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 1 May 2019   (all times are UTC)

00:07
<Sleaker>
I should make sure it's not my monitor doing wierd things.
00:11
hmm yah I think my monitor was doing wierd things and like auto-swapping to another display. rebooted again and back up.
00:11
or just every-other boot is bad?
00:20
hmm ok so playing around with epoptes now. it looks like the server gets connected to from the client, but then it immediately DCs the client. and the client seems to connect before it even has finished booting into x
00:22
oh. that's probably just because it fetches the cert from the server during my initial scripts.
00:22
it looks like the epoptes-client service isn't connecting properly to the server.
00:29
hmm doesn't look like epoptes-client starts properly when called via systemctl.
00:31
ahhh the actual issue is the init script hard exits if /etc/ltsp_chroot doesn't exist.
00:32
seems odd...
00:34
<vagrantc>
i didn't think we had anything that relied on that still ... is that leftovers from an old install?
00:35
<Sleaker>
no this is a fresh image on the client
00:35
let me check which version it's grabbing
00:35
ii epoptes-client 0.5.10-2 all Computer lab management tool (client)
00:36
so I didn't change the boot line from what I previously had.
00:36
maybe this is part of the issue. so I also still have init=/sbin/init-ltsp in my pxe config
00:37
I should doublecheck what the autogenerated one sets up with now I guess
00:37
since I just updated the location of the image file
00:39
<vagrantc>
ah, yeah, older epoptes had the check only for backwards compatibility
00:39
e.g. ltsp 5.3.x
00:39
<Sleaker>
ii ltsp-server 5.5.9-2 all basic LTSP server environment
00:39
ii ltsp-server-standalone 5.5.9-2 all complete LTSP server environment
00:39
hmm but that stuff is getting included when building with 5.5 ?
00:43
ah is there no stretch-backports for 1.x?
00:44
well if I disable the 2 old-ltsp checks it shows up properly in the epoptes server window and I can connect to it just fine
00:45
<vagrantc>
the comment in the older epoptes startup script suggests that the if-up events aren't getting triggered, then
00:51
<Sleaker>
I don't know if ifupdown is even isntalled on this client image atm
00:51
I had to manually tell it to install a bunch of networking stuff
00:52
yah the stretch image doesn't default a lot of the networking stuff except what's like super necessary
00:57
probably just goes away if/when I push buster though I guess.
00:58
<vagrantc>
admittedly, i've done a lot less testing of buster's ltsp
00:58
<Sleaker>
ahh well.. I guess I'll have a lot of fun when it comes up :D
00:58
I ran into some wierd issues with xterm input where we had to ratelimit usb scanners :-/
00:59
<vagrantc>
though a lot of bugs have been fixed in theory :)
00:59
<Sleaker>
not related to ltsp..
05:08kjackal has joined IRC (kjackal!~quassel@2a02:587:793e:5d00:cd36:ac7a:accf:1899)
06:13woernie has joined IRC (woernie!~werner@p5B296A22.dip0.t-ipconnect.de)
06:17kjackal has left IRC (kjackal!~quassel@2a02:587:793e:5d00:cd36:ac7a:accf:1899, Ping timeout: 258 seconds)
06:21bcg has left IRC (bcg!~b@82-128-233-166.bb.dnainternet.fi, Ping timeout: 246 seconds)
06:27bcg has joined IRC (bcg!~b@82-128-233-166.bb.dnainternet.fi)
06:30kjackal has joined IRC (kjackal!~quassel@2a02:587:793e:5d00:cd36:ac7a:accf:1899)
06:32
<alkisg>
Ah, ltsp 5.18 isn't available in stretch-backports...
06:34
<vagrantc>
back when i used it all the time, i had a lot of motivation to backport everything ... now...
06:34* vagrantc shrugs
06:35
<alkisg>
Sure, understandable
06:40kjackal has left IRC (kjackal!~quassel@2a02:587:793e:5d00:cd36:ac7a:accf:1899, Ping timeout: 258 seconds)
07:15ricotz has joined IRC (ricotz!~ricotz@ubuntu/member/ricotz)
07:24
<alkisg>
vagrantc: were there any discussions on debian dropping the i386 arch? Buster still supports it for all DEs, right?
07:24
E.g. ubuntu-mate doesn't produce i386 CDs anymore; so I guess debian could be an alternative to people still using thin clients in the future
07:25
<vagrantc>
alkisg: technically, it's i686, but haven't heard any recent discussions about dropping it any time soon
07:25
<alkisg>
Great
07:25
Yeah sure, not real i386, as long as it runs on pentium 4's...
