[00:15] <manusheel> dfarning: Hi David.
[00:15] <dfarning> manusheel,  hello, how are you?
[00:16] <manusheel> dfarning: Good. Back to work.
[00:17] <manusheel> dfarning: Thank you for sharing the links.
[00:17] <manusheel> dfarning: Going through them.
[00:18] <dfarning> manusheel, I took a nap this after noon so I could stay up later tonight if needed.
[00:19] <manusheel> dfarning: Great. Sure, David. I hope you are also taking fine rest at night. Important for long term productivity.
[00:20] <dfarning> dfarning, other wise I get grumpy and don't think clearly:(
[12:29] <manusheel> kandarpk: Hi Kandarp.
[14:50] <kandarpk> bernie: there ?
[16:09] <manusheel> kandarpk: Wish to ask you whether you have completed the five issues in reference to Sugar on Debian.
[16:09] <kandarpk> manusheel sir: the reading issue could not be solved
[16:10] <kandarpk> dfarning said we need to ask debian people to update the package
[16:10] <kandarpk> so that it can be downloaded
[16:19] <dfarning> kandarpk, I think our biggest blocker is you and neeraj learning to the debian/git packaging process from dipankar.
[16:20] <kandarpk> manusheel sir: I am learning git usage right now
[16:20] <manusheel> kandarpk: Ok. I'll communicate this to the Debian person.
[16:20] <kandarpk> just need to understand branching
[16:20] <manusheel> kandarpk: Sure, Kandarp. Please get an understanding on git.
[16:21] <manusheel> dfarning: Dipankar has been facing internet issues. I am not sure why you referred to it as a blocker. Dipankar is one of the most consistent team members here.
[16:24] <dfarning> manusheel, Did not mean that as a criticism or either dipankar of anyone else.  It is just the most important task.  Any criticism goes to me for not being aware Luke was leaving so soon.
[16:27] <manusheel> dfarning: Completely agree. It is indeed the most important task right now. No, I didn't mean it as a criticism. We generally don't use "blocker" for a delay.
[16:27] <manusheel> dfarning: And, that too, not under Dipankar's control.
[16:29] <manusheel> kandarpk: What all have we covered in git packaging uptill now, in terms of concepts?
[16:30] <kandarpk> manusheel sir: not packaging but usage of git in general
[16:30] <dfarning> manusheel, ahh blocker must have a much more negative connotation then I thought.
[16:30] <kandarpk> manusheel sir: I've understood initiating a local repo
[16:31] <manusheel> kandarpk: Ok, sure. Do you have sufficient resources on understanding git? Let me send you a couple of tutorials too.
[16:31] <kandarpk> working on files, commiting changes, making branches
[16:31] <kandarpk> manusheel sir: I am using tutorials from git official site
[16:32] <kandarpk> *they are quite good
[16:32] <manusheel> dfarning: Absolutely fine.
[16:32] <manusheel> kandarpk: Neat. Appreciate it.
[16:34] <kandarpk> dfarning, Hi
[16:34] <dfarning> kandarpk, Hello
[16:35] <kandarpk> dfarning: got some understanding of git
[16:35] <kandarpk> wanted to share with you
[16:35] <kandarpk> let me know if I am wrong somewhere
[16:35] <dfarning> kandarpk,  sounds good.
[16:36] <kandarpk> dfarning: I think we make branches so that in one branch ( say Mainstream )
[16:37] <kandarpk> we have got all the things going and all sources working
[16:37] <kandarpk> and another branch ( say Devel )
[16:37] <kandarpk> which is used for development of a project
[16:37] <kandarpk> and when we are satisfied with the final code in Devel
[16:38] <kandarpk> we merge it with Mainstream
[16:38] <kandarpk> is it ok ?
[16:38] <kandarpk> dfarning: any other uses of branching  ?
