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


Channel log from 28 May 2019   (all times are UTC)

00:31adrianorg has joined IRC (adrianorg!~adrianorg@186.213.159.184)
00:34adrianor1 has left IRC (adrianor1!~adrianorg@187.113.251.145, Ping timeout: 258 seconds)
02:27adrianor1 has joined IRC (adrianor1!~adrianorg@179.179.77.59)
02:30adrianorg has left IRC (adrianorg!~adrianorg@186.213.159.184, Ping timeout: 272 seconds)
02:51adrianor1 has left IRC (adrianor1!~adrianorg@179.179.77.59, Ping timeout: 258 seconds)
02:57adrianorg has joined IRC (adrianorg!~adrianorg@177.18.97.153)
03:19
<vagrantc>
alkisg: well, ltsp-build-client is biting again ... options for how to specify an unsigned apt repository have changed. :/
03:20
and now ltsp needs to be updated to handle it...
03:20* vagrantc sighs
03:20
<vagrantc>
alkisg: mostly just more encouragement to keep moving forward! :)
04:19
<alkisg>
Morning all!
04:20
vagrantc: I've been thinking, maybe this will work for both of us? I keep a packaging/debian dir upstream, and a symlink debian -> packaging/debian. Since I'll never touch the symlink, rebasing with upstream should be easy every time you need to do it, right?
04:28
(trying to see what I'll need to do in a couple of days/weeks when I'll need to upload to the PPA again, so that we won't have to go through it yet another time :))
04:31
<vagrantc>
maybe
04:31
makes it a lot harder to cherry-pick your packaging changes, though
04:32
<alkisg>
Should I try to keep the commits to packaging/debian separate from the respective code commits? Will that help?
04:32
(and to the fedora packaging etc)
04:33
<vagrantc>
it just requires manual work to merge, rather than relying on git to handle what git is reasonably good at
04:33
<alkisg>
Why? You can just merge ALL the commits, no need to cherry pick
04:33
You can discard the packaging folder why building
04:34
git merge start..end, where start==the last time you merged
04:34
The debian symlink will never be referenced there
04:34
<vagrantc>
if you make a packaging change i want, then i have to manually merge it, git won't do it for me
04:34
<alkisg>
Sure, but that will be true wherever I maintain the packaging
04:35
It's caused because you don't want a common collaborative debian folder; not by cvs
04:35
<vagrantc>
if you maintained the packaging separately, i could just cherry-pick and/or merge
04:35
<alkisg>
But don't you need to merge both upstream and packaging anyway?
04:35
Why does it help if they are 2 separate repositories?
04:35
*both code...
04:35
<vagrantc>
upstream never conflicts and requires no thought to merge
04:36
<alkisg>
It wouldn't conflict now either, since it'd use a different dir
04:36
<vagrantc>
git merge v5.x -> done!
04:36
<alkisg>
And then after you merge the code, in both cases, you need to cherrypick debian
04:36
No difference there either
04:37
<vagrantc>
yes, but then i would have to look at what packaging changes you made and use the amazing vcs "cp" and loose history.
04:37
<alkisg>
Because I would use a different path?
04:37
<vagrantc>
you do make changes i think are worth merging the vast majority of the time
04:38
<alkisg>
While if it were a separate repository, I'd use debian/, which would be in the same path?
04:38
*using
04:38
<vagrantc>
yes, i can't "git cherry-pick packaging/debian into debian/"
04:38
but because it's git, i can just "git merge the whole branch, possibly.
04:39
<alkisg>
That makes it very hard for upstream to manage packaging though; the history is split between repositories, and each commit can't reference the respective debian or fedora commit
04:39
So upstream history is hindered, just to keep a downstream debian history
04:39
I'd much rather prefer the loss of history using "cp" than to lose the upstream history
04:40
<vagrantc>
this is well known
04:40
<alkisg>
:D
04:41
To be clear, you'd still prefer packaging/debian upstream, vs debian/ folder upstream, right?
04:41
<vagrantc>
not entirely sure.
04:41
which eye would you like the hot poker in?
04:42
i mean, i don't like either option, and you don't like the status quo
04:42
so ... here we are
04:42
<alkisg>
For ltsp5, you've been the main contributor to debian/; of course I'll respect that, and do it your way; I'll use a different repository
04:43
For ltsp19 and epoptes, where I'll be the main contributor to debian/, I don't want to lose the main history in order to keep a downstream one,
04:43
so ok we can postpone that decision until then...
04:43
<vagrantc>
i get that ...
04:44
i wish dgit worked better, as that would solve most of this.
04:44
<alkisg>
So for now I'll cp debian/ into a new repository in launchpad (I don't want the source there too); I'll lose all history; and I'll be maintaining that one
04:45
<vagrantc>
you'd think git could be smart enough to implement a copy operation so history could be preserved
04:46
alkisg: out of all the workflows proposed ... separate debian dir was the most frustrating to me
04:46
but we have different aims
04:47
<alkisg>
But if I'm maintaining the upstream debian/ dir, there's no point in having a temp copy of upstream code too
04:47
I'd have to spend hours to next contributors to explain why all this is done to cover some debian tool inefficiency
04:48
<vagrantc>
what i like about the workflow i've been doing is it's clear what packaging changes are relevent to an upstream release ... but with builds not from tagged releases, that isn't as attractive and it makes sense to merge the packaging changes with the upstream changes
04:48
<alkisg>
I think you're not using to having upstream contributors to the debian/ dir
04:48
It was distro-specific all along
04:48
*not used
04:48
<vagrantc>
i've survived such scenarios and they formed my opinions :P
04:48
<alkisg>
And a solo work too
04:49
E.g. if you had to maintain debian, ubuntu, kali, linuxmint etc releases, your workflow would change
04:49
Now you only focus on debian, while i focus in central maintenance
04:49
<vagrantc>
perhaps
04:50
<alkisg>
You wouldn't be able to have a "frozen debian dir" for months, while you'd need to do releases for other distros
04:51
<vagrantc>
back when ltsp was more active, i did maintain multiple concurrent debian dirs for debian
04:51
<alkisg>
And did you have a "master" one?
04:51
<vagrantc>
and it was so nice once i settled on this workflow
04:51
<alkisg>
Or was "master == sid"?
04:51
<vagrantc>
debian/master is typically sid or experimental, whichever is newest
04:52
<alkisg>
But now you have master=buster, while ppa (considering it another distro temporarily) needs to release, hence the issue
04:52
<vagrantc>
typically fork off a debian/CODENAME branch around the time of the freezes, depending on upsteam release cycles
04:53
<alkisg>
I.e. you don't allow for a collaborative debian branch, where all maintainers can contribute, and distro-maintainers can then fork when they have releases or need to backport patches
04:53
It's because you were only doing solo work, without debian/ contributors
04:53
<vagrantc>
haven't had a project where people didn't balk at debian policy requirements
04:54
mostly, sure
04:54
<alkisg>
I find debian policy well written and a means to increase software quality
04:55
<vagrantc>
i mostly haven't had much time and energy for this particular project as a volunteer, either
04:55
could change, as it gets exciting again
04:55
so workflows that don't increase my workload and how i spend my free time also have value
04:56
<alkisg>
True; but imposing twice as much work to people that maintain upstream packaging isn't good for development either
04:56
OK to sum up and leave it behind, around July we'll need to decide if ltsp19 will have debian/ or packaging/debian.
04:56
For now, I'll just do that debian/ repository in launchpad
04:57* alkisg puts aside all those hard negotiations and tries to get some coding done! :)
04:57
<vagrantc>
just stick it in debian for now
04:57
<alkisg>
I can't; I can't build the recipes
04:57
<vagrantc>
for ltsp19 ?
04:57
<alkisg>
I'd have to undo your commits to build
04:57
Ah sorry
04:57
<vagrantc>
i'm not touching it
04:57
<alkisg>
I though you were talking about ltsp5
04:57
OK, great
04:57
<vagrantc>
yeah, ltsp5, let's be more cautious
04:58
<alkisg>
OK, for ltsp19 I'll start with debian/, in a month or so when I'll be ready,
04:58
for ltsp5, a debian/ repository in launchpad
04:58
<vagrantc>
for me, having upstream code and packaging in the same branch is so much less painful than a packaging-only branch
04:59
so i'll figure out a way to manage it
04:59
presuming i'm still around :)
04:59
<alkisg>
Let's hope so :)
05:00
<vagrantc>
might swear off computers, never know :)
05:08
<alkisg>
vagrantc: ...although now ltsp-client will need to be inherently cross-distro, as the same ltsp.img will be used in all distros
05:09
Even if that means adding "if distro == xx then else" in the code
05:09
<vagrantc>
that's going to be very interesting...
05:12
<alkisg>
Btw, go-md2man doesn't align options properly; but OK no big deal... https://github.com/cpuguy83/go-md2man/issues/45
05:13
Whoops, this link instead: https://github.com/cpuguy83/go-md2man/issues/27
05:26ricotz has joined IRC (ricotz!~ricotz@ubuntu/member/ricotz)
05:30* vagrantc isn't too fussy about manpages, but ok
05:34* vagrantc waves
05:34vagrantc has left IRC (vagrantc!~vagrant@unaffiliated/vagrantc, Quit: leaving)
05:35kjackal has joined IRC (kjackal!~quassel@2a02:587:3106:1000:a895:839f:87b0:61d8)
06:03os_a has joined IRC (os_a!~Thunderbi@195.112.116.22)
07:25woernie has joined IRC (woernie!~werner@p57A0E866.dip0.t-ipconnect.de)
07:53statler has joined IRC (statler!~Georg@gwrz.lohn24.de)
08:40josefig has left IRC (josefig!~josefig@unaffiliated/josefig, Ping timeout: 244 seconds)
08:53zavag has joined IRC (zavag!506ab707@ouxe21.static.otenet.gr)
08:53
<zavag>
alkisg καλημέρα
08:53
<alkisg>
Καλημέρα zavag
08:54
Κάνε λίγο vnc να δω:
08:54
!vnc-edide
08:54
<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
08:55
<zavag>
δεν τρεχει ...
08:56
<alkisg>
Δοκίμασε αυτό:
08:56
!vnc-dide
08:56
<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.
08:56
<alkisg>
x11vnc -noshm -connect srv1-dide.ioa.sch.gr
08:56
Σκέτο αυτό ^
09:55zavag has left IRC (zavag!506ab707@ouxe21.static.otenet.gr)
11:40spaced0ut has joined IRC (spaced0ut!~spaced0ut@unaffiliated/spaced0ut)
11:48Faith has joined IRC (Faith!~Paty_@unaffiliated/faith)
11:53woernie has left IRC (woernie!~werner@p57A0E866.dip0.t-ipconnect.de, Ping timeout: 258 seconds)
11:53woernie has joined IRC (woernie!~werner@p57A0E866.dip0.t-ipconnect.de)
12:14josefig has joined IRC (josefig!~josefig@unaffiliated/josefig)
14:16
<uumas>
I'm also using nfs3 for homedirs and going to switch to kerberized nfs4 soon, so lmk if some help is needed to test it
14:19
There are also a few schools around here using opisys/puavo ltsp
14:33kjackal has left IRC (kjackal!~quassel@2a02:587:3106:1000:a895:839f:87b0:61d8, Ping timeout: 258 seconds)
14:33kjackal has joined IRC (kjackal!~quassel@ppp-2-86-50-83.home.otenet.gr)
14:43
<alkisg>
uumas: thanks; ltsp19 won't have any integrated support for nfs4/kerberos, but what would need testing is, that it will work once the user has set up nfs4/kerberos/ldap by himself
14:44
If someone can write a small how to, I could do the initial testing + preparation myself, and once I have something running, make it public for more testing
14:44
re: puavo, since ltsp19 will be a rewrite, if they rely on upstream ltsp, they'll need to update their codebase somewhat
14:47
<uumas>
I already have freeipa (ldap+kerberos) setup and am going to reinstall our nfs server (running freebsd 10) with debian and set it up to use kerberos.
14:47
Will see how that works then
14:49
I guess opinsys are aware of ltsp19 already (I'm sure they follow the mailing lists), but based on past experience with them I believe they may take more than a year to switch over.
15:09josefig has left IRC (josefig!~josefig@unaffiliated/josefig, Quit: Ping timeout (120 seconds))
15:12josefig has joined IRC (josefig!~josefig@unaffiliated/josefig)
16:49kjackal has left IRC (kjackal!~quassel@ppp-2-86-50-83.home.otenet.gr, Ping timeout: 246 seconds)
17:04statler has left IRC (statler!~Georg@gwrz.lohn24.de, Remote host closed the connection)
17:40vagrantc has joined IRC (vagrantc!~vagrant@unaffiliated/vagrantc)
18:51kjackal has joined IRC (kjackal!~quassel@2a02:587:3106:1000:a895:839f:87b0:61d8)
19:41kjackal has left IRC (kjackal!~quassel@2a02:587:3106:1000:a895:839f:87b0:61d8, Ping timeout: 258 seconds)
19:41kjackal has joined IRC (kjackal!~quassel@2a02:587:3106:1000:a895:839f:87b0:61d8)
20:11woernie has left IRC (woernie!~werner@p57A0E866.dip0.t-ipconnect.de, Remote host closed the connection)
20:53Faith has left IRC (Faith!~Paty_@unaffiliated/faith, Quit: Leaving)
21:41ricotz has left IRC (ricotz!~ricotz@ubuntu/member/ricotz, Quit: Leaving)
21:45kjackal has left IRC (kjackal!~quassel@2a02:587:3106:1000:a895:839f:87b0:61d8, Ping timeout: 252 seconds)
23:43GodFather has left IRC (GodFather!~rcc@143.59.184.72, Ping timeout: 258 seconds)