IRC chat logs for #ltsp on irc.libera.chat (webchat)
Channel log from 20 March 2020 (all times are UTC)
| 00:16 | GodFather has left IRC (GodFather!~rcc@d53-64-7-141.nap.wideopenwest.com, Ping timeout: 250 seconds) | |
| 00:48 | <map7> alkisg & markit I've now installed Debian 10 & 11 localling on the same fat-client and tested booting
| |
| 00:49 | Debian 10 5400RPM drive Grub -> login 43seconds with 11 seconds blank screen
| |
| 00:49 | Debian 11 5400RPM drive Grub -> login 52 seconds with 20 seconds blank screen
| |
| 00:50 | Debian 11 PXE boot iPXE -> login 57 seconds with 27 seconds blank screen
| |
| 00:50 | Debian 11 PXE boot with dd caching iPXE -> login 41 seconds with 6 seconds blank screen
| |
| 00:51 | So alkisg you are right it does show similar results to a rotational drive.
| |
| 00:52 | weird that Debian 11 is much slower but then again it's "testing" so maybe more logging or less compression is happening in pacakges.
| |
| 01:12 | For completeness I also built a Debian10 with LTSP 20.3 and did these tests
| |
| 01:13 | Debian 10 PXE boot iPXE -> login 29 seconds with 10 seconds blank screen
| |
| 01:13 | Debian 10 PXE boot with dd caching iPXE -> login 28 seconds with 3 seconds blank screen
| |
| 03:38 | <markit> hi map7
| |
| 03:40 | I've tried to boot 4 client VM, and initrd.img decompression is taking 4x the time, so 45 seconds! Wondering what will happen with 24 clients turned on almost at the same time
| |
| 03:42 | don't know if removing dnsmasq and using tftpd-hpa can alleviate the problem
| |
| 03:43 | <map7> hi markit
| |
| 03:44 | <markit> 4.45 am here :(
| |
| 03:45 | <map7> 14:45 here
| |
| 03:47 | Well I only have 8 fat-clients to boot so I won't have the same problems
| |
| 03:47 | and I have a 10Gbps link to the server
| |
| 03:48 | all the clients are 1Gbps to the 10Gbps link to the server
| |
| 03:49 | markit I'm just going to install Debian 10 with the dd cache and do a chrootless image
| |
| 03:49 | I'm happy with the speed and with my setup this should be fine.
| |
| 03:50 | <markit> map7: oh, I see, you have a very different user case than mine (school lab)
| |
| 03:50 | <map7> I will have to upgrade memory on two of the machines but that should be easy
| |
| 03:50 | yeah I manage a multi-tennant office (4 x companies on the one level)
| |
| 03:51 | all running LTSP5 thin-clients at the moment, moving to the fat-clients very soon
| |
| 03:51 | so our thin-clients stay on for weeks sometimes months and don't get rebooted
| |
| 03:52 | <markit> maybe I have to ask the school if has budget for a 10Gb uplink switch, but I've 3 labs so would cost a lot (switch + lan adapter )
| |
| 03:52 | <map7> that's why I'm not too worried about the boot speed, but I can see why you are.
| |
| 03:52 | <markit> yep, also small children start screaming "computer is broken" if does not start fast :)
| |
| 03:55 | <map7> hahaha, we don't want children screaming at computers
| |
| 03:55 | or anything really
| |
| 03:56 | I found debian 10 booted much quicker than debian 11 "testing"
| |
| 04:14 | <markit> tftpd-hpa creates 4 processes instead of the single dnsmasq one, but does not seem to speed things up... maybe is a lan bandwidth limit
| |
| 04:14 | I'm very scared :|
| |
| 04:14 | vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc) | |
| 04:15 | <markit> sleep time, see you, good luck with your setup :)
| |
| 04:15 | markit has left IRC (markit!~marco@mail.ammdomus.it, ) | |
| 04:26 | <map7> I'm just testing ltsp 20.3 using a chroot image for the first time and having trouble booting
| |
| 04:27 | I get the error 'mount: No such file or dir' and then I'm left at the initramfs prompt
| |
| 04:30 | Here is what I've done on a fresh install of debian 10: https://gist.github.com/map7/bdcef100f7f9ecba38a439fed12bea53
| |
| 05:51 | <alkisg> map7: can you get a picture of that error screen, so that we also see which ltsp command failed?
