.gvt.net.br)
16:40ZAJDAN has left IRC (ZAJDAN!~zdenek@77.48.149.75, Ping timeout: 245 seconds)
16:41ZAJDAN has joined IRC (ZAJDAN!~zdenek@77.48.149.75)
16:42adrianor1 has left IRC (adrianor1!~adrianorg@179.177.213.173.dynamic.adsl.gvt.net.br, Ping timeout: 244 seconds)
16:51statler has left IRC (statler!~Georg@gwrz3.lohn24.de, Remote host closed the connection)
17:24
<quinox>
how incompatible are fish and bash when it comes to scripts?
17:28
it looks nice
17:33
<uumas>
quinox: Incompatible for other than really, really basic stuff. I just write my scripts in bash even though I use fish as my main shell.
17:33
<quinox>
okay
17:36
it seems like a really neat shell, that tab completion with man-page info looks superb
17:40kjackal_v2 has left IRC (kjackal_v2!~quassel@2a02:587:3110:a100:7c2c:406a:3a98:c136, Ping timeout: 276 seconds)
17:48mmarconm has joined IRC (mmarconm!~mmarconm@unaffiliated/mmarconm)
17:49
<mmarconm>
MWALTERS: do u use shared folder mapped to user ?
17:58
<alkisg>
vagrantc: I'm about to release ltsp 19.09 beta, and I was thinking to start using tags. Are you using the "lightweight" or the "annotated" ones?
18:01
I'm thinking: update changelog; git commit -am 'Version 19.09'; git tag -a v19.09 -m 'Version 19.09'
18:02mmarconm has left IRC (mmarconm!~mmarconm@unaffiliated/mmarconm, Quit: Leaving)
18:08
<vagrantc>
i've always just used: git tag -s TAG
18:08
for signed
18:08
and it pops up an editor, so i can verify that i'm signing the right thing
18:09
e.g. "git commit -p && git tag -s TAG" ... because i like reviewing my changes
18:09
<alkisg>
Nice, ty
18:10
<vagrantc>
i guess because i use signed tags it implies git tag -a ?
18:10
<alkisg>
No idea, first time reading about tags :D
18:12
<vagrantc>
signed tags were a big motivation for me to switch to git
18:13
so i don't know if i've ever even use unsigned annotated tags
18:17kjackal has joined IRC (kjackal!~quassel@2a02:587:3110:a100:7c2c:406a:3a98:c136)
19:56section1 has left IRC (section1!~section1@178.33.109.106, Quit: Leaving)
20:17alexxtasi[m] has left IRC (alexxtasi[m]!alexxtasim@gateway/shell/matrix.org/x-oiqmennsoysvyeyr, *.net *.split)
20:21alexxtasi[m] has joined IRC (alexxtasi[m]!alexxtasim@gateway/shell/matrix.org/x-asuykdwunegmpwxq)
20:47Faith has left IRC (Faith!~Paty_@unaffiliated/faith, Quit: Leaving)
20:48ricotz has left IRC (ricotz!~ricotz@ubuntu/member/ricotz, Remote host closed the connection)
20:51MWALTERS has left IRC (MWALTERS!~ubox@c-73-152-61-86.hsd1.va.comcast.net, Ping timeout: 245 seconds)
20:53MWALTERS has joined IRC (MWALTERS!~ubox@c-73-152-61-86.hsd1.va.comcast.net)
20:58josefig has joined IRC (josefig!~Jose@unaffiliated/josefig)
22:57kjackal has left IRC (kjackal!~quassel@2a02:587:3110:a100:7c2c:406a:3a98:c136, Ping timeout: 252 seconds)
22:57kjackal has joined IRC (kjackal!~quassel@athedsl-173226.home.otenet.gr)