[16:39] <dfarning> kandarpk, yes, that is correct and we will use branching for one other purpose
[16:41] <dfarning> we will keep a number of permanent branches:  debian (mainstream) -> ubuntu 10.04 -> ubuntu 10.10
[16:41] <kandarpk> yeah
[16:41] <dfarning> the ubuntu branches will hold ubuntu specific patches for each release
[16:42] <kandarpk> realise it now
[16:43] <dfarning> but the very cool thing is that if any one commits a fix to debain (mainstream) we can pull in to either of the ubuntu branches.
[16:43] <kandarpk> yeah.
[16:44] <kandarpk> Things becoming very clear now
[16:44] <dfarning> kandarpk, this is why it is not worth it to fix bugs in the current ubuntu packaging..... because we will just have to fix it again upstream in debain.
[16:45] <dfarning> kandarpk, better to fix it once upsteam in debian and pull to the branches.
[16:46] <kandarpk> dfarning: definitely.
[16:46] <dfarning> kandarpk, my problem is that I don't know to debian packaging process well enough to explain them.
[16:46] <kandarpk> dfarning: thats no problem
[16:47] <dfarning> Jonas the olpc-debian maintainer invented the system himself so there is no documentation yet:)
[16:47] <kandarpk> dfarning: your guidance is very helpful
[16:49] <kandarpk> dfarning: what happens if we merge two branches having different files as well ?
[16:51] <dfarning> kandarpk, this is where git is so clever.  It can automatically detect changes in both the origin and destation branches.  As long the the changed don't directly conflict it will apply changes as necessary.
[16:52] <kandarpk> dfarning: I mean, if in branch 1 I have file f1
[16:52] <kandarpk> and in branch 2 i have file f1 and f2
[16:52] <kandarpk> what will happen if we try to merge them ?
[16:55] <dfarning> kandarpk, please take a look at the webcast on http://book.git-scm.com/3_basic_branching_and_merging.html it explains it better than I can.
[16:55] <kandarpk> Ok.
[17:01] <dfarning> dipankar, hello
[17:02] <dipankar> dfarning, hello. Sorry about yesterday. My net was down.
[17:02] <dfarning> dipankar, no problem at all.  Some things we can't fix:)
[17:03] <dipankar> dfarning, yeah. So what is the topic today?
[17:03] <dipankar> dfarning, * I had a very tiring day today. Hope you don't mind if I leave early today
[17:04] <dfarning> dipankar, no problem
[17:05] <dipankar> dfarning, tomorrow we are having bug bashing. Manu Sir must have told you.
[17:05] <dfarning> dipankar, do  you think you could explain to kandarpk and I how to build a debain package using git?
[17:06] <dipankar> dfarning, I will try my best to tell you guys what Luke explained to me
[17:06] <dipankar> :)
[17:06] <dfarning> yes, dipankar I am starting to understand and respect the methodical workflow and QA you guys use.
[17:07] <dfarning> dipankar, thanks
[17:07] <dipankar> dfarning, :)
[17:08] <dipankar> dfarning, But I would suggest following the IRC log side by side
[17:09] <dfarning> dipankar, ok i'll start rereading the logs.
[17:10] <dfarning> dipankar, in the mean time could you try packaging/updating another activity in the git repo?
[17:11] <kandarpk> dipankar: if you try, let me know
[17:12] <dipankar> dfarning, How about I tell you and Kadarp sir to update a package? I guess that way I can revise things too.
[17:13] <dfarning> dipankar, I am not sure shat you mean.
[17:13] <dipankar> dfarning, I meant : I tell you and Kandarp Sir the steps in updating a package and you try the steps out on your machines.
[17:15] <dfarning> dipankar, we can all do the same steps on our local machine.... just not the final push.
[17:16] <dipankar> dfarning, yes. Good idea.
[17:17] <dipankar> dfarning, which package do you want me to work upon?
[17:18] <dfarning> dipankar, which ever looks easiest.  We can do the complicated stuff later.
[17:18] <dipankar> dfarning, ha ha! But I don't know which package needs what modification :( Luke only told what to change.