07:25
<vagrantc>
yeah, but won't run on fake pentium 4's
07:25
:)
07:25
<alkisg>
:D
07:26
<vagrantc>
maybe it still suppoirts i586, actually... i forget ... at least twice the baseline went up and nobody noticed for a few years
07:26
e.g. too late to really do anything
07:27
<alkisg>
I only saw a couple of 586-compatible PCs around here, between thousands... so I don't really care about <i686
07:27
Now if only there was a browser capable of surfing with 1 GB RAM... :P
07:30
I'd like to test my new laptop, with i7-8750h = 12480 passmark, in a classroom of thin clients... at just 750€, it should make a decent server
07:32
<vagrantc>
how much ram, how many cores?
07:32
<alkisg>
6 real cores, 12 logical
07:32
just 8 gb ram, but it takes up to 24
07:33
I has this optimus thing with double intel/nvidia, but fortunately it's easy to completely disable nvidia from the uefi settings
07:34
<vagrantc>
multiple video cards?
07:34
<alkisg>
Yes, dynamically switchable
07:34
<vagrantc>
whoah
07:34
<alkisg>
I.e. when one plays games, nvidia is used; when he closes the game, intel is used
07:34
without changing any cables or anything at all
07:43vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Quit: leaving)
09:07ltsp has joined IRC (ltsp!bot@ltsp.org)
10:06kjackal has left IRC (kjackal!~quassel@2a02:587:793e:5d00:cd36:ac7a:accf:1899, Ping timeout: 276 seconds)
10:09kjackal has joined IRC (kjackal!~quassel@ppp-2-84-244-151.home.otenet.gr)
11:56bcg has left IRC (bcg!~b@82-128-233-166.bb.dnainternet.fi, Ping timeout: 255 seconds)
11:58bcg has joined IRC (bcg!~b@82-128-233-166.bb.dnainternet.fi)
12:55
<mwalters>
re: browser... what about palemoon? It's a fork of firefox from way back
12:55
standard browser on manjaro-i3
12:55
(probably standard on their other DE packages, as well)
12:58
https://www.palemoon.org/
12:59
It seems perfectly usable... but we're a gSuite shop, so I have chromium installed on my laptop
13:07
<alkisg>
mwalters: it's not really the browsers that are the problem nowadays, but the web itself
13:08
HTML5, multimedia, javascript... they need modern cpu and lots of ram
13:13
<mwalters>
for sure
13:13
(I used to be a frontend dev :) )
13:13
well.. "full stack"
13:13
beyond some really complicated ETL type stuff on the backend... the frontend is what kicked my butt most of the time
13:15
It's amazing how far frontend development has come from trying to figure out tables on geocities/tripod ;)
13:15
...and how needlessly complicated it's become
13:18
<alkisg>
True; to me it seemed like ...visual basic developers wanted to reinvent the forms/windows controls etc, and didn't like the windows api or gtk or qt, and decided to reimplement them all in a slow untyped interpreted language
13:18
I don't know yet why they think that javascript is so superior to visual basic :P Pascal ftw!
13:19
<mwalters>
HAH! I did *some* pascal back in the BBS days, hacking on WWIV
13:20
As a language, I do actually really like javascript... It definitely "gives you enough rope", but it can be written extremely elegantly if you're so inclined
13:21
A good middleground for someone like me who doesn't want to be bothered learning a "hardcore" functional language, but still wants the expressiveness in a "c styled" language
13:21
I've been digging golang more and more these days for systems stuff, though... nodejs is just tooooooooo sloooooooow
13:22
and electron is a hot mess
13:22
<alkisg>
I didn't see anyone inventing "pythoncoffeescript"; why was there need for coffeescript?
13:22
<mwalters>
there wasn't
13:22
people just didn't want to learn javascript ;)
13:22
<alkisg>
Haha
13:22
I think they needed a more typed language
13:22
And, something more cross-browser
13:22
<mwalters>
Typescript on the other hand... typescript is pretty great. The extra functionality it brought to IDEs and the "compiler" checks were great for my productivity
13:23
<alkisg>
In 1991, I could set up breakpoints "when this variable is read", and land on the code and step through it etc
13:23
I don't think I can do that now with javascript
13:23
<mwalters>
Firefox and V8 are pretty comparable these days... it's been a LONG time since I ran into browser inconsistencies with javascript
13:24
You still can
13:24
<alkisg>
Which environment allows you to stop execution and debug, when a[8] is accessed?
13:24
<mwalters>
Firefox/Chrome support source mapping even... so if you minify/package, you upload the original sources too and you can work in those
13:24
breakpoints, variable watching, the whole 9 yards
13:25
press f12 next time you're in chrome ;)
13:25
the dev tools for javascript are pretty great
13:25
you can even connect chrome to a nodejs process and debug server side code with the chrome dev tools
13:25
I found javascript/node way easier to debug than python
13:26
<alkisg>
breakpoints = source lines
13:26
variable watching = when you step
13:26
I'm talking about data breakpoints
13:26
I haven't seen that
13:26
<mwalters>
oh, break if a variable is a certain value?
13:26
<alkisg>
Right
13:26
Or when it's read
13:26
<mwalters>
read, probably not
13:26
<alkisg>
So it's funny that it was possible 30 years ago, and it's not possible now, with all those advancements
13:27
<mwalters>
I'm pretty sure you can do "conditional breakpoints", but I haven't had a need to
13:27
<alkisg>
It's just an example; I'm missing a whole lot more stuff
13:27
<mwalters>
it's probably something with the memory management in the JIT compiler
13:27
<alkisg>
And of course the worse part is that it's slower than what pascal in 30 year old pcs
13:28
*pascal was in...
13:28
I could do bottomless list boxes with millions of lines instantly
13:28
I can't do that now, neither in javascript nor in gtk
13:28
<mwalters>
you can do "watch expressions" with the chrome dev tools... I'm not sure you can watch an address space to test if its been read, though
13:29
I'm sure the way around that is to do OOP (*shudder*) and watch getter methods ;)
13:30
<alkisg>
Hehe... if you have control over all the stack, sure, but if you just got into a project of others with 100.000 lines of code, good luck
13:30
<mwalters>
absolutely.
13:31
<alkisg>
What I do like nowadays though is that a plain console.log(object) gives you the whole object in an extendable view
13:31
<mwalters>
ah, you need to use the chrome dev tools ;)
13:31
stack exploration when you break
13:31
<alkisg>
I have... I had to do a bit of web development last year
13:31
<mwalters>
expand the entire thing
13:31
<alkisg>
And except for internet explorer support, it went rather well
13:32
<mwalters>
memory use explodes... it's fun ;)
13:32
<alkisg>
#($*&(#* oh my why doesn't this IE read svg.innerHTML... ok, let's not support IE, end of story :P
13:32
And this CORS thing was rather painsome
13:33
<mwalters>
A couple jobs ago I was building a business management system for a postal carrier here... Angular2 frontend, mostly nodejs backend... it was pretty elegant
13:33
Moving a bunch of processes from... Access databases ;)
13:33
<alkisg>
I started with canvasjs, as that was what adobe/microsoft recommended
13:33
3 months later, I had to rewrite everything with plain html5
13:34
<mwalters>
oic... frontend was a hot mess like 8-10 years ago
13:34
now-a-days you just have 900 different frameworks and css layout libraries
13:34
but for the most part they all "just work" as they describe on the package
13:34
<alkisg>
And you need 900 days to learn all that :D
13:34
<mwalters>
not like 10 years ago where you had to do a bunch browser feature checking and stuff
13:35
<alkisg>
Example, for kindergarden kids learning to move the mouse: https://ermis.dschool.edu.gr/preschool/erase/
13:35
<mwalters>
oh, this is super cool
13:36
<alkisg>
Just keeping it maximized with standard aspect ratio needed a week, to make it cross-browser
13:36
<mwalters>
My wife is what we call an "Instructional technology resource teacher" working in elementary schools here
13:36
<alkisg>
Heh, great; I'm a teacher myself, although in secondary schools
13:36
<mwalters>
Ah ok. I worked for Pearson/Powerschool a few years back
13:36
and a local district
13:37
moved on to for-profit business, got screwed a couple times, now I'm back in the not for profit space ;)
13:37
<alkisg>
:D
13:38
<mwalters>
multimedia like this in html is a "hard problem", yes
13:38
I think a couple libraries have come out since flash started it's death throes, though
13:38
(for better or worse)
13:38
<alkisg>
Oh I tried like 10 libraries
13:39
Each had its pros and cons
13:39
<mwalters>
you shouldn't *need* a library... but you know ;)
13:39
absolutely
13:39
<alkisg>
In the end, I used plain HTML5, and it worked better than all the others
13:39
Essentially I developed my own little library
13:39
<mwalters>
can I share this? is there grade-one/first year, etc?
13:40
<alkisg>
After that, the next activities were developed very fast
13:40
This is the development space: https://ermis.dschool.edu.gr/preschool/
13:40
You can find a few more
13:40
After they're developed, they move to http://photodentro.edu.gr, e.g. http://photodentro.edu.gr/photodentro/pattern_v1.0_pidx0054967/
13:40
<mwalters>
I don't think she'll expect the kids to learn greek, but that mouse one is super cool... I'm sure that'd be useful even in kindergarten in some cases
13:41
<alkisg>
Only one of them requires reading (by the teacher)
13:41
It's kindergarden, so students can't read
13:41
<mwalters>
I'll tell *her* to learn Greek, then ;)
13:41
<alkisg>
Some were inspired by gcompris, e.g. https://ermis.dschool.edu.gr/preschool/mosaic/
13:41
<mwalters>
chromium translate seems to work pretty well, though
13:42bcg has left IRC (bcg!~b@82-128-233-166.bb.dnainternet.fi, Quit: bcg)
13:42
<mwalters>
thanks, I'm sure she'll enjoy digging around!
13:43
<alkisg>
np
14:05kjackal has left IRC (kjackal!~quassel@ppp-2-84-244-151.home.otenet.gr, Ping timeout: 245 seconds)
14:11kjackal has joined IRC (kjackal!~quassel@ppp-2-84-244-151.home.otenet.gr)
14:25kjackal has left IRC (kjackal!~quassel@ppp-2-84-244-151.home.otenet.gr, Ping timeout: 268 seconds)
14:42bcg has joined IRC (bcg!~b@dfx4btyyyyyyyyyyyyyyt-3.rev.dnainternet.fi)
15:39vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc)
16:03bcg has left IRC (bcg!~b@dfx4btyyyyyyyyyyyyyyt-3.rev.dnainternet.fi, Quit: bcg)
16:06bcg has joined IRC (bcg!~b@dfx4btyyyyyyyyyyyyyyt-3.rev.dnainternet.fi)
17:08
<mwalters>
Welp, I'm giving up on my bash script for passwords... too many gotchas and I don't have the brainpower to think through all of them :D
17:09
I guess I'll give FreeIPA another look.
17:11vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Quit: leaving)
17:57
<alkisg>
mwalters: meh, just let them change it locally, then sync shadow with the server :)
17:58
ah ldm, never mind
17:58
<mwalters>
I also want to keep it in sync with 4 ltsp servers
18:01vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc)
18:01
<quinox>
keeping passwords in sync between servers... if only there was some kind of authorative server that could be used for that ... ;)
18:10
for me it works great
18:12
<mwalters>
no more servers D:
18:12
<quinox>
of course you're introducting a new dependency, but for me OpenLDAP has been setup-and-forget; I have no experience with FreeIPA but it looks solid as well
18:21
it's servers all the way down
18:22kjackal has joined IRC (kjackal!~quassel@ppp-2-84-244-151.home.otenet.gr)
18:22
<quinox>
then you're making new server to control your other servers, and a server to monitor all your servers, and then you need another server to monitor your monitor server
18:22
<vagrantc>
such a rudimentary service!
18:33pppingme has left IRC (pppingme!~pppingme@unaffiliated/pppingme, Excess Flood)
18:34pppingme has joined IRC (pppingme!~pppingme@unaffiliated/pppingme)
18:55kjackal has left IRC (kjackal!~quassel@ppp-2-84-244-151.home.otenet.gr, Ping timeout: 246 seconds)
19:06GodFather has left IRC (GodFather!~rcc@143.59.184.72, Remote host closed the connection)
19:31woernie has left IRC (woernie!~werner@p5B296A22.dip0.t-ipconnect.de, Remote host closed the connection)
20:45ricotz has left IRC (ricotz!~ricotz@ubuntu/member/ricotz, Remote host closed the connection)
23:47mwalters has left IRC (mwalters!~ubox@c-73-152-61-86.hsd1.va.comcast.net, Ping timeout: 246 seconds)
23:47mwalters has joined IRC (mwalters!~ubox@c-73-152-61-86.hsd1.va.comcast.net)