17:00:25 <jlkinsel> #startmeeting
17:00:25 <cs-meetbot> Meeting started Wed Sep 12 17:00:25 2012 UTC.  The chair is jlkinsel. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:25 <cs-meetbot> Useful Commands: #action #agreed #help #info #idea #link #topic.
17:00:34 <jlkinsel> #topic CloudStack meeting for 9/12/2012
17:01:02 <jlkinsel> Good day everyone, this is the weekly CloudStack IRC meeting
17:01:25 <jlkinsel> as a reminder, this is for discussion of relevant topics, but final decisions are not to be made here but on the cloudstack development mailing list
17:02:19 <rohityadav> how about maven?
17:02:22 <jlkinsel> #chair bhaisaab chipc cvittal edison_cs iswc Japje ke4qqq nslater rohityadav topcloud vogxn widodh
17:02:22 <cs-meetbot> Current chairs: Japje bhaisaab chipc cvittal edison_cs iswc jlkinsel ke4qqq nslater rohityadav topcloud vogxn widodh
17:02:38 <jlkinsel> let's go around in alphabetical order by nickname so we can keep some order to things
17:02:50 <rohityadav> +1
17:02:58 <nslater> hi guys :)
17:03:00 <jlkinsel> I'll mention each person's nick, if they have any topics to discuss, raise them, otherwise we'll move to next person
17:03:04 <jlkinsel> 'morning nslater
17:03:09 <jlkinsel> bhaisaab: anything to discuss?
17:03:18 * rohityadav is bhaisaab
17:03:27 <rohityadav> jlkinsel: maven, debs/rpms?
17:03:41 <jlkinsel> heh sorry :)
17:03:49 <chipc> That's the topic I think should be first too
17:04:00 * bhaisaab is rohityadav
17:04:02 <jlkinsel> sounds good. where do we stand today?
17:04:19 <edison_cs> we are using maven to install dependence
17:04:25 <edison_cs> ant to build jars
17:04:26 <rohityadav> bhaisaab is my logger bot, jfyi http://bhaisaab.org/irc/cloudstack/
17:04:33 <edison_cs> waf to build deb/rpm
17:04:59 <edison_cs> better to just stay the current status in 4.0
17:05:18 <rohityadav> edison_cs: tying couple of maven plugins, am able to build debs with jdeb (rpm wip), see diff on api submodule: http://pastebin.com/SB9GiHnx
17:05:36 <rohityadav> tomcat plugin won't work… any luck with mvn to deploy, debug?
17:05:47 <rohityadav> vogxn tells me jetty works for him
17:05:54 <chipc> rohityadav: did you bring this up on the list?
17:06:21 <rohityadav> chipc: sure, as soon as I'm done with a proof of concept pom/code, I'll share…
17:06:22 <chipc> I think we would need at least a lazy consensus about making that change, especially with all the current churn
17:06:30 <rohityadav> rpm is tricky deb was easy
17:06:32 <chipc> and the goal of getting to 4.0
17:06:43 <widodh> rohityadav: I don't agree that deb works that way
17:06:51 <rohityadav> widodh: what way?
17:06:56 <widodh> the problem with maven-deb is that postinst isn't working
17:07:03 <widodh> and we actually have a couple of postinst files
17:07:17 <rohityadav> widodh: I'm using another plugin called jdeb, viz the diff
17:07:30 <rohityadav> widodh: https://github.com/tcurdt/jdeb
17:07:43 <widodh> rohityadav: Ok, but what is the real gain for using jdeb instead of invoking dpkg-buildpackage?
17:07:55 <widodh> WAF will still generate the artifacts, right?
17:08:18 <chipc> rohityadav: is it build platform independent?
17:08:27 <edison_cs> rohityadav: still doubt the capability of maven plugin, spec/control is more powerful than maven
17:08:31 <rohityadav> mvm does everything, we can get rid of waf… I'm in favour of having maven do everything for us if possible, building, deploying and packaging… if that's possible
17:09:03 <rohityadav> chipc: i hope so with that diff, you do mvn clean install -X… they are no arch by default
17:09:07 * ke4qqq shows up late
17:09:09 <edison_cs> rohityadav: the first thing we need to do is to install artifacts build by ant or maven
17:09:15 <widodh> rohityadav: I agree that I'd like to see WAF do, I think we all want to
17:09:16 <edison_cs> so that we can get rid of waf
17:09:18 <widodh> edison_cs: +1
17:09:25 <rohityadav> ok, just an experiment I did, and it seemed to work
17:09:26 <rohityadav> :D
17:09:38 <widodh> rohityadav: No, good work! Very much appriciated
17:09:59 <chipc> rohityadav: +1 to wido's comment - great to try things!
17:10:05 <edison_cs> rohityadav: +1
17:10:11 <rohityadav> widodh: okay, this is not it, there are two other plugin but they are tricky, so as soon as I've something concrete I'll post a patch or review...
17:10:23 <chipc> rohityadav: how about doing what I suggested…  propose it on the dev list as a suggestion?
17:10:26 <rohityadav> next thing, about maven-waf and maven branches, what to do with them?
17:10:35 <rohityadav> okay chipc!
17:10:42 <widodh> I'd like to see it in action
17:10:45 <rohityadav> will do that soon!
17:10:55 <widodh> and see what it does, but I'm sceptical at this point
17:11:00 <edison_cs> hugo and I will work on maven-waf branch in the next few weeks
17:11:13 <widodh> The control file is very powerful for Debian packages
17:11:26 <edison_cs> hugo said he will come to my office, maybe at end of this month
17:11:35 <rohityadav> widodh: yes that is used by that plugin
17:11:36 <jlkinsel> #action rohityadav to propose building package files with maven
17:11:39 <rohityadav> cool!
17:11:47 <edison_cs> ok, move on
17:11:50 <chipc> widodh / edison_cs - your intent wasn't to to get it ready by 4.0 (the maven-waf branch)
17:12:01 <chipc> correct?
17:12:04 <edison_cs> yes
17:12:17 <chipc> and on the maven branch, IIRC we haven't seen Darren around working on it recently
17:12:26 <chipc> it was a start for what's currently in master
17:12:36 <chipc> and could probably be proposed for deletion now
17:12:43 <edison_cs> maven still works
17:12:53 <edison_cs> we just don't use it for build
17:12:53 <chipc> the maven branch isn't needed now though, right?
17:13:05 <chipc> I'm talking about the branch
17:13:09 <rohityadav> maven branch works? so merge useful commits then?
17:13:14 <rohityadav> and delete the branch?
17:13:24 <widodh> jlkinsel: Sorry to get back to it
17:13:30 <widodh> but is the action correct for rohityadav
17:13:31 <edison_cs> I think huog already merged some of patches from maven branch into master
17:13:43 <widodh> I think he's proposing to try jdeb and build the packages with maven
17:13:57 <chipc> edison_cs - exactly what I believe to be true
17:14:12 <chipc> last update to the maven branch was 2 weeks ago
17:14:18 <jlkinsel> yep will update one sec
17:14:27 <rohityadav> widodh: +1 nothing concrete yet, but that seems to work
17:14:56 <jlkinsel> #undo
17:14:56 <cs-meetbot> Removing item from minutes: <MeetBot.items.Action object at 0x23b63d0>
17:15:23 <edison_cs> chipc: we just don't use maven to build and package, as hugo did in maven-waf branch
17:15:31 <jlkinsel> #action rohityadav to propose on dev list to try jdeb and building packages with maven
17:15:46 <vogxn> may i propose we should try the debs generated by jdeb on a 12.04 LTS ? that's the final target for the deb after all
17:16:05 <jlkinsel> #unlurk
17:16:20 <rohityadav> vogxn: okay we'll try them on your machines tomorrow
17:16:44 <vogxn> if it doesn't work out well. i think we can stick to waf?
17:17:07 <widodh> vogxn: rohityadav: Be prepared, WAF does a lot
17:17:08 <edison_cs> rohityadav: I think we'd better focus on how to install artifacts first… Without it, we can't get rid of waf
17:17:12 <rohityadav> vogxn: I never said we need to nuke waf now, only when maven gets matured enough
17:17:25 <widodh> all the artifacts like , , etc
17:17:34 <widodh> and replacing all those things, that's a lot of work
17:17:36 <rohityadav> edison_cs: +1 totally agree with you, let's keep that on priority
17:17:54 <rohityadav> no premature breaking of build system, we already had that once
17:18:44 <widodh> btw, we should note that it's common on Debian systems, that whenever a "debian" directory is available, you can just hit "dpkg-buildpackage" and it will sort itself out
17:18:47 <vogxn> rohityadav: testing the artifacts will give us a good idea about the jdeb plugin. so let's add that as part of the action. i can test the debs done by rohityadav
17:18:53 <edison_cs> #action: rohityadav investigates how to install artifacts
17:18:57 <edison_cs> is that ok?
17:19:09 <rohityadav> cool, +1
17:19:13 <widodh> you can have dpkg-buildpackage invoke maven for building the artifacts
17:19:20 <rohityadav> vogxn: okay, first thing tomorrow morning
17:19:22 <widodh> like it's currently invoking waf
17:19:24 <jlkinsel> edison_cs: +1
17:19:52 <widodh> I'll shut up now over the packages :)
17:19:57 <rohityadav> widodh: jdeb is little different does not used dpkg-buildpackage, that's why I was able to do that on mac osx, that needs testing though :D
17:20:20 <edison_cs> how to build rpm/debs is the next stage
17:20:24 <edison_cs> :0
17:20:28 <edison_cs> let's move on
17:20:30 <jlkinsel> ok, moving on...
17:20:32 <jlkinsel> chipc: ?
17:21:19 <chipc> I'm going to suggest (on list) that we make a final decision about how to bundle binaries for 4.0 (for ASF infra distribution)
17:21:32 <chipc> we need to drive that part to closure
17:21:35 <jlkinsel> yeah
17:21:49 <edison_cs> the options we have right now?
17:21:54 <jlkinsel> #action chipc to suggest on list that we make a final decision about how to bundle binaries for 4.0
17:22:09 <vogxn> chipc: what about non-OSS binaries? low priority now?
17:22:22 <chipc> we can't do non-OSS binaries from ASF
17:22:38 <edison_cs> vogxn: non-oss is low priority, it doesn't work right now
17:22:43 <chipc> edison_cs - I'll try to summarize the thread back on the list, so that we can get to either consensus or have a vote
17:22:46 <vogxn> got it
17:22:48 <edison_cs> ok
17:22:54 <ke4qqq> yes, we might get away with non-ASF friendly convenience binaries, but certainly not non-oss
17:23:27 <chipc> other topic - we have not yet got an answer from legal-discuss on the config files in patches
17:23:53 <chipc> I'll poke them again, but I think we may need to ask for mentor advice on how to get to closure on that issue
17:23:56 <ke4qqq> did we actually outline the config files that we have - I thought they were waiting on us?
17:24:05 <chipc> ke4qqq: chiradeep did that
17:24:10 <ke4qqq> ok, I missed it
17:24:18 <chipc> perhaps we could provide a more exhaustive list
17:24:30 <chipc> and perhaps *we* need to do the work to ask the source projects
17:24:44 <ke4qqq> perhaps
17:25:13 <chipc> #action chipc to start a DISCUSS thread on the dev list about how we want to approach coming to resolution on the config files in the patches dir
17:25:15 <chipc> ;-)
17:25:26 <chipc> otherwise it will hang out there forever
17:25:35 <edison_cs> that's the only legal issue may block our release?
17:25:48 <ke4qqq> no
17:25:57 <ke4qqq> we still have export control filings
17:26:28 <ke4qqq> I started the thread on legal-discuss to figure out the ASF's process
17:26:35 <chipc> plus, I'm still not satisfied that we have completely reviewed all files in the tree for compliance
17:27:23 <ke4qqq> me either - I am still parsing through it - and based on what we have hit so far, there's still lots of potential problems
17:27:34 <jlkinsel> that one seems to not go away (src compliance)
17:27:46 <edison_cs> chipc: is there a process to review all the files?
17:28:13 <vogxn> jenkins job that does RAT? will that help?
17:28:25 <chipc> edison_cs: yes, literally looking at them…  ke4qqq has been working on it as well
17:28:30 <ke4qqq> vogxn: that would help going forward.
17:28:39 <chipc> unfortunately, it isn't something that's easy to automate until we know our baseline
17:28:51 <chipc> there's just too much code from too many sources
17:28:54 <ke4qqq> edison_cs: less/vim/emacs/grep/find
17:29:03 <vogxn> ouch
17:29:05 <chipc> +1 - yeah
17:29:08 <ke4qqq> it's mind thrashing
17:29:10 <rohityadav> vogxn: I did propose a tool I found, but I was suggested to do manual QA and then later we will have RAT/whisker
17:29:32 <ke4qqq> rohityadav: more eyes, and more ways looking at it is good - go find problems and file bugs
17:29:57 <edison_cs> rohityadav: I see your tool, does it work?
17:30:09 <chipc> jlkinsel: that's it from me
17:30:13 <jlkinsel> ok
17:30:21 <rohityadav> ke4qqq: okay, btw license checker log for last time: http://bhaisaab.org/cloudstack-checklicense.txt (files without Apache license)
17:30:40 <jlkinsel> cvittal: anything to discuss?
17:30:45 <cvittal> no
17:30:48 <rohityadav> edison_cs: seems so, but I'll do manual QA this weekend, that too much manual labour :D
17:31:02 <jlkinsel> thx
17:31:19 <jlkinsel> edison_cs: anything to discuss?
17:31:26 <edison_cs> the build
17:31:37 <edison_cs> seems deb build is broken
17:31:47 <edison_cs> widodh: deb build works for you?
17:31:59 <widodh> edison_cs: Yes, I've build deps 30 minutes ago
17:32:11 <widodh> dpkg-buildpackage does the job
17:32:23 <edison_cs> that's weird, build on jenkins failed all the time
17:32:25 <ke4qqq> edison_cs: the problems on jenkins don't seem really 'build' related - it's outside of the actual build
17:32:45 <ke4qqq> disk space?
17:32:51 <edison_cs> yah, maven failed to download the deps
17:33:16 <edison_cs> maybe network is too slow on that build machine
17:33:22 <widodh> edison_cs: I've been having a fight with maven for a couple of weeks now
17:33:32 <edison_cs> yah, painful
17:33:35 <widodh> sometimes it tries to fetch stuff from ws.zones.apache.org
17:33:37 <edison_cs> I do know
17:33:40 <rohityadav> edison_cs: setup all reps, and use -o to have maven build in offline mode
17:33:43 <rohityadav> *deps
17:34:00 <widodh> ws.zones.apache.org has been dead for a long time
17:34:11 <edison_cs> that's weird, why build on ubuntu will try to download from ws.zones.apache.org
17:34:31 <edison_cs> while on centos 6.3, seems no such issue at all
17:34:34 <chipc> maven version?
17:34:47 <widodh> No idea, I have maven 3.0.4 on Ubuntu
17:35:01 <edison_cs> rohityadav: how to setup reps?
17:35:40 <vogxn> edison_cs: i build on ubuntu 12.04 without problems today.
17:35:41 <ke4qqq> rohityadav: +1 offline mode will let us go much faster anyway
17:35:55 <rohityadav> sorry that was deps… get all the dependencies that maven needs into ~/.m2 and do mvn -o compile or mvn -o install
17:36:05 <ke4qqq> rohityadav: only concern is when deps change, but that should (at this point) be rare
17:36:34 <chipc> +1 - and when they do, we'll know because the build will break
17:36:37 <rohityadav> this won't get poms/jar every time maven is called as it's painful that everytime you do anything with maven it download them from maven repos
17:37:14 <rohityadav> ke4qqq: do refresh every 12/24 hours or so, no need to build in online mode, I use -o that saves time
17:37:18 <vogxn> +1 for offline maven. seems like a good checkpoint
17:37:33 <edison_cs> ok, i'll try offline mode
17:37:39 <edison_cs> good to learn
17:37:40 <rohityadav> just keep this option in mind, just in case building is taking a lot time
17:38:02 <rohityadav> edison_cs: mvm -o <life cycle target, like compile install package>
17:38:12 <rohityadav> *mvn, -X to debug
17:38:19 <edison_cs> ok, move on
17:38:28 <jlkinsel> thx sir
17:38:36 <jlkinsel> iswc: are you around?
17:39:51 <jlkinsel> guessing no.
17:39:56 <jlkinsel> Japje: around/anything to add?
17:39:58 <Japje> im here for the first time and was intrested in attending/lurking a cs meeting. i have no topics to discuss ;-)
17:40:10 <jlkinsel> good enough, thanks for the response and welcome. :)
17:40:13 <jlkinsel> ke4qqq
17:40:57 <chipc> #action edison_cs to explore builds using maven's offline mode for jenkins.c.o
17:41:06 <jlkinsel> ah yes thx
17:41:09 <chipc> ;-)
17:41:18 <jlkinsel> guessing ke4qqq stepped away, will come back to him later.
17:41:22 <jlkinsel> nslater: anything from you?
17:41:33 <nslater> nope. not this week
17:41:38 <jlkinsel> thx
17:41:41 <jlkinsel> topcloud?
17:42:01 <topcloud> me already?
17:42:08 <jlkinsel> yes sir
17:42:13 <topcloud> i just have some things to report.
17:42:30 <topcloud> reviews: a lot of activities there but review counts have not dropped since initial cleanout.
17:42:40 <chipc> topcloud: thanks for doing all the daily reminders, btw
17:42:47 <vogxn> +1
17:42:51 <topcloud> bugs: most bugs are build (been discussed today) and document so far.
17:43:19 <ke4qqq> topcloud: /me is somewhat concerned with folsk committing things outside their core competency in a rush to get things cleaned out - but love the daily reminders
17:43:21 <topcloud> qa: is progressing on testing but rather slowly due not nightly build not really setup.
17:43:50 <topcloud> if there are specific concerns, please let me know.
17:43:57 * vogxn echoes ke4qqq's concern
17:44:09 <topcloud> majority of the people i've picked to commit are people i know who knows the code.
17:44:21 <topcloud> if i missed on anything specific, please bring it to my attention.
17:44:36 <ke4qqq> topcloud: I'll ping you off list - and you can tell me whether I am paranoid or not.
17:44:44 <ke4qqq> don't want to call folks out needlessly
17:44:55 <topcloud> my biggest concern right now is how slow qa is picking up testing the release.
17:45:15 <topcloud> so to me our biggest concern has to be able to give qa stable nightly builds.
17:45:15 <edison_cs> i'll have a automate test in this week
17:45:34 <topcloud> so that's my list for now.
17:45:41 <edison_cs> setup up machines in our lab for asf automate test
17:45:47 <topcloud> ke4qqq: definitely send me a list and i'll go over them.
17:45:54 <jlkinsel> #info significant concern that we have stable nightly builds for QA
17:46:00 <edison_cs> it will be integrated into jenkins.c.s
17:46:24 <chipc> topcloud: suggestion for you - the Citrix QA team's involvement is great, but they are kind of unknown on the list
17:46:37 <chipc> and any plans they might have aren't on the list either
17:46:52 <chipc> it might be good for them to say "hi, here's us and here's what we'd like to do"
17:47:01 <chipc> just IMO
17:47:16 <jlkinsel> +1
17:47:17 <widodh> chipc: +1
17:47:24 <topcloud> chipc: good point.  i'll relay that to sudha
17:47:33 <topcloud> i do believe the plans are public on the old wiki.
17:47:34 <vogxn> i think the wiki page of qa lists the folks testing from citrix side
17:47:41 <vogxn> cwiki has the link
17:47:45 <jlkinsel> I guess there's a line somewhere between what citrix does and what CS does, but..
17:48:05 <topcloud> no in this case all citrix qa is working in the interest of CS.
17:48:09 <topcloud> there should be no line.
17:48:15 <chipc> yeah, we get that…  it's awesome
17:48:30 <chipc> just like devs though, interacting and planning on the list is key
17:48:30 <topcloud> i'll ask people to introduce themselves on the list.
17:48:36 <chipc> thanks!
17:48:36 <rohityadav> +1
17:48:46 <vogxn> https://cwiki.apache.org/confluence/display/CLOUDSTACK/ASF+CS+4.0+QA
17:48:46 <jlkinsel> ok
17:49:01 <jlkinsel> vogxn: you've got the conch
17:49:21 <vogxn> jlkinsel: no actions from my side. will be helping edison_cs with the automate stuff
17:49:25 <jlkinsel> great
17:49:31 <jlkinsel> widodh
17:49:33 <vogxn> and rohityadav to test with the artifacts from jdeb
17:49:38 * vogxn done
17:49:45 <jlkinsel> thx :)
17:49:45 <widodh> ok :)
17:50:06 <rohityadav> :) vogxn
17:50:12 <edison_cs> vogxn: +1
17:50:17 <widodh> We discussed the Debian packaging already. But I'd like to come to a decision about the Debian repo. ke4qqq started a thread on the ml regarding this
17:50:35 <widodh> imho (expressed it a couple of times, feel like repeating ;) ) the install.sh isn't working
17:50:52 <chipc> widodh - it works pretty well on RHEL
17:50:52 <widodh> The discussion about a Debian repo seems to die over and over
17:50:53 <edison_cs> widodh: is it ok to host a rpm repo also?
17:51:06 <topcloud> vogxn: are you prasanna?
17:51:15 <widodh> chipc: Yes, but there are so many drawbacks on the longer run
17:51:18 <vogxn> topcloud: that's me
17:51:26 <widodh> edison_cs: Yes, I prefer a DEB and RPM repo
17:51:39 <chipc> *laughs* - we need a name to nick translation page
17:51:40 <rohityadav> widodh: yeah it does work for cent/rhel
17:51:44 <widodh> Now, we need space. I've offered it a couple of times, but I want to call a vote on the ml to push through
17:51:54 <widodh> Wether we do it or stick with install.sh
17:51:59 <ke4qqq> chipc: we need committer id == irc nick :)
17:52:06 <edison_cs> widodh: nice
17:52:09 <rohityadav> chipc: /whois?
17:52:11 <topcloud> vogxn: can you take that feedback to the QA team?
17:52:30 <vogxn> #action have QA formally introduced and do daily planning in list
17:52:53 <vogxn> topcloud: sounds good?
17:53:03 <topcloud> vogxn: yup...thx
17:53:03 <edison_cs> widodh: if we go to repo, may need a way to integrate with jenkins ubild
17:53:09 <edison_cs> build
17:53:15 <widodh> edison_cs: Yes, as a second step that would be nice
17:53:21 <edison_cs> ok
17:53:21 <chipc> widodh - I have an action from earlier to try to get that to the point of decision
17:53:34 <widodh> it would make the documentation much easier. #1 add the repo, #2 apt-get install / yum install XXXX
17:53:42 <chipc> widodh: you have me convinced, after thinking about it…
17:53:45 <widodh> And updates would just be: apt-get upgrade or yum upgrade
17:53:51 <Japje> +1 widodh
17:54:08 <widodh> Have apt and yum sort everything out
17:54:10 <chipc> I actually would suggest that AFTER that's done, we have a scripted install that does exactly what you just said for the folks that are really new
17:54:30 <jlkinsel> yep +1 that'd be great
17:54:41 <widodh> chipc: Sure, it's a couple of commands and you're done
17:54:49 <edison_cs> chipc: that's true, but that script will be much simpler that install.sh
17:55:04 <widodh> I'd vote for not having a script, it's 4 commands
17:55:40 <widodh> http://pastebin.com/ZqLHuAJ7
17:55:49 <chipc> heh, yeah, I hear you
17:56:04 <vogxn> widodh: that's really useful for unattended installs
17:56:05 <chipc> let's drive to conclusion on list
17:56:11 <widodh> The first 3 steps go into: Configurting the cloudstack repositories
17:56:21 <widodh> chipc: Yes, but there is one blocker
17:56:35 <widodh> We can't host this on ASF infra due to the nature of a deb/rpm repo
17:56:55 <chipc> ok - so then perhaps our proposal is that we, as an ASF project, only provide source binaries?
17:56:57 <edison_cs> I am also concerned about it
17:56:58 <rohityadav> widodh: host repo on launchpad, just add-apt-repo pap:<> and apt-get install :)
17:57:08 <ke4qqq> widodh: svn does the same thing
17:57:13 <ke4qqq> look at their binary downloads page
17:57:19 <ke4qqq> and see where the downloads are hosted
17:57:22 <edison_cs> what's the artifacts we need to release for a ASF project
17:57:28 <ke4qqq> edison_cs: source
17:57:30 <widodh> ke4qqq: I'll check that out
17:57:37 <edison_cs> source code?
17:57:42 <widodh> I'm offering to host the mirror, but we need to call a vote for that
17:57:47 <chipc> edison_cs - ASF releases are officially source code only
17:57:47 <ke4qqq> edison_cs: that is the ASF releace - source code
17:57:57 <chipc> ^ what he said
17:57:58 <jlkinsel> #info look at how svn provides binary downloads
17:58:23 <rohityadav> widodh: +1
17:58:32 * ke4qqq can't type
17:58:39 <edison_cs> but usually, the asf project also provides tar.gz on their website
17:58:46 <nslater> widodh: you dont need to call a vote on that
17:58:58 <nslater> widodh: remember that you are offering this as a convenience for users, on your own behalf
17:59:09 <edison_cs> like http://tomcat.apache.org/download-60.cgi
17:59:11 <nslater> anyone could provide a place to download binary packages
17:59:18 <widodh> nslater: True, but if we mention it in the docs?
17:59:19 <chipc> widodh: and the project could choose to link to you from our website
17:59:28 <nslater> that is fine also. you dont need to ask permission for everything
17:59:33 <rohityadav> nslater: i think we're looking for something rather semi-official?
17:59:42 <nslater> i find it better to work under "it's better to ask forgiveness than it is to ask permission"
17:59:56 <nslater> rohityadav: you cannot have semi-official binary packages
17:59:59 <widodh> nslater: I don't want to go rogue, set up the mirror and add it on the docs
18:00:03 <chipc> nslater - we should at least discuss the overall idea on list
18:00:03 <nslater> i know what you are saying but you have to be careful about your wording
18:00:25 <widodh> s/mirror/repo
18:00:25 <nslater> here's how we would do it on couchdb, as a reference point:
18:00:30 <ke4qqq> nslater: thought on pointing repo.cs.o to widodh's host?
18:00:35 <nslater> * Bob announces he wants to host binary packages
18:00:37 <ke4qqq> or does that make it seem to official?
18:00:42 <nslater> * We have a discussion on the list, it all seems fine
18:00:46 <nslater> * Bob puts up the packages
18:00:55 <nslater> * A committer updates the homepage, or whatever to point to them, with a note
18:01:20 <chipc> nslater: that's good advice
18:01:21 <nslater> no voting, no permission getting, just an open discussion where one person says "i want to do this" and a committer says "looks good, i will update a link"
18:01:24 <rohityadav> +1 widodh?
18:01:35 <rohityadav> that's nice advice
18:01:42 <widodh> +1
18:01:45 <edison_cs> +2
18:02:04 <nslater> an apache project is a collection of people doing stuff, and talking about the stuff they are doing. if you want to change something, change it. discuss it first if you think it needs discussing. mistakes can be corrected
18:02:16 <widodh> I'll continue with the repo and put it on the list.
18:02:19 <nslater> (about the only exception to this is releases. releases... cant be corrected.)
18:02:21 <jlkinsel> if the bandwidth usage becomes significant, we can consider mirrors...
18:02:26 <widodh> nslater: Thanks!
18:02:52 <nslater> ke4qqq: about it seeming too official, let me dig something up
18:03:04 <ke4qqq> ok
18:03:14 <chipc> widodh / edison_cs - can we get it all tied together, so that jenkins is building the same type of artifact that you'll host?
18:03:24 <chipc> (you in that being widodh)
18:03:25 <nslater> see the wording here:
18:03:25 <Japje> +1 on having mirrors' i hate it when there is just 1 place to download something and its down/unreachable
18:03:26 <nslater> http://www.apache.org/dist/couchdb/
18:03:45 <widodh> Ok, my action will be that I'll continue with the mirror on my own, try to work with edison on that and make it easier. I'll put this on the ml again to get it clear
18:03:47 <nslater> (i maintain this dir... im the couchdb rm)
18:03:49 <widodh> what my actions will be
18:03:56 <nslater> we have two dirs
18:03:57 <nslater> -
18:03:59 <nslater> Binary Packages
18:03:59 <nslater> The binary packages are provided for your convenience only.
18:03:59 <nslater> They have not been voted on and do not consititue an official Apache CouchDB release.
18:03:59 <nslater> Source Releases
18:03:59 <nslater> The official Apache CouchDB source releases are voted on by the community and are approved by the PMC.
18:04:00 <nslater> -
18:04:14 <nslater> putting binaries under dist works for us. but it's not required
18:04:20 <widodh> nslater: Yes, that works for me
18:04:24 <nslater> if you look at our homepage we also link to a windows download *on github*
18:04:31 <widodh> and it makes life for users much easier
18:04:39 <nslater> (i am actually not too thrilled by this, and am working to try and get that under /dist)
18:04:52 <widodh> jlkinsel: I'm done :)
18:05:07 <widodh> Actually have to go, I'm already late for a dinner!
18:05:16 <chipc> thanks widodh
18:05:24 <nslater> widodh: your plan seems cromulent to me
18:05:27 <nslater> widodh: thanks!
18:05:30 <ke4qqq> nslater: so provided we make it clear it's not official - no other problems seen in point repos.cs.o to wido or similar
18:05:32 <jlkinsel> thanks wido
18:05:40 <edison_cs> thanks widodh
18:05:41 <jlkinsel> |ind-cloud|: anything to add?
18:05:42 <nslater> ke4qqq: not that i am aware of
18:05:43 <ke4qqq> s/point/pointing
18:05:55 <ke4qqq> nslater: awesome - thanks for the sanity check
18:06:00 <nslater> np
18:06:01 <widodh> I'm afk!
18:06:01 <jlkinsel> yeah
18:06:09 <jlkinsel> ke4qqq: missed you earlier, anything to discuss?
18:06:19 <ke4qqq> jlkinsel: finally got software grant submitted
18:06:23 <ke4qqq> still waiting on export control
18:06:25 <ke4qqq> that is all
18:06:31 <jlkinsel> I didn't see the attachment on that email, fyi
18:06:46 <chipc> ke4qqq: what are you waiting for on the export control side?  getting a copy of the orig CS one?
18:06:47 <ke4qqq> jlkinsel: it got stripped by the list - happy to send you a copy directly if you wish
18:06:57 <jlkinsel> nah that's ok, as long as those who need it got it
18:07:03 <nslater> chipc: who can i work with to get the ip clearance stuff sorted?
18:07:06 <ke4qqq> chipc: no I have that - but the export control page at the asf says 'email the list'
18:07:11 <nslater> chipc: we'll need various things signing and faxing
18:07:24 <ke4qqq> nslater: I am happy to pick that up
18:07:29 <chipc> nslater - I'm not from Citrix
18:07:30 <chipc> ;-)
18:07:38 <ke4qqq> nslater: I'll shoot you an email with my contact info
18:07:43 <chipc> ke4qqq is your guy
18:07:47 <nslater> ke4qqq: you're happy to work with me on that, you mean?
18:07:52 <nslater> ke4qqq: from the citrix side?
18:07:59 <ke4qqq> nslater: yes
18:08:08 <nslater> sweet. my email is nslater@apache.org
18:08:09 <nslater> thanks!
18:08:24 * ke4qqq doesn't know how the ip clearance stuff works, but I've made lots of friends at citrix legal in the past 6 months
18:08:33 <jlkinsel> heh
18:08:39 <nslater> i dont, either. not very well. ive just done my first one from couchdb
18:08:44 <nslater> but we can muddle through it together
18:09:04 <jlkinsel> ok guys - anything else?
18:09:08 <chipc> ke4qqq / nslater - that has to be done before we do a release, right?
18:09:18 <nslater> chipc: only if you want those tests in your next release
18:09:24 <chipc> ah, it was for the tests
18:09:25 <chipc> got it
18:09:30 <nslater> chipc: if the community is happy shipping a release without tests, thats fine
18:09:34 <chipc> no, those can make it in whenever they make it in
18:09:38 <chipc> IMO
18:09:38 <nslater> *without THOSE tests
18:09:43 <nslater> cool
18:09:56 <ke4qqq> nslater: email sent
18:10:06 <nslater> i presume the main bulk of work has already passed ip clearence
18:10:10 <nslater> however..!
18:10:13 <ke4qqq> nslater: uhhh
18:10:14 <nslater> i checked the repos this morning. empty?
18:10:34 <ke4qqq> nslater: depends - what is the process for ip clearance? :)
18:10:43 <chipc> ke4qqq / nslater - that was my actual question - did the orig CS code go through it?
18:10:47 <chipc> or does it need to?
18:10:49 <nslater> ah, no idea!
18:10:51 <nslater> yes, it does!
18:10:52 <ke4qqq> and the fact that we don't know - means it probably hasn't
18:11:00 <nslater> i presumed that an existing mentor had been doing this
18:11:00 <chipc> ^ right
18:11:12 <nslater> your repos is empty, so i guess you've not done the import yet
18:11:25 <ke4qqq> nslater: so brett raelized not SGA existed, and we finally got that in place
18:11:30 <ke4qqq> nslater: whch repo?
18:11:35 <nslater> the asf one
18:11:41 <chipc> we're in git
18:11:42 <ke4qqq> we are on git, not svn
18:11:45 <chipc> it's not empty
18:11:49 <nslater> oh my that explains it
18:11:51 <nslater> hehe
18:11:56 <nslater> i think your podling page is out of date
18:11:57 <ke4qqq> or if it is, we are all grandly delusional :)
18:12:03 <chipc> *whew*
18:12:06 <ke4qqq> podling has links to both svn and git
18:12:16 * nslater facepalms
18:12:24 <nslater> cool
18:12:35 <nslater> so i am guessing this stuff has already cleared the incubator's ip stuff
18:12:46 <nslater> either that of brett messed something up. which i doubt
18:12:53 <ke4qqq> nslater: it is - but a lot of the signoff stuff I didn't feel comfortalbe checking, particularly after the sga stuff
18:12:53 <chipc> Can I suggest an action to make sure that the orig submission is ok with ip clearance?
18:12:58 <nslater> what is SGA?
18:13:07 <ke4qqq> software grant agreement
18:13:20 <nslater> chipc: it should be noted on your podling page. checking
18:14:28 <nslater> hmm. my newbness is showing again
18:14:32 <nslater> i think the ip clearence stuff is just for code seperate to the main project code
18:14:40 <nslater> might be worth checking
18:14:48 <nslater> ill post something to the list
18:15:23 <chipc> is there an action in there for nslater?
18:15:36 <ke4qqq> fix all our IP problems? :)
18:15:39 <chipc> ha
18:15:42 <rohityadav> :)
18:15:47 <jlkinsel> please, thanks. ;)
18:16:09 <chipc> #action nslater to review the state of the orig cloudstack code submission (IP clearance and SGA)
18:16:12 <chipc> how's that?
18:16:43 <jlkinsel> looks good to me
18:16:45 <nslater> sent
18:16:48 <nslater> and yes
18:17:18 <chipc> jlkinsel: Is that a wrap?
18:17:24 <jlkinsel> i believe so
18:17:39 <jlkinsel> thanks everybody for attending. I'll post minutes/logs to the ml and wiki
18:17:42 <topcloud> thanks everyone.
18:17:42 <jlkinsel> #endmeeting

  • No labels