| |
| 05:51 | And put it to imgur.com or somewhere
| |
| 05:54 | "Why not use the root image? Because it's 180GB!" ==> maybe you have /data and you didn't put it to ltsp image.excludes, to omit it from the generated image?
| |
| 05:54 | LTSP images rarely are larger than 1-5 GB
| |
| 05:55 | map7: also, you forgot to install ltsp in the chroot
| |
| 06:09 | Finally, it's best to use a VM that you install graphically, not a chroot, as debian packages sometimes do not declare all of their dependencies :)
| |
| 06:59 | adrianorg has joined IRC (adrianorg!~adrianorg@187.115.108.32) | |
| 07:02 | adrianor1 has left IRC (adrianor1!~adrianorg@177.204.159.122.dynamic.adsl.gvt.net.br, Ping timeout: 256 seconds) | |
| 07:29 | vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Quit: leaving) | |
| 07:31 | shored has left IRC (shored!~shored@87-92-92-55.bb.dnainternet.fi, Read error: Connection reset by peer) | |
| 07:35 | shored has joined IRC (shored!~shored@87-92-92-55.bb.dnainternet.fi) | |
| 07:48 | woernie has joined IRC (woernie!~werner@pD9E8BE9E.dip0.t-ipconnect.de) | |
| 08:23 | shored has left IRC (shored!~shored@87-92-92-55.bb.dnainternet.fi, Read error: Connection reset by peer) | |
| 08:24 | shored has joined IRC (shored!~shored@87-92-92-55.bb.dnainternet.fi) | |
| 10:51 | markit has joined IRC (markit!~marco@mail.ammdomus.it) | |
| 11:10 | <markit> alkisg: hi, I've tried 4 client boot, and really is 4x slower than a 1 client boot, wondering how could it work with 24 clients. Do you think that I do really need a server with 10Gb uplink to the switch? Do you already have tested ltsp 20 in a real lab?
| |
| 11:10 | <alkisg> markit: yes, I have many schools using ltsp 20 both in 18.04 and 20.04
| |
| 11:11 | The speed compared to ltsp5 is better, not worse
| |
| 11:11 | So if you already used ltsp5, it should be 10% faster now
| |
| 11:12 | The server usually has 1 gbps and the clients either 1 gbps or 100 mbps
| |
| 11:12 | I also have 10 schools with 2 gbps server, bond
| |
| 11:12 | <markit> I've seen dnsmasq during tftp providing kernel images is single thread, while tftp-hp opens a thread for each client
| |
| 11:13 | but performances look not much improved, maybe with 24 clients they are
| |
| 11:13 | <alkisg> That's not a problem. One is using threads, the other "select", they should perform the same
| |
| 11:13 | <markit> (compared with dnsmasq)
| |
| 11:13 | <alkisg> Are you using PXE on the clients, or iPXE?
| |
| 11:13 | Some PXE implementations for tftp clients are a bit slow, and then switching to http will be faster
| |
| 11:14 | <markit> in real clients they have "on board" PXE, VM instead use (default of proxmox) iPXE
| |
| 11:14 | <alkisg> You said "decompressing initrd" is slow
| |
| 11:14 | I guess you mean "fetching initrd is slow"
| |
| 11:14 | As decompression should happen in a couple of seconds in recent CPUs
| |
| 11:14 | <markit> does it fetch and THEN decompress? I see the % completion
| |
| 11:14 | <alkisg> Yes, that % is fetching
| |
| 11:15 | While decompressing is done by the kernel later on
| |
| 11:15 | It's not done by ipxe
| |
| 11:15 | In many cases, TFTP works at 100 mbps
| |
| 11:15 | Which means that booting 10 clients will make the server send 1 gbps
| |
| 11:15 | <markit> so when I read /ltsp/x86_64/initrd.img... 30% is 30% fetching? sigh
| |
| 11:16 | <alkisg> If that time is acceptable, end of story; otherwise you could try fetching initrd via http
| |
| 11:16 | Yes, fetching 80 MB over 100 mbps link is slow :)
| |
| 11:16 | But with tftp is even slower
| |
| 11:16 | Time it and tell me how much it needs
| |
| 11:18 | <markit> I've setup tftp-hp now, let me try with it
| |
| 11:19 | <alkisg> There's no real difference, it just wastes more RAM
| |
| 11:19 | <markit> ok, with the "virtual" 2.5Gb/s nic, it's 8 seconds
| |
| 11:20 | let me try with 4 clients, just to see if scales linear (I was really tired yesterday night better re-check my tests)
| |
| 11:20 | <alkisg> Virtual doesn't count though. Try with real clients.
| |
| 11:21 | (at least the virtual shouldn't be on the same server)
| |
| 11:21 | <markit> I can't easely, we have "corona virus" and can't go around (or better not to)
| |
| 11:23 | <alkisg> Are you testing with 4 virtual clients?
| |
| 11:23 | <markit> yes
| |
| 11:24 | <alkisg> All using the same physical NIC?
| |
| 11:24 | This then devices their bandwidth
| |
| 11:24 | <markit> yes, virual bridge
| |
| 11:24 | <alkisg> So the "boot speed for 4 clients is 4 times less", is inaccurate
| |
| 11:24 | As the real clients would be 1 gbps, not 1 gpbs div 4
| |
| 11:24 | <markit> if I test with epoptes, all clients transfer at 800Mb/s
| |
| 11:25 | <alkisg> Normally, if you boot a client in 40 secs, and then you boot 10 clients, they should boot in 50 secs
| |
| 11:25 | <markit> now initrd.image completed after 58"
| |
| 11:25 | <alkisg> That's not a usual timing
| |
| 11:25 | <markit> booting also seem slow like hell... that's strange indeed
| |
| 11:25 | <alkisg> Booting 12 clients here needs less than a minute, including the initrd
| |
| 11:25 | With 1 gpbs on the server
| |
| 11:26 | It's a problem in your test emulation
| |
| 11:26 | <markit> oh, so happy to hear it, thanks a lot
| |
| 11:26 | <alkisg> There *are* NICs that are very slow in tftp
| |
| 11:26 | If you do have such NICs (you'll only be able to tell with real testing), then you might prefer http instead of tftp
| |
| 11:28 | <markit> does dnsmasq or tftp-hp switch automatically to http or you told me before but I missed it?
| |
| 11:28 | <alkisg> HTTP booting needs manual configuration, and a web server
| |
| 11:28 | It's not automated at all
| |
| 11:29 | <markit> ok, is there a config file of ltsp where I can set additional ltsp image parameters? like the "-comp" or "--ionice" ones
| |
| 11:29 | <alkisg> Yes, ltsp.conf
| |
| 11:29 | <markit> is it documented? I read the doc but was not able to find it, let me check again
| |
| 11:29 | <alkisg> https://github.com/ltsp/community/issues/85#issuecomment-576326061
| |
| 11:30 | man ltsp image
| |
| 11:30 | Gives you the --mksquashfs-params documentation
| |
| 11:30 | And then you can put any parameter in ltsp.conf, under [server]
| |
| 11:30 | !ltsp-image
| |
| 11:30 | <ltspbot> ltsp-image: Generate a squashfs image from an image source: https://ltsp.org/man/ltsp-image
| |
| 11:31 | <markit> was reading https://ltsp.org/man/ltsp.conf/
| |
| 11:31 | <alkisg> The command parameters are documented in ltsp commands
| |
| 11:32 | https://ltsp.org/man/ltsp/#files
| |
| 11:32 | "All the long options can also be specified as variables in the ltsp.conf configuration file in UPPER_CASE, using underscores instead of hyphens."
| |
| 11:34 | <markit> it's easy to find only once you know it
| |
| 11:34 | <alkisg> ltsp sysadmins are supposed to read ALL the man pages eventually :)
| |
| 11:34 | I spent a lot of days writing them! :P
| |
| 11:35 | Lunch time, later...
| |
| 11:36 | <markit> read <> understand... when you read https://ltsp.org/man/ltsp/#files you are focused on something else, then you look for compression, find that ltsp image can have additional parameters, and then look at ltsp.conf documentation to check if there is one suitable
| |
| 11:36 | alkisg: thanks a lot :)
| |
| 12:21 | Teridon has joined IRC (Teridon!~Teridon@dragon.teridon.com) | |
| 12:24 | <markit> alkisg: shame on me, I've found the problem... long time ago I tested with 16.04 in Proxmox, but now I've some "heavy cpu intensive" VM running, so with 4 clients boot the CPU is saturated and each vm can't run at max speed
| |
| 12:27 | <alkisg> markit: yeah usually virtual nics in localhost are only limited by cpu, not real bandwidth, e.g. I get more than 5 gbps here, even though I only have gigabit nic
| |
| 13:27 | NecTal has joined IRC (NecTal!c27e6a95@194.126.106.149) | |
| 13:29 | NecTal has left IRC (NecTal!c27e6a95@194.126.106.149, Remote host closed the connection) | |
| 13:44 | GodFather has joined IRC (GodFather!~rcc@d53-64-7-141.nap.wideopenwest.com) | |
| 13:45 | NecTal has joined IRC (NecTal!c27e6a95@194.126.106.149) | |
| 13:46 | <NecTal> Hello.
| |
| 13:46 | Does ltsp initrd take users and groups that i see with getent passwd and getent group?
| |
| 13:49 | I'm asking because i joined ltsp server to my domain to get users from there. I see all users with getent passwd and i'm able to login over ssh with users but when i try to login with thinclient it says "Invalid passwd/shadow for user"
| |
| 13:50 | Mapped users with sssd.
| |
| 14:04 | <alkisg> NecTal: which ltsp version, 5 or the new one?
| |
| 14:04 | Ah, nvm, ltsp initrd
| |
| 14:04 | So, the new ltsp actually puts your passwd and group in initrd
| |
| 14:05 | If you want to use ldap etc, you need to configure the image to use it. The accounts need to "exist" in the image, only the authentication happens via ssh
| |
| 14:06 | <NecTal> Hmm okay.
| |
| 14:06 | <alkisg> There are many discussion topics regarding ldap
| |
| 14:06 | !community-issues
| |
| 14:06 | <ltspbot> community-issues: In https://github.com/ltsp/community/issues you may ask (or answer) LTSP19+ related questions. Use that instead of mailing lists/forums/stackexchange/launchpad questions; LTSP developers also participate in community issues
| |
| 14:06 | <alkisg> !ltsp-issues
| |
| 14:06 | <ltspbot> ltsp-issues: https://github.com/ltsp/ltsp/issues is where you file bug reports or feature requests that can be addressed from the LTSP code base. See also !community-issues for more general LTSP questions or discussions
| |
| 14:06 | <alkisg> Also there, and in the wiki
| |
| 14:06 | !wiki
| |
| 14:06 | <ltspbot> wiki: The LTSP community wiki can be found at https://github.com/ltsp/community/wiki
| |
| 14:07 | <NecTal> Okay. Very good. I will take a look around topics.
| |
| 14:07 | Thank you.
| |
| 14:07 | <alkisg> np
| |
| 14:07 | NecTal has left IRC (NecTal!c27e6a95@194.126.106.149, Remote host closed the connection) | |
| 14:59 | vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc) | |
| 15:17 | GodFather has left IRC (GodFather!~rcc@d53-64-7-141.nap.wideopenwest.com, Ping timeout: 250 seconds) | |
| 15:24 | vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Quit: leaving) | |
| 15:29 | vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc) | |
| 15:33 | jgee has left IRC (jgee!~jgee@190.159.118.121, Quit: Ping timeout (120 seconds)) | |
| 15:33 | jgee has joined IRC (jgee!~jgee@190.159.118.121) | |
| 15:40 | <markit> alkisg: "ADD_IMAGE_EXCLUDES="/etc/ltsp/add-image.excludes"" does excluted THAT specific file, or reads that file and uses it's content?
| |
| 15:42 | if the former, an example like ="/etc/ltsp/myfile" would be less confusing
| |
| 15:43 | <alkisg> !ltsp.conf
| |
| 15:43 | <ltspbot> ltsp.conf: Configuration file for LTSP: https://ltsp.org/man/ltsp.conf/
| |
| 15:43 | <alkisg> "They can either be filenames or multiline text."
| |
| 15:43 | I.e. if it's a single file that exists, its contents are read
| |
| 15:44 | If there are multiple files or one file that doesn't exist, it's multiline text
| |
| 15:44 | test -f "$ADD_IMAGE_EXCLUDES" => it's a file, otherwise it's text
| |
| 15:48 | <markit> alkisg: so if I want to exclude a single particular file?
| |
| 15:49 | <alkisg> If that file exists, you need to either create /etc/ltsp/add-image.excludes and put it in a line there, or use a wildcard
| |
| 15:49 | e.g. ADD_IMAGE_EXCLUDES="/bin/tr?e" or /bin/true* etc etc
| |
| 15:50 | actually excludes are relative
| |
| 15:50 | So without the initial /, I doubt it would exist anyway
| |
| 15:50 | So ADD_IMAGE_EXCLUDES="bin/true"
| |
| 15:52 | <markit> still confused "So without the initial /, I doubt it would exist anyway"
| |
| 15:53 | ADD_IMAGE_EXCLUDES="bin/true" will exclude, since I do ltsp image /, the existing file /bin/true ?
| |
| 15:54 | <alkisg> Yes
| |
| 15:54 | <markit> "I doubt it would exist anyway" where? in what position?
| |
| 15:55 | <alkisg> Where you run the ltsp command
| |
| 15:55 | markit, it's just a "test -f"
| |
| 15:55 | Try it, run test -f "whatever" && echo exists
| |
| 15:55 | <markit> ok, I see now what you meant
| |
| 15:57 | btw, virtualization server unoticed CPU saturation explains why ltsp server and clients seemd to be never saturated but still slow. I wasted tons of hours in 2 days just trying to figure out what was so wrong in ltsp :(
| |
| 15:58 | Teridon1 has joined IRC (Teridon1!~Teridon@dragon.teridon.com) | |
| 16:01 | <alkisg> So now everything's OK in your LTSP setup?
| |
| 16:01 | Teridon has left IRC (Teridon!~Teridon@dragon.teridon.com, Ping timeout: 256 seconds) | |
| 16:09 | <markit> alkisg: I think so, I need real hardware to test, and re-test my account management scripts (and add "ltsp initrd" every time I create / delete accounts)
| |
| 16:10 | I've seen you exclude dirs related to lightdm so I'm looking for similar sddm related dirs
| |
| 16:11 | THEN I do really hope there isn't some nasty bug related to KDE that ruins all that
| |
| 16:13 | <alkisg> I like that KDE doesn't use any snaps by default :)
| |
| 16:14 | It makes it even lighter than the default ubuntu mate
| |
| 16:14 | Although I purge snapd right after installation :D
| |
| 16:14 | <markit> is canonical polluting desktops with snaps? I thought were only intended for local user's installations
| |
| 16:15 | <alkisg> They even stopped offering chromium unless you install snap
| |
| 16:15 | <markit> apt install chromium does not work?
| |
| 16:15 | <alkisg> If you purge snapd, then go to software center to install chromium, they install snap for you
| |
| 16:15 | No, it installs snap too :/
| |
| 16:15 | <markit> argh! But maybe for "security reasons" :)
| |
| 16:16 | <alkisg> markit, try to vote for this and give publicity: https://bugs.launchpad.net/ubuntu/+source/chromium/+bug/1855594
| |
| 16:16 | No packages in "universe" have security updates
| |
| 16:16 | chromium is in universe
| |
| 16:16 | Debian manages to keep chromium up to date just fine
| |
| 16:17 | They could just sync the debian version
| |
| 16:17 | <markit> yep but AFAIK, snapd are "confined", jailed apps
| |
| 16:17 | snaps
| |
| 16:18 | btw, I've teachers that are using LTSP lab only for Google Suite access :(
| |
| 16:18 | <alkisg> I don't think they're able to jail big apps like chromium
| |
| 16:18 | chromium does jail its tabs though :)
| |
| 16:18 | <markit> so the "FOSS everywhere" battle I fought seems lost to me
| |
| 16:19 | <alkisg> Commercial software does have a right to exist :)
| |
| 16:19 | <markit> I install tons of useful apps in ltsp, but they simple ignore them
| |
| 16:19 | yep, but you shuold teach children foss, not proprietary programs dependency OMHO
| |
| 16:20 | <alkisg> You can't really expect foss to drive evolution, it's usually what comes after companies introduce evolution
| |
| 16:20 | <markit> and in any case, now the school is becoming totally dependant upon Google
| |
| 16:21 | you can't teach children "you can't learn how things work, or built or modify yourself what you are using, or keep your own data"
| |
| 16:22 | and FOSS can lead evolution, is just a matter of investments
| |
| 16:23 | i.e. schools are wasting tons of money for proprietary software that US company owns, while our programmers don't find jobs here in italy.
| |
| 16:23 | <alkisg> Invention needs profit. Open inventions get stolen too quickly to be profitable; a period must exist when they should be protected, either with patents or with closed source.
| |
| 16:24 | <markit> you can change this with public intervention or more citizen awareness
| |
| 16:24 | <alkisg> I don't think Italian or Greek programmers can construct Google. It's possible to do some things though, but not very widescale
| |
| 16:25 | So, we can give foss in some places, but also use commercial apps where it makes sense
| |
| 16:25 | <markit> its foss vs proprietary, not foss vs commercial, if source is available is fine
| |
| 16:26 | vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Quit: leaving) | |
| 16:26 | <markit> it's a complicated business model, that's why I think we need legislation and public money to make it work
| |
| 16:26 | but would be better for all of us
| |
| 16:30 | <alkisg> I haven't seen any government-funded project produce anything of exceptional quality here in Greece
| |
| 16:31 | vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc) | |
| 16:31 | <alkisg> Even the tax sites barely work
| |
| 16:33 | vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Client Quit) | |
| 16:49 | Teridon1 has left IRC (Teridon1!~Teridon@dragon.teridon.com, Remote host closed the connection) | |
| 16:49 | <markit> that's why public sector is not working well enough, but the solution is make it work, not relay on private dependency in fundamental, strategic or really public useful things. It's madness
| |
| 16:49 | <fiesh> in an ideal world, all public money spent on software would be *required* to only go to FOSS
| |
| 16:49 | <markit> public is not working -> fix public, NOT -> surrende to private
| |
| 16:49 | <fiesh> that would change a lot, and it makes perfect money that the benefits of public money are publicly available
| |
| 16:49 | <markit> fiesh: public money, public code :)
| |
| 16:50 | <fiesh> but there's too little understanding and too much lobbying for this to happen
| |
| 16:50 | vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc) | |
| 16:50 | <markit> fiesh: yes, "bad guys" have a lot of money to promote their interests, "good citizens" don't know (that's why I think school should teach it) and don't care
| |
| 16:51 | fiesh: a funny thing is how many complain here because Autocad costs A LOT, but no one is founding FreeCad 3d project
| |
| 16:52 | they don't know, or look at it and say "is not good enough", or if is good enough use it but don't give money back... like for LibreOffice
| |
| 16:52 | but in any case, "Cloud" is removing computation power from our hands and put in few, bug ones that will control everything
| |
| 16:55 | <fiesh> markit: alas especially with things like CAD, there is no good free alternative
| |
| 16:55 | markit: opencascade, the geometry kernel underlying FreeCAD, is very very poor compared to the commercial solutions
| |
| 16:55 | <markit> fiesh: also has a nasty license
| |
| 16:56 | <fiesh> markit: but again, a lot of public money is going into CAD systems, if that had been directed differently, we would have a good open source geometry kernel by now
| |
| 16:56 | <markit> fiesh: but freecad started some years ago, with few programmers (I make a yearly donation of few euros to Yorik), I know people spend SOME THOUSAND of license every year for Autocad
| |
| 16:57 | if they had donated 1/10 of that sum to Freecad since was started, you would have a very good program, I'm sure
| |
| 16:57 | i.e. Blender is well founded -> great program
| |
| 16:58 | also DWG proprietary format is a problem, that's why OPEN format should be required always, but here public sectors are ok with .dwg, doc or docx (false open format) or whatever
| |
| 16:59 | fiesh: so also private money could do the difference as well in a lot of situations
| |
| 17:01 | alkisg: any project to have Epoptes soon be improved and tell info about CPU architecture of the clients? Maybe adding a "multi select" client rapport with a spreadsheet generated with clients caracteristics
| |
| 17:01 | just to check easely which clients are 32 bit only
| |
| 17:01 | <alkisg> markit: for example, similar commercial projects to epoptes give their developers thousands of euros per year. I get 0 euros per year for epoptes development
| |
| 17:02 | Sustaining foss apps isn't easy
| |
| 17:03 | So, something big that would use epoptes, maybe a whole country that uses it in its schools, should fund development permanately
| |
| 17:03 | Until then... commercial works :)
| |
| 17:03 | (I mean closed source, as open source would be stealed too easily)
| |
| 17:04 | <markit> alkisg: in fact, your ltsp project and epoptes too is an example
| |
| 17:05 | <alkisg> Right, these see no money at all for developement. Only google sent money for that. The only money ltsp and epoptes get is for *additional* user support
| |
| 17:05 | Not for development, but extra work for helping users
| |
| 17:05 | <markit> I've read i.e. about Libreoffice adoption in some public sector... they same some million euros in license and give ZERO back to the project
| |
| 17:06 | <alkisg> Right. They have no reason to fund. "Let someone else fund it, it's free; if it doesn't sustain itself, we'll find something else"
| |
| 17:06 | A very high level of perspective is required for foss to work in this cases. Until it does... closed source is fine too
| |
| 17:06 | <vagrantc> this is what the snowdrift.coop model is attempting to solve ... not sure it'll ever get off the ground
| |
| 17:07 | <markit> alkisg: they don't understand the FOSS at all. I.e. in Italy there are a lot of people that put effort to help others for free
| |
| 17:08 | but if you talk about "found Gimp", they tell you "why? I crack Photoshop for free!"
| |
| 17:08 | <alkisg> That one is easy; fines for piracy here is 1000 euros per program
| |
| 17:08 | <markit> they don't get it at all, that's why I would like to spread awareness
| |
| 17:09 | <alkisg> A teacher using office, photoshop etc in her school can be fined for 30000 euros easily
| |
| 17:09 | It doesn't actually happen, but it's a good excuse not to risk piracy
| |
| 17:09 | <markit> alkisg: no one I know has been persecuted (maybe some company in the past, none of my "smart" friends has been)
| |
| 17:10 | alkisg: we have schools that relay only on piracy, when I told about it in the past at the "chief" of the school (don't know the english term), seemd that I was the one causing troubles
| |
| 17:11 | <alkisg> Oh my mouth has gotten me to trouble too often. My intentions were good but... "oh lord, please don't let me be misunderstood" fits here :)
| |
| 17:11 | <markit> I've installed LibreOffice and found some days later M$ Office installed in the same pc, where they just do some simple text and spreasheet, and LiBo was in any case overkill for them
| |
| 17:11 | alkisg: lol, nice song
| |
| 17:11 | well, enough ranting today, back to ltsp ;P
| |
| 17:12 | I've to improve my installation and configuration notes, I've such a bad memory I need them
| |
| 17:12 | (there is a lot of copy/paste from this irc channel too)
| |
| 17:13 | alkisg: we need a "no log irc" command when we discuss not techical related stuff... we will be persecuted in the future I'm sure :P
| |
| 17:13 | <alkisg> Hehe
| |
| 17:13 | We can just join #ltsp-offtopic
| |
| 17:29 | <markit> vagrantc: did't know about it, thanks for the info
| |
| 17:31 | vagrantc: trying to check the difference with Patreon / Liberapay, maybe this one is only FOSS focused
| |
| 18:10 | <vagrantc> markit: they actually have an interesting model to encourage everybody to contribute as much as they can ...
| |
| 18:10 | markit: it's sort of like a multi-party match
| |
| 18:12 | and yes, definitely FOSS focused
| |
| 18:13 | well, actually, they coined their own acronym as it's beyond software
| |
| 18:13 | give it a read ... interesting stuff
| |
| 18:18 | ok, have audio
| |
| 18:18 | no idea if my mic works, but it isn't complaining about it
| |
| 18:23 | woernie has left IRC (woernie!~werner@pD9E8BE9E.dip0.t-ipconnect.de, Remote host closed the connection) | |
| 18:31 | woernie has joined IRC (woernie!~werner@pD9E8BE9E.dip0.t-ipconnect.de) | |
| 18:43 | GodFather has joined IRC (GodFather!~rcc@d53-64-7-141.nap.wideopenwest.com) | |
| 18:45 | markit has left IRC (markit!~marco@mail.ammdomus.it, Ping timeout: 240 seconds) | |
| 18:45 | markit has joined IRC (markit!~marco@mail.ammdomus.it) | |
| 19:12 | <alkisg> telep4ss
| |
| 19:13 | Eh :D
| |
| 19:13 | At least it's for a discardable VM :D
| |
| 19:23 | GodFather has left IRC (GodFather!~rcc@d53-64-7-141.nap.wideopenwest.com, Ping timeout: 250 seconds) | |
| 19:32 | <markit> alkisg: beside "standard" ltsp.conf, is there available the one you use for your schools? Maybe you have already done stuff I need also, like not having clients with the HPLIP icon on the tray area or stuff like that
| |
| 19:33 | (if I can ask)
| |
| 19:42 | Sleaker has left IRC (Sleaker!quasselcor@172.110.7.65, Quit: http://quassel-irc.org - Chat comfortably. Anywhere.) | |
| 19:56 | <alkisg> markit: this is concentrated in "sch-scripts"
| |
| 19:56 | I tried to internationalize it as "ltsp-manager", but then I couldn't put the greek-specific stuff I wanted, and there was not much general interest
| |
| 19:56 | So apt install sch-scripts, installs ltsp, removes snapd, does a whole lot of settings etc
| |
| 20:23 | sutula has left IRC (sutula!~sutula@184.100.166.160, Read error: Connection reset by peer) | |
| 20:25 | sutula has joined IRC (sutula!~sutula@184.100.166.160) | |
| 20:51 | GodFather has joined IRC (GodFather!~rcc@d53-64-7-141.nap.wideopenwest.com) | |
| 21:25 | GodFather has left IRC (GodFather!~rcc@d53-64-7-141.nap.wideopenwest.com, Ping timeout: 240 seconds) | |
| 21:28 | vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Quit: leaving) | |
| 21:52 | woernie has left IRC (woernie!~werner@pD9E8BE9E.dip0.t-ipconnect.de, Remote host closed the connection) | |
| 22:22 | GodFather has joined IRC (GodFather!~rcc@d53-64-7-141.nap.wideopenwest.com) | |
| 22:58 | GodFather has left IRC (GodFather!~rcc@d53-64-7-141.nap.wideopenwest.com, Ping timeout: 250 seconds) | |
| 23:25 | Klimm has left IRC (Klimm!~Georg@p54897080.dip0.t-ipconnect.de, Ping timeout: 246 seconds) | |
| 23:25 | GKlimm has joined IRC (GKlimm!~Georg@p54897080.dip0.t-ipconnect.de) | |
| 23:36 | alkisg1 has joined IRC (alkisg1!~alkisg@ubuntu/member/alkisg) | |
| 23:37 | alkisg has left IRC (alkisg!~alkisg@ubuntu/member/alkisg, Ping timeout: 246 seconds) | |
| 23:37 | map7 has left IRC (map7!~user@103.232.216.31, Ping timeout: 246 seconds) | |
| 23:37 | map7 has joined IRC (map7!~user@103.232.216.31) | |