[17:21] <dfarning> how about a simple activity that is already pacakged that needs a newer verison from ASLO
[17:22] <dipankar> dfarning, kandarpk : I guess we can start with the logviewer-activity :) That file needed a control.in file made.
[17:22] <dfarning> dipankar, sounds good
[17:22] <dipankar> dfarning, ^^ That was my task actually. :P I guess we can work on it right now.
[17:23] <kandarpk> dipankar, sure
[17:24] <kandarpk> dipankar: but please start with some background to control.in
[17:24] <kandarpk> I read about it
[17:24] <kandarpk> but not very certain
[17:25] <dfarning> kandarpk, https://wiki.ubuntu.com/PackagingGuide/Complete#control
[17:26] <dipankar> kandarpk, please also read through this irc : http://irclogs.ubuntu.com/2010/07/01/%23ubuntu-sugarteam.html
[17:26] <dfarning> kandarpk, it is where we set the variable which determine how the package is built.
[17:28] <dipankar> kandarpk, adding to David, it also contains the build dependencies without which the packaging cannot be complete
[17:28] <kandarpk> dipankar: I think there was some difference b/w control and control.in
[17:28] <dfarning> dipankar, most of the stuff is pretty standard like the package name source url
[17:29] <dfarning> opps wrong person:(
[17:29] <dipankar> dfarning, ? Didn't get you.
[17:30] <dfarning> kandarpk, the part we will worry about the most is the various dependancies-- build time and run time:
[17:30] <dfarning> dipankar, i met to alert kandarpk no you a couple of mesages ago.
[17:31] <dfarning> dipankar, did luke explain the difference between control and control.in?
[17:31] <dipankar> dfarning, gee.. I forgot my self. Let me see the log.
[17:31] <dipankar> :(
[17:32] <kandarpk> dipankar: you mean there is some difference, right ?
[17:33] <dfarning> kandarpk, dipankar this part is kind of confusing.
[17:33] <dipankar> dfarning, kandarpk : there is a difference. control file can be regenerated from a control.in file
[17:33] <dfarning> normally when you run debbuild it uses the contents of control.
[17:34] <kandarpk> dipankar: I believe we wont require debuild in git
[17:34] <kandarpk> or is it ?
[17:35] <dipankar> kandarpk, nope. The command for git package building is : 'git-buildpackage'
[17:35] <kandarpk> hmmm
[17:36] <dipankar> kandarpk, till now I have never used it got git repo
[17:36] <dfarning> but, as dipankar just pointed out when useing the debgit tool the git-buildpackage first parses the  control.in to regenerate the control file.
[17:37] <kandarpk> dipankar: can you please start with how file are changed
[17:38] <dfarning> dipankar, this give the package the ability to do some fancy abstraction at buildtime
[17:38] <kandarpk> instead of telling which commands to use
[17:38] <dipankar> ohk. I am pasting the required IRC transcript.: http://paste.ubuntu.com/458814/. This will help a bit regarding control.in and control file
[17:40] <kandarpk> Ok.
[17:41] <kandarpk> dfarning, dipankar: how does maintaining a separate control file helps if all changes can be generated from control.in ?
[17:41] <dfarning> dipankar, thanks I did not realize that we commited the control file too.
[17:42] <dfarning> kandarpk, I was wondering the same thing.  One traditionally does not commit stuff which can be dynamically recreated:(
[17:42] <dipankar> dfarning, :)
[17:42] <dipankar> kandarpk, This might help : http://www.debian.org/doc/maint-guide/ch-dreq.en.html
[17:43] <dfarning> kandarpk, must be a quirk
[17:44] <dfarning> dipankar, there is no way I will remember 'DEB_MAINTAINER_MODE=1 fakeroot debian/rules clean' two days from now:(
[17:45] <kandarpk> dfarning, dipankar: I think control.in has CDBS scripts which automatically detects dependencies
[17:45] <kandarpk> *as Luke mentioned
[17:45] <kandarpk> and then places those in control file
[17:45] <kandarpk> *wild guess
[17:45] <dipankar> dfarning, me too, That command is ugliest I have ever seen.
[17:46] <kandarpk> dfarning: can that be a possibility ?
[17:46] <dfarning> dipankar, I think I now under stand the difference.
[17:46] <dipankar> kandarpk, dfarning : Actually I am very very tired right now. I might not act quickly and with full attention
[17:46] <dipankar> :(
[17:47] <dfarning> kandarpk, yes that sounds correct.
[17:47] <dfarning> dipankar, do you want to start again tomowwor or monday?
[17:48] <kandarpk> and as Luke said, pushing both of them lets you know what changes were made by you, and what made by CDBS
[17:48] <dfarning> I would like if we could set a time so kandarpk and I can both be there if possiable?  I can work around your scheuddal.
[17:48] <kandarpk> dfarning, dipankar: I don't want to leave here
[17:49] <kandarpk> please just for another 25-30 minutes
[17:49] <dfarning> dipankar, are you up for another 30 minutes:)
[17:50] <dipankar> dfarning, kandarpk : I will try my best.
[17:51] <dfarning> the command to regenerate the control file is shown on http://wiki.debian.org/Sugar/GettingStartedGuide
[17:51] <dfarning> dipankar, what next:)
[17:52] <dipankar> dfarning, we have to start off yet
[17:52] <dipankar> :)
[17:52] <kandarpk> dipankar, dfarning : we'll need to modify only control.in ?
[17:53] <dipankar> not quite kandarpk. We have to actually make it.
[17:53] <kandarpk> dipankar: yeah, that is what I meant
[17:54] <kandarpk> but only the control.in file, right ?
[17:54] <dipankar> kandarpk, : yup! :)
[17:54] <dipankar> kandarpk, dfarning : First start by moving to a suitable directory.
[17:54] <dipankar> (in terminal)
[17:55] <dfarning> ok, that I can do pretty well:)
[17:56] <dipankar> :P
[17:56] <manusheel_> dfarning : :-)
[17:56] <kandarpk> dipankar: next ?
[17:58] <dipankar> run the command: 'apt-get source sugar-logviewer-activity'
[17:58] <kandarpk> dipankar: instead of telling the commands
[17:58] <kandarpk> *that might take long
[17:58] <kandarpk> please let us know what files need changing
[17:59] <kandarpk> just as an overview
[17:59] <kandarpk> dfarning: will that be ok ?
[18:00] <dipankar> kandarpk, actually I am guiding in setting up a local git repo in the directory using a git repo on ssh. That is crucial.
[18:00] <dipankar> dfarning, ^^
[18:00] <kandarpk> dipankar: Ok.
[18:00] <dfarning> ok
[18:00] <kandarpk> sorry for interrupting
[18:00] <dipankar> after that just some file changes and committing :)
[18:00] <dipankar> not a problem :P
[18:01] <dipankar> also try running 'debcheckout sugar-logviewer-activity'
[18:01] <dipankar> see if it gives an error
[18:01] <kandarpk> apt-get source sugar-logviewer-activity didn't run
[18:02] <kandarpk> ERROR:
[18:02] <kandarpk> E: You must put some 'source' URIs in your sources.list
[18:03] <dfarning> dipankar, kandarpk are you doing this on your normal ubuntu 10.04 machines?
[18:03] <kandarpk> dfarning: I am using Debian
[18:03] <dipankar> dfarning, yes. I do packaging from my laptop.
[18:04] <kandarpk> *git is used for debian only ?
[18:04] <dfarning> kandarpk, you will need to go into synatic and enable the source repos.
[18:05] <dfarning> kandarpk, will be using it for both ubuntu and debain... with a branch for each.
[18:05] <kandarpk> dfarning: there is no sources in repositories list ...
[18:06] <kandarpk> dfarning: to add new, I'll need a URI
[18:07] <dfarning> kandarpk, isn't there a place to tick 'sources' in synaptic -> setting -> repoitories.
[18:07] <kandarpk> dfarning: I mean, various repositories are listed, but none as sources
[18:08] <dfarning> kandarpk, let's let dipankar go he was not feeling well yesterday
[18:09] <kandarpk> dipankar: please wrap it up then
[18:09] <kandarpk> dipankar: in say 2-3 minutes
[18:09] <dipankar> :)
[18:09] <dipankar> dfarning, kandarpk : The main thing here is importing the source code.
[18:10] <kandarpk> dipankar: and do tell something which keeps us busy for next hour
[18:10] <dfarning> kandarpk, let's start this again later and all use ubuntu 10.04... so our systems are as similar as possiable.
[18:10] <dipankar> and then setting up a local repo for the package
[18:10] <kandarpk> dipankar: setting up local hier.. will be OK
[18:11] <kandarpk> what next
[18:11] <kandarpk> ?
[18:13] <dipankar> dfarning, kandarpk : then you have to make a control.in file with the following things:
[18:13] <dipankar> lfaraone	dipankar: then, I'd like you to clone the sugar-terminal-activity.git repository using the "git+ssh" url listed on the page you're seeing now. Edit the debian/control.in file and update the description of the package using similar text to what is used in the description of http://git.debian.org/?p=collab-maint/sugar-calculate-activity.git;a=blob;f=debian/control.in;h=ae9a2b9989f80377c248ca774c869e0f052d7242;hb=dc85b18722eaac7f0457597a6
[18:13] <dipankar> a90b83
[18:14] <dipankar> http://git.debian.org/?p=collab-maint/sugar-calculate-activity.git;a=blob;f=debian/control.in;h=ae9a2b9989f80377c248ca774c869e0f052d7242;hb=dc85b18722eaac7f0457597a6a90b83
[18:14] <dipankar> dfarning, kandarpk : use the link above for the control.in file
[18:15] <dipankar> once you are done with the control.in file then you can commit the changes to the local repo.
[18:15] <kandarpk> ok
[18:16] <dipankar> after that regenerate the control file from the control.in file and commit the changes again
[18:16] <dipankar> then the final task is to push to git ssh repo
[18:17] <kandarpk> dipankar: we didn't make any changes to source code ?
[18:17] <dfarning> dipankar, so we are just cutting and pasting the contents of the control.in for calulate to our new logviewer package.  And then modifing the control.in file as necessary
[18:17] <dipankar> kandarpk, dfarning : when I asked Luke about it, he told me that this activity is available on ubuntu server
[18:18] <kandarpk> dipankar: isn't that as if we didn't change anything
[18:18] <dipankar> so what we are doing here is moving the activities to Debian
[18:18] <dipankar> kandarpk, I agree with you.
[18:19] <kandarpk> dipankar: there's a difference
[18:19] <dipankar> The task right now is to make packages according to Debian guidelines
[18:19] <kandarpk> for debian, we require control.in
[18:19] <kandarpk> dipankar: right.
[18:19] <dfarning> kandarpk, ok so this is a case where we have already done the basic packaging work in ubuntu.  now we just need to modify it slight and upload it to debain.
[18:20] <dipankar> dfarning, You told me a couple of days back that some packages were made in shortcut and they don't follow the guidlines
[18:21] <kandarpk> dfarning, dipankar : I think control.in will check for dependencies for the package on Debian
[18:21] <dipankar> *PS: I meant to say, after the activities are upon the Debian, then we can use it both for Debian & Ubuntu in a single go
[18:21] <kandarpk> in ubuntu we manually specified that
[18:21] <kandarpk> dipankar: am I right ?
[18:22] <dfarning> dipankar, I would not trust any packages _I_ made,  I think luke set this up and an example for us.
[18:22] <dipankar> kandarpk, Like I said I am out of senses right now :P. I will go with what you are suggesting right now :)
[18:23] <dipankar> I will look deeply into the control file tomorrow morning
[18:23] <kandarpk> dipankar: lets wait for luke then
[18:23] <kandarpk> he has better expertise
[18:23] <dipankar> yup.
[18:23] <manusheel_> kandarpk: Let me ask Luke if he can join us now.
[18:23] <manusheel_> dipankar: Thank you.
[18:24] <manusheel_> Appreciate your support.
[18:24] <kandarpk> manusheel_ sir: that will help
[18:24] <dipankar> dfarning, But your packages work don't they?
[18:24] <dfarning> dipankar, yes they work, but they are not up to the debian standard.
[18:25] <kandarpk> dfarning: why is that ?
[18:25] <dipankar> dfarning, that's why Luke is suggesting the changes..
[18:25] <bernie> kandarpk: did you call?
[18:25] <kandarpk> bernie: hi
[18:25] <bernie> kandarpk: sorry that my comments reached you much later... I was offline yesterday. working from schools is HARD
[18:25] <kandarpk> thanks for being considerate
[18:26] <dfarning> debain is very strict about enforcing certain standards for technical and legal reasons.
[18:26] <dipankar> dfarning, thats new.
[18:26] <kandarpk> dfarning: leaving the legal reasons for now,
[18:26] <dipankar> dfarning, I have a Debian system now too!
[18:27] <kandarpk> dfarning: how are they ( packages you make for ubuntu )
[18:27] <kandarpk> not technically right for debian
[18:27] <kandarpk> ?
[18:27] <dfarning> kandarpk, I just wanted to get a large set of activities available for you to see when you started using sugar... and would not be disapointed:)
[18:27] <dipankar> manusheel sir,  I guess Luke is on Mobile. He might not come on IRC
[18:28] <kandarpk> dfarning: no, I mean
[18:28] <bernie> kandarpk: hehe... welcome to the devian VS ubuntu diatribe
[18:28] <kandarpk> whats the difference
[18:28] <kandarpk> bernie: :)
[18:28] <manusheel> dipankar: You are right. Luke is on mobile.
[18:28] <dipankar> bernie, Hello.
[18:29] <dipankar> bernie, I like this kind of fight (OS vs OS)
[18:29] <dipankar> all, It was earlier used to Linux vs Windows.
[18:29] <kandarpk> bernie, dipankar : I am not comparing here
[18:29] <dfarning> dipankar, bernie likes these types of arguments too.... he can go on for days.
[18:30] <dipankar> now its Debian vs Ubuntu : and I support Ubuntu
[18:30] <manusheel> dipankar: No.
[18:30] <manusheel> thats not true.
[18:30] <kandarpk> dfarning, dipankar, bernie : no Ubuntu VS Debian right now
[18:30] <manusheel> dipankar: We need to support both, and you need to be a master in both.
[18:30] <kandarpk> let me know the difference
[18:30] <bernie> kandarpk: be warned that, behind the technicalities of policy and quality, there's often a hidden argument of the two communities not getting along too well.
[18:31] <manusheel> dipankar: We should not worry about Debian vs Ubuntu. We should have Sugar on both Debian and Ubuntu.
[18:31] <bernie> dfarning: hehe
[18:31] <manusheel> bernie: Hi Bernie. How have you been?
[18:31] <bernie> manusheel: indeed
[18:31] <manusheel> bernie : +1
[18:32] <dfarning> dipankar, from our point of view the only differences are debian is upstream to ubuntu and the have slightly different release cycles.
[18:32] <kandarpk> *Power cut
[18:32] <kandarpk> *will be back
[18:32] <bernie> manusheel: I've been working from the field for 3 days in a row, now I need two days of sleep to recover.
[18:32] <dfarning> ubuntu releases every 6 months and debain releases less frequetly about every two years.
[18:32] <kandarpk> bernie: thats bad news for me :(
[18:33] <dfarning> there are pros and cons to both.
[18:33] <manusheel> kandarpk: Thats not a bad news.
[18:33] <bernie> wow kandarpk was so scared he dropped off
[18:33] <manusheel> bernie: You seem to have the same lifestyle as always :-)
[18:33] <bernie> ;-)
[18:33] <manusheel> Take a break.
[18:33] <bernie> manusheel: haha. "life".
[18:33] <bernie> what is it?
[18:33] <manusheel> bernie: When are you going to Brazil?
[18:34] <bernie> manusheel: 8-18 Jul... but it seems I won't go to Salvador any more
[18:34] <dipankar> dfarning, got it!
[18:34] <bernie> manusheel: the plane from belo horizonte was $360 so we opted for a bus to Rio
[18:35] <dfarning> dipankar, thanks for you help I need to get some lunch
[18:35] <bernie> manusheel: I already feel guilty for leaving 10 days just before the final release of F11-0.88... we still have many bugs to fix
[18:36] <dipankar> dfarning, no problem :)
[18:36] <manusheel> We do need a stable 0.88 release.
[18:36] <manusheel> bernie: Indeed.
[18:37] <dipankar> dfarning, enjoy your lunch
[18:37] <dipankar> manusheel Sir, bernie : I am heading to bed. Good night everybody!
[18:37] <manusheel> bernie: Interesting. I hope you figured out a good bus. I remember some of our friends at MIT experimented with some Chinese bus from MIT to New York. And, they were scared with the bus routes.
[18:38] <manusheel> dipankar: Good night Dipankar.
[18:39] <kandarpk> hi all, I missed out on anything ?
[18:39] <manusheel> kandarpk: Glad that you are back.
[18:40] <kandarpk> dfarning: please let me know about that packaging issue
[18:41] <bernie> manusheel: haha I got the Fung Wa bus too...
[18:42] <bernie> manusheel: until cjb told me "they have a habit of catching fire on the highway"... and then we googled for it and there were indeed, like, 10 accidents some of which worse than just catching fire
[18:42] <bernie> such as loosing the two back wheels
[18:42] <manusheel> bernie: Yes, friends told me that it is not safe to travel through that bus.
[18:42] <bernie> or crashing into the guardrail between the highway and an exit!!!
[18:42] <manusheel> bernie: Loosing two back wheels?
[18:42] <manusheel> :-)
[18:43] <bernie> manusheel: http://en.wikipedia.org/wiki/Fung_Wah_Bus_Transportation#Incidents_and_accidents
[18:43] <bernie> lol
[18:43] <bernie> anuary 3, 2007: In Framingham, Massachusetts, a New York-bound bus lost its back two wheels. No injuries were reported. [9]
[18:44] <manusheel> bernie: Looks like the buses are developed by some high school students :-) One cannot cut down costs like that.
[18:44] <manusheel> :-)
[18:44] <bernie> the plane is definitely safer
[18:44] <manusheel> kandarpk: It seems David has head for lunch.
[18:44] <kandarpk> :(
[18:44] <manusheel> bernie: Much safer.
[18:44] <kandarpk> bernie: you like OS vs OS
[18:45] <manusheel> kandarpk: No worries.
[18:45] <kandarpk> bernie: please explain the differences in Ubuntu and Debian
[18:45] <kandarpk> on technical grounds
[18:46] <manusheel> bernie: Around?
[18:46] <bernie> manusheel: yes
[18:46] <manusheel> bernie: Let us discuss Ubuntu vs Debian on technical grounds.
[18:46] <manusheel> Very good question by Kandarp.
[18:47] <bernie> kandarpk: I like to read the OS vs OS and language VS language politics in the news... but it's not fun when you get in the middle.
[18:47] <bernie> kandarpk: and you can't get your job done
[18:47] <manusheel> kandarpk: Yes, that is a very important question. Bernie, Kandarp is talking about it on the technical level.
[18:47] <manusheel> Not at policy level.
[18:48] <kandarpk> bernie: dfarning said his packages for Ubuntuwont work on Debian
[18:48] <manusheel> kandarpk: Where Debian does well, and where Ubuntu does well.
[18:48] <kandarpk> manusheel sir: No
[18:48] <manusheel> bernie: We are interested in Sugar on Debian and Sugar on Ubuntu.
[18:48] <manusheel> kandarpk: Ok.
[18:48] <kandarpk> right now I only want to know why dfarning's packages wