User Tools

Site Tools


2015-02:day-2015-02-05

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

2015-02:day-2015-02-05 [2015/03/04 18:12] (current)
h1 created
Line 1: Line 1:
 +====== Developer Meeting 1/2015 ======
 +===== Who is taking part? =====
 +  * <todo @all>​About you and X2Go</​todo> ​
  
 +===== Topics =====
 +==== X2Go Project'​s Hardware ====
 +  * List all Hardware the Project is using
 +    * <todo @h1>​Who'​s able to administrate japsand.x2go.org?</​todo> ​
 +      * <todo @all>​action needed [x]Yes [ ]No </​todo>​
 +    * <todo @mike>​Who'​s able to administrate Mike's server?</​todo>​
 +      * <todo @all>​action needed [x]Yes [ ]No </​todo>​
 +  * <todo @all>​Where are backups of those server installations stored?</​todo> ​
 +    * <todo @all>​action needed [x]Yes [ ]No </​todo>​
 +  * <todo @all>Is there a fail-back solution?</​todo> ​
 +    * <todo @all>​action needed [x]Yes [ ]No </​todo>​
 +
 +==== X2Go Project'​s Software ====
 +  * <todo @all>​List all "​mission critical"​ Software the Project is using</​todo> ​
 +  * <todo @all>​Where are backups of those instances/​installations stored?</​todo> ​
 +  * <todo @mihai>​Status git.x2go.org</​todo> ​
 +    * <todo @mihai>​Who'​s got access to git.x2go.org</​todo>​
 +    * <todo @mihai>​What changes did you made to git.x2go.org</​todo>​
 +  * <todo @all>Are there a fail-back solutions?</​todo> ​
 +    * <todo @all>​action needed [x]Yes [ ]No </​todo>​
 +
 +==== X2Go Project'​s Services ====
 +  * <todo @h1>List all x2go.org subdomains</​todo>​
 +  * <todo @mike>​Future of our startssl certifcate</​todo> ​
 +  * <todo @mike>​What about the Mailinglists?</​todo> ​
 +  * <todo @all>Is there a fail-back solution?</​todo> ​
 +
 +===== Future of Develper Meetings =====
 +  * <todo @h1>Find a reccurring date</​todo>​
 +  * <todo @h1>​Include research tasks and presentation</​todo>​
 +  * <todo @h1>Add reviews and planning sessions</​todo>​
 +
 +====== Raw Chat Log ======
 +
 +<​file>​
 +(16:50:35) ***Stefan Baur waves at neverpanic h1Org Ionic sunweaver tan Chakka flexiondotorg_ and theuser2
 +(16:51:18) ***theuser2 yawns
 +(16:51:41) ***Stefan Baur slides a cup of coffee over to theuser2 ..............................c(__)
 +(16:54:04) Heinz-M. Graesing: OK... I' finally arrived at a tiny place eith a internet connection...
 +(16:56:16) Orion Poplawski [~orion@NORTHWEST-R.edge3.Denver1.Level3.net] hat den Raum betreten.
 +(16:57:11) Stefan Baur: sunweaver will be late
 +(16:58:23) Heinz-M. Graesing: my connection speed is not capable of audio streams... so we should use this IRC room for communication
 +(16:58:33) Stefan Baur: h1Org: So no together.js at all?
 +(16:59:00) Heinz-M. Graesing: MyNameIsRetro:​ the wiki page and maybe the co-browsing ide may help us
 +(16:59:42) Heinz-M. Graesing: Ionic: sry - appear.in will be the same problem for me at the moment
 +(17:00:12) Stefan Baur: h1Org: sunweaver expects to be with us in about 5 minutes, he is still stuck in traffic
 +(17:01:09) Stefan Baur: h1Org: Also, sunweaver asked that we take all the agena items that involve him and reschedule them to the very beginning o the meeting, as he will have to leave around 18:45 (at least I think that's what he said, the connection was bad)
 +(17:01:56) Heinz-M. Graesing: MyNameIsRetro:​ I don't see a problem with that...
 +(17:02:40) Mihai Moldovan: h1Org: yeah, sure, 3G.
 +(17:06:28) ***Mike#1 waves.
 +(17:07:19) Mike#1: I am ready.
 +(17:07:24) Stefan Baur: _o/
 +(17:07:44) Mike#1: Who else takes part. Participants please raise their hands...
 +(17:07:45) ***Mihai Moldovan oceans
 +(17:07:51) ***Mike#1 raises hands
 +(17:07:54) Stefan Baur: _o/
 +(17:07:56) Heinz-M. Graesing: Ionic: [»X] Welcome to our first Developer Meeting 2015 [»X]
 +(17:08:07) Mihai Moldovan: and raises his hand
 +(17:08:15) Heinz-M. Graesing: Please respect that this chatroom will be used as conference room for the next hour(s) and make sure that this communication will not be disturbed by any questions until the meeting is over!
 +(17:08:24) Heinz-M. Graesing: Thank you!
 +(17:08:31) Stefan Baur: sunweaver: If you raise *both* hands, that's a GESCHÄFTSORDNUNGSANTRAG and not a vote ;)
 +(17:08:48) Heinz-M. Graesing: For our participants:​ please make sure that a statement/​post of a person won't be interrupted! There will be one active spokesman at once and he will tell when his post will be completed (EOS - End Of Statement).
 +(17:08:50) ***Mike#1 has to leave at 5pm UTC (sorry, miscalculated between local time und UTC).
 +(17:09:03) Stefan Baur: h1Org: Might want to change the channel topic for the duration of the meeting?
 +(17:09:46) Mike#1: noone is ChanOp at the moment. Let's not bother.
 +(17:09:48) Heinz-M. Graesing: MyNameIsRetro:​ You're not a channel operator
 +(17:09:53) theuser2: h1Org, why aren't we simply opening a new room?
 +(17:09:55) Heinz-M. Graesing: ok
 +(17:10:06) theuser2: #x2go.conf ?
 +(17:10:13) Stefan Baur: h1Org: That's why I'm usggesting it rather than doing it myself :D
 +(17:10:32) Mihai Moldovan: lol, .conf
 +(17:10:34) Mike#1: wait. We would prevent people from reading this channels/​sessions backlog.
 +(17:10:59) theuser2: well its better that cluttering it up with general support requests and what not
 +(17:11:06) Heinz-M. Graesing: please start with our agenda - we can improve everything in our next meetings
 +(17:11:13) Mike#1: ack
 +(17:11:14) Heinz-M. Graesing: at the moment we should just start
 +(17:11:44) Heinz-M. Graesing: Thank you very much for beeing online!
 +(17:12:24) Heinz-M. Graesing: There are some facts unkown to a lot of people about our infrastructure
 +(17:13:03) ***Mike#1 can give a summary, if needed.
 +(17:13:08) Heinz-M. Graesing: an as the infrastructure will change, we should be aware about what is done in which way an where
 +(17:13:38) Heinz-M. Graesing: sunweaver: ok - please tell us about your HARDWARE
 +(17:13:59) Mike#1: the X2Go project runs two machines at the moment.
 +(17:14:06) Heinz-M. Graesing: sunweaver: in this case about who can access it and where are the backup stored too
 +(17:14:11) Mike#1: one is physical and sponsored by h1Org.
 +(17:14:19) Mike#1: The other is virtual and hosted as VM by me.
 +(17:14:39) Mike#1: the host is called ymir.das-netzwerkteam.de and provides several vhosts.
 +(17:14:47) Mike#1: git.x2go.org,​ code.x2go.org,​ lists.x2go.org,​ bugs.x2go.org.
 +(17:15:04) Mike#1: there are two dedicated X2Go logins: x2go and x2go-admin.
 +(17:15:34) Mike#1: several people can log into those accounts (h1Org, ncryer, mikedep333, Ionic, me).
 +(17:15:46) Mike#1: the "​x2go"​ user decides if people can commit to Git.
 +(17:16:02) Mike#1: The x2go-admin user decides whether people can stored webstuff on code.x2go.org/​releases/​
 +(17:16:24) Mike#1: the named reall persons have access to both account via SSH keys.
 +(17:16:35) Mike#1: for bugs.x2go.org,​ you need root access to ymir.
 +(17:16:44) Mike#1: similar for mailman administration on the fs level.
 +(17:17:04) ***Clemens also has access to x2go-admin@code.x2go.org to push OS X releases
 +(17:17:08) Mike#1: root access has been granted to me, to h1Org, to kasimo and phobie (co-admins at DAS-NETTZWERKTEAM).
 +(17:17:20) Mike#1: neverpanic: right.
 +(17:17:37) Mike#1: tan (Terje Andersen) can also commit to Git as user "​x2go"​
 +(17:18:07) Mike#1: for lists.x2go.org (mailman) several people know the listadmin passwords for all x2go-* lists.
 +(17:18:30) Mike#1: The mailman master PW is only know by me, kasimo and phobie (colleagues of mine not related to X2Go, see above).
 +(17:18:44) Mike#1: any questions concerning ymir?
 +(17:18:46) dowdle [~dowdle@scott.coe.montana.edu] hat den Raum betreten.
 +(17:18:56) Mike#1: EOS
 +(17:19:10) Mihai Moldovan: yeah, so ymir is a VM?
 +(17:19:21) ***Mike#1 nodes
 +(17:19:23) Mike#1: nodes
 +(17:19:24) Mike#1: nods
 +(17:19:31) Mike#1: KVM VM
 +(17:19:32) Heinz-M. Graesing: sunweaver: is there a failback system anywhere and is there a backup stored outside the same provider rooms?
 +(17:19:42) Mike#1: on a host named freyr.das-netzwerkteam.de,​ IIRC
 +(17:20:07) Mike#1: all DAS-NETZWERKTEAM servers are backed up on a separate machine locate in Kiel (LOGO EDV-Systeme GmbH).
 +(17:20:13) Mike#1: so is ymir
 +(17:20:23) Mike#1: there is no fallback system atm.
 +(17:20:40) Heinz-M. Graesing: Thank you!
 +(17:20:40) Alex|2 [~kvirc@dslb-178-010-090-039.178.010.pools.vodafone-ip.de] hat den Raum betreten.
 +(17:20:58) Heinz-M. Graesing: Alex|2: is that you?
 +(17:21:05) Alex|2: Hi
 +(17:21:13) Alex|2: yes. who is Alex#1 ?
 +(17:21:23) Heinz-M. Graesing: sorry for starting the conversation without you!
 +(17:21:31) Alex|2: no problem
 +(17:21:56) Heinz-M. Graesing: Alex|2: I'll give you the lines above per mail!
 +(17:21:58) Stefan Baur: Unless my computer fails me, we'll have a chat log online later on so Alex|2 can read up on everything he missed.
 +(17:22:01) Mike#1: shall I give the same summary on Japsand?
 +(17:22:21) Heinz-M. Graesing: ok sunweaver: we should continue with our second server, japsand
 +(17:22:27) Mike#1: right.
 +(17:22:32) Heinz-M. Graesing: sunweaver: If you like: please start
 +(17:22:42) Mike#1: Japsand is a physical server hosted at Hetzner (like the VM host of ymir, actually).
 +(17:22:57) Mike#1: the host is mainly used as the build system for X2Go packages.
 +(17:23:02) Mike#1: plus: wiki.x2go.org
 +(17:23:20) Heinz-M. Graesing: sunweaver: and our mike#2 windows vm
 +(17:23:21) Mike#1: wiki.x2go.org lives in an schroot based chroot.
 +(17:23:51) Mike#1: yeah, the WinBuilder VM has actually become part of the build system meanwhile.
 +(17:24:18) Mike#1: on the old Japsand used to be a Scientific Linux 6 VM, but that one was not migrated IIRC.
 +(17:24:21) Mike#1: It is obsolete anyway.
 +(17:24:36) Mike#1: The reason for having Japsand as a physical server is one part of the build system.
 +(17:24:44) Mihai Moldovan: what is schroot?
 +(17:24:52) Mike#1: Ionic: man schroot
 +(17:24:57) Heinz-M. Graesing: Ionic: devster31bian way of choot
 +(17:25:02) Heinz-M. Graesing: debian
 +(17:25:15) Stefan Baur: Ionic: https://​wiki.debian.org/​Schroot
 +(17:25:20) Mihai Moldovan: No manual entry for schroot :p
 +(17:25:31) Heinz-M. Graesing: ok, please continue
 +(17:25:32) Stefan Baur: silence please.
 +(17:25:36) Mike#1: we build packages for Ubuntu, Debian, Fedora, EPEL, openSUSE and SLE at the moment.
 +(17:25:46) Mike#1: The UBuntu packages are built at launchpad.net/​~x2go
 +(17:25:53) Mike#1: the rest is built on jenkins.x2go.org.
 +(17:25:57) Mike#1: aka japsand.
 +(17:26:04) Mihai Moldovan: hum, okay
 +(17:26:15) Mike#1: the Debian package build script (git://​code.x2go.org/​buildscripts.git)
 +(17:26:23) Mike#1: uses qemubuilder.
 +(17:27:06) Mike#1: in theory, we can build X2Go packages for various architectures with qemubuilder,​ but in reallife, the non-intel archs suck up a lot of CPU performance,​ so that never got activated.
 +(17:27:27) Mike#1: due to the usage of qemubuilder,​ the build system has to be a native system (and not a VM).
 +(17:27:48) Mike#1: because qemubuilder requires VT technology to perform well for amd64 and i386 builds.
 +(17:28:24) Mike#1: Git commits arrive on git.x2go.org and trigger Jenkins jobs via CURL requests against jenkins.x2go.org
 +(17:28:35) Mike#1: each Git push triggers several jenkins jobs.
 +(17:29:02) Mike#1: Problems with the build system: sometimes qemubuilder jobs freeze and need to be killed (from Jenkins, from console).
 +(17:29:38) Mike#1: UNfortunately,​ killing qemubuilder leaves some parts of the process in an non-intact state (open lo devices, bind-mounted dirs, non-removed tmp folders).
 +(17:29:44) Mike#1: so this needs to be look at soon.
 +(17:29:53) ***Heinz-M. Graesing adds a picture about what japsand looks like: http://​www.hetzner.de/​hosting/​produkte_rootserver/​ex40
 +(17:30:02) Mike#1: for Fedora and EPEL packaging, the mock utility is used.
 +(17:30:20) Mike#1: for openSUSE and SLE the obs-build tool gets used.
 +(17:30:30) Mike#1: I am involved in maintenance of those packages in Debian.
 +(17:30:50) Mike#1: questions till here (EOS)
 +(17:30:55) Heinz-M. Graesing: sunweaver: we should not forget about the NIC issue...
 +(17:31:03) Mike#1: yep.
 +(17:31:20) Mike#1: the onboard NIC requires a proprietary driver to perform well with high network loads.
 +(17:31:24) Mihai Moldovan: we should not go for hetzner again, btw...
 +(17:31:26) Mike#1: perform well means: not crash
 +(17:32:12) Mike#1: the buildscripts.git requires some explanation... Shall I?
 +(17:32:17) Heinz-M. Graesing: Ionic: traffic inside hetzners network is free - this is a very good point for rsync jobs!
 +(17:32:24) Heinz-M. Graesing: sunweaver: Yes, go for IT!
 +(17:32:42) Mike#1: Git commit -> jenkins trigger in post-receive hook -> jobs starts
 +(17:32:43) Mihai Moldovan: h1Org: yes, but the hardware is utter crap (seriously, you've experienced it multiple times too)
 +(17:32:53) Heinz-M. Graesing: but please, add one sentence about jenkins...
 +(17:33:05) Mike#1: each Debian packaging jobs starts ~jenkins/​bin/​x2go-build+upload-deb-package.
 +(17:33:20) Stefan Baur: Ionic: Please let's not discuss this now. It's sunweaver'​s turn.
 +(17:33:44) Mike#1: that packages sets up a build env, performs a Git checkout, perpares the package for build (add a changelog entry with our fancy version strings, etc.)
 +(17:34:21) Mike#1: each RPM build job fires up obs-build or mock via x2go-build+upload-rpm-package,​ also in ~jenkins/​bin.
 +(17:34:35) Mike#1: the scripts requires some pre-requisites.
 +(17:34:45) Mihai Moldovan: and cleans up the workspace... (which is prety important, to not have old packages accumulate!)
 +(17:35:07) Mike#1: (a) packages get signed, so in ~jenkins/​.gnupg/​ there is a GPG key for signing the packages. The GPG key is stored there and no password is required for signing stuff (CAVE)
 +(17:35:35) Mike#1: (b) the build scripts push built packages to packages.x2go.org/<​distro>/<​path>​
 +(17:35:44) Mike#1: (packages.x2go.org is also hosted on ymir).
 +(17:36:02) Mike#1: so jenkins@japsand can SSH to x2go-admin@code.x2go.org without PW.
 +(17:36:20) Mike#1: sometimes it comes in handy to test builds or the buildscripts themselves.
 +(17:36:53) Mike#1: so the x2go-admin@japsand.x2go.org can run the same build jobs by hand like what the jenkins use can trigger via Jenkins software.
 +(17:37:46) Mike#1: the build scripts are rather peculiar and historically grown, but they fulfill their purpose for the various distros and versions we build for.
 +(17:38:05) Mike#1: mikedep333 and Ionic have an idea how to maintain that stuff, I guess.
 +(17:38:15) Mike#1: EOS
 +(17:38:49) Mihai Moldovan: yeah, I've added SRPM package support to the RPM build script
 +(17:38:53) Heinz-M. Graesing: oK - there is a mirror "​packages2.x2go.org"​ on annother machine inside my realm
 +(17:39:09) Mike#1: relevant software: mock, qemubuilder,​ obs-build, rpm, createrepo, reprepro, dupload
 +(17:39:48) Heinz-M. Graesing: Are there any questions about what jenkins is?
 +(17:40:08) Mihai Moldovan: only one
 +(17:40:18) Mihai Moldovan: why is wiki.x2go.org on jenkins and not on ymir?
 +(17:40:25) Mihai Moldovan: err, japsand
 +(17:40:31) Mihai Moldovan: on japsand and not on ymir
 +(17:40:48) Mihai Moldovan: wouldn'​t it make more sense to have all HTTP vhosts on one machine?
 +(17:40:57) Mike#1: it is good to have informative platforms not all on one machine.
 +(17:41:02) Heinz-M. Graesing: wiki.x2go, is now rsynced to annother server (boxy) to avoid downtimmes o the wiki
 +(17:41:05) Heinz-M. Graesing: times
 +(17:41:08) Mike#1: but wiki.x2go.org should actually be on machine#3
 +(17:41:08) Stefan Baur: Ionic: And how do you inform people that your main web server is down? ;)
 +(17:41:18) Mihai Moldovan: hum, okay
 +(17:41:22) Heinz-M. Graesing: but japsand is a server exclusive for the project
 +(17:41:27) Mihai Moldovan: MyNameIsRetro:​ https://​web.archive.org
 +(17:41:31) Stefan Baur: Ionic: google "all eggs in one basket"​
 +(17:41:47) Heinz-M. Graesing: so "​everybody"​ related can access it
 +(17:41:47) Mihai Moldovan: yeah, got it
 +(17:42:16) Stefan Baur: alright. "ORDER IN THE COURT!"​ - h1Org, lead on
 +(17:42:41) Heinz-M. Graesing: MyNameIsRetro:​ I'll configure a failback wiki - but this can't be done for our complete build engine
 +(17:43:08) ***Mike#1 has to leave in 18min...
 +(17:43:12) Heinz-M. Graesing: so - this is a very important message: at the moment, japsand is a single point of failure
 +(17:43:31) Heinz-M. Graesing: what about git.x2go.org?​
 +(17:43:42) Mihai Moldovan: git.x2go.org is code.x2go.org and on ymir, right?
 +(17:43:44) Mike#1: git.x2go.org also is a SPOF
 +(17:43:46) Mihai Moldovan: which is backupped internally
 +(17:43:48) Mihai Moldovan: hum
 +(17:43:51) Mike#1: on ymir
 +(17:44:10) ***Stefan Baur should leave around the same time as sunweaver, maybe +30 minutes
 +(17:44:24) Mihai Moldovan: unfortunately,​ me too
 +(17:45:02) Stefan Baur: Alright. So let's keep gathering as much information as possible, I'll post the chat log to the wiki, then we can draw conclusions "​offline"​ and post them to the mailing list.
 +(17:45:05) Heinz-M. Graesing: sunweaver: what action is needed in the next months?
 +(17:45:36) Heinz-M. Graesing: sunweaver: will there be any changes to the machines itself?
 +(17:45:38) Mike#1: it is needed that people pick up tasks that I have been working on in the previous 2-4 years.
 +(17:45:50) Heinz-M. Graesing: will they keep up running on new administration?​
 +(17:46:03) Mike#1: I will not distract us by removing hardware resources from the project.
 +(17:46:15) Heinz-M. Graesing: OK, then go on for the tasks
 +(17:46:29) Mike#1: I continue to have an eye on ymir, I'd be happy if someone else takes over maintaining Jenkins and stuff related to taht.
 +(17:46:59) Mike#1: (1) Mihai has already started being in charge of git.x2go.org and the actual X2Go code.
 +(17:47:15) Stefan Baur: (props to Mihai!)
 +(17:47:26) Mike#1: Ionic: make sure you read x2go-commits ML.
 +(17:47:56) Mihai Moldovan: sunweaver: I do?
 +(17:48:03) Mike#1: In the past years I reviewed every commit with a second pair of eyes (except my own, of course). This is a task someone has to pick up.
 +(17:48:26) Mike#1: Ionic: please nag people about code style, wrong changelog entries.
 +(17:48:29) Stefan Baur: sunweaver: --verbose
 +(17:48:33) Stefan Baur: ah
 +(17:48:43) Mike#1: It is good if the git master knows what bugs exist on X2Go BTS.
 +(17:49:05) Mike#1: every dev should be responsible for checking if a commit will close an open bug and mark that in changelog.
 +(17:49:17) Mike#1: The git master may perform a second check on this.
 +(17:49:18) Mihai Moldovan: hum, okay, I don't read every-everything,​ that's true
 +(17:49:23) Mihai Moldovan: I probably should
 +(17:49:32) Stefan Baur: Ionic: you WILL now. *g*
 +(17:49:38) Mike#1: From the people in the room, Mihai has the cleanest coding style!!! SO please respect his feedback well!!!
 +(17:50:14) Mike#1: (2) we should name X2Go component owners for Git repos on git.x2go.org
 +(17:50:18) Mihai Moldovan: having all bug reports memorized is a difficult task
 +(17:50:29) Mike#1: a component owner is responsible for triaging incoming bugs to the X2Go BTS.
 +(17:50:36) Mike#1: Ionce: I know.
 +(17:50:42) Alex|2: how many people have write access to git?
 +(17:50:52) Alex|2: and who gives this access?
 +(17:50:53) Mike#1: about 7 or 8
 +(17:51:22) Mike#1: everyone with access git x2go@code.x2go.org can grant Git access to others (via SSH pubkeys in ~x2go/​.ssh/​authorized_keys)
 +(17:51:24) ***theuser2 thinks Ionic is a whitespace n.,..!
 +(17:51:34) Mihai Moldovan: theuser2: NAZI
 +(17:51:37) Mihai Moldovan: theuser2: SAY IT
 +(17:51:40) Mike#1: guys: stop.
 +(17:51:41) Mihai Moldovan: theuser2: I DARE YOU
 +(17:51:53) Stefan Baur: Ionic, theuser2, we don't have time for that.
 +(17:52:06) Mike#1: copmonent owners...
 +(17:52:11) Stefan Baur: Alex|2: There'​s a Wiki page documenting who has git access, but I'm not sure if it is up to date. sunweaver?
 +(17:52:20) Mihai Moldovan: actually, as the git-admin, I'm supposed to add new people to the access list and remove inactive ones, right?
 +(17:52:26) Mike#1: the component owners have to be responsible about people reporting bugs to X2Go BTS.
 +(17:52:46) Mike#1: In most cases those people really face a problem, so they should receive a reply at least. At best: a solution / bug fix.
 +(17:52:52) Mihai Moldovan: sunweaver: is that a smart idea? any committer adding other people
 +(17:53:01) Mike#1: Ionic: yes.
 +(17:53:19) Mike#1: it is not a good idea and it worked well for a small group of devs.
 +(17:53:49) Mike#1: but::: I would not introduce workflow changes now, until all relations and underlying mechanisms have been understood.
 +(17:54:07) Heinz-M. Graesing: sunweaver: I really will try to come up with new meeting dates that are accessible to you. Thank you very much for all your explanations!
 +(17:54:09) Mihai Moldovan: ok, but I will keep changing the authorized_keys file's permissions in mind
 +(17:54:12) Mike#1: also: infrastructure comes second atm, development/​project/​communication tasks should come first
 +(17:54:26) Alex|2: hm, I think it works as long not many people have such access.
 +(17:55:06) Mike#1: (3) one person needs to be reponsible for bugs.x2go.org in general (reassign bugs, triage bugs, dig out bugs with no feedback, etc.)
 +(17:55:26) Mike#1: tis person should be a person like MyNameIsRetro that is good at poking others.
 +(17:55:47) ***Stefan Baur will have to stock up on cattle prods then
 +(17:55:59) Mike#1: EOS, any questions?
 +(17:56:16) Heinz-M. Graesing: we all should be aware about the difference between a patch and a feature... This should be a topic for the next meeting regarding git.x2go.org
 +(17:56:57) Mike#1: (4) the git-admin should triage bugs with patches...
 +(17:57:16) Heinz-M. Graesing: sunweaver: thank you for adding your opinions and suggestions
 +(17:57:19) Mike#1: at least tag those bugs with "​patch"​ if the patch looks sane and request review from the component owner.
 +(17:57:26) Mike#1: or directly apply the patch if it makes sense.
 +(17:58:02) Heinz-M. Graesing: are there still pristine tar.gzs build?
 +(17:58:19) Mike#1: ohhh....
 +(17:58:23) Mike#1: the release workflow...
 +(17:58:33) Mike#1: that should be on the top list for next meeting.
 +(17:58:39) Mike#1: THanks for reminding.
 +(17:58:47) Heinz-M. Graesing: sunweaver: sory for interrupting you - when will you leave?
 +(17:58:52) Mike#1: 2min.
 +(17:59:14) ***Mike#1 sits in a coffee place that closes at 18.00 CET
 +(17:59:40) Heinz-M. Graesing: This won't be sufficiant for a whole question...
 +(17:59:50) Mike#1: sorry.
 +(17:59:50) Stefan Baur: Goodbye then, sunweaver, and thanks for being here tonight.
 +(17:59:55) Mike#1: welcome
 +(18:00:00) Mike#1: CU!
 +(18:00:03) Mike#1: _o/
 +(18:00:04) Alex|2: ok, next week again?
 +(18:00:12) ***Mike#1 will read the backlog...
 +(18:00:15) Heinz-M. Graesing: sunweaver: will it be ok, to find a new date they way we did it this time?
 +(18:00:32) Mike#1: yeah, but Thursday afternoon does not work well for me.
 +(18:00:51) Stefan Baur: h1Org: Is there anything we need to discuss today, with sunweaver gone? Or can we declare the meeting finished?
 +(18:00:52) Mike#1: only every two weeks and then from 1400 to 1600 UTC.
 +(18:01:14) Mike#1: please discuss the tasks I mentioned above, if possible.
 +(18:01:18) Mike#1: _o/
 +(18:01:26) ***Stefan Baur would welcome the ide of "​meeting"​ once every two weeks (or even less often)
 +(18:01:30) Stefan Baur: *idea
 +(18:01:54) Mihai Moldovan: actually, hum
 +(18:01:56) Mihai Moldovan: well...
 +(18:01:58) Alex|2: I want to speak about git access. I Don't think we need sunweaver for that
 +(18:02:00) Heinz-M. Graesing: for all remaining participants:​ these were very important informations!
 +(18:02:11) Mihai Moldovan: sunweaver: you still need to provide an all-components release, right?
 +(18:02:19) Alex|2: Ionic, if you have two mor eminutes?
 +(18:02:26) Mihai Moldovan: sure
 +(18:02:38) Heinz-M. Graesing: we should now care about those informations
 +(18:02:49) Heinz-M. Graesing: the should become part of our knowledgebase
 +(18:02:54) Stefan Baur: h1Org: it seems we will ever be able to arrange for a meeting that everyone can attend if we stick to the same weekday. ​ My suggestion: rotate through the weekdays and live with the fact that we can't always have everyone around.
 +(18:03:03) Stefan Baur: s/​ever/​never/​
 +(18:04:12) Alex|2: I don't think it's a good idea, that every one who can write to git should be possible to grant such access to other people
 +(18:04:13) Heinz-M. Graesing: MyNameIsRetro:​ at the moment, this choosen date (thursday) is not suitable for me, but we'll find a solution!
 +(18:04:22) Mihai Moldovan: Alex|2: me neither
 +(18:04:43) Stefan Baur: Alex|2, Ionic, sounds like something that you two can solve together
 +(18:05:01) Heinz-M. Graesing: I'll make a copy of this chatlog, but it should be spread to the wiki page locations
 +(18:05:16) Alex|2: yes it's true. Ionic, PM?
 +(18:05:57) Stefan Baur: h1Org: As I said, the simplest solution would be to live with the fact that we can't achieve perfection and a 100% attendance rate - and rotate through the weekdays (one week Monday, two weeks later Tuesday, etc.) so everyone gets a fair chance to attend.
 +(18:06:32) Heinz-M. Graesing: MyNameIsRetro:​ I'll keep that in my mind!
 +(18:06:53) Stefan Baur: Alex|2, Ionic I trust that you can come up with a good solution, just don't forget to make it public once you've agreed on one (Wiki, Mailing List)
 +(18:07:44) Heinz-M. Graesing: OK thank you very much for your attention! I'll close now this first meeting, but stay online untill I'll reach for my train!
 +(18:07:53) Stefan Baur: h1Org: We use the same principle for a real-life business meeting where I live and it works out rather well (the system is in place for more than 10 years)
 +(18:08:06) Alex|2: Yes, it's actually not an important thing for now. There are not so many people at the momemnt who want write to git...
 +(18:08:09) Heinz-M. Graesing: maybe you - developers can add this
 +(18:08:12) Heinz-M. Graesing: [»X]
 +(18:08:17) Heinz-M. Graesing: to your signature
 +(18:08:38) Heinz-M. Graesing: a miniature x2go-logo :)
 +(18:08:54) Stefan Baur: h1Org: --verbose - on what signature. ML? 
 +(18:10:11) Heinz-M. Graesing: MyNameIsRetro:​ I'm still working on a ascii seal - the movin X is made for the core team
 +(18:11:26) Heinz-M. Graesing: [»X] Thak you for taking part in our first Developer Meeting 2015 [»X]
 +(18:11:44) Heinz-M. Graesing: Thank
 +(18:11:56) Heinz-M. Graesing: 3G is annoying!
 +</​file>​
2015-02/day-2015-02-05.txt · Last modified: 2015/03/04 18:12 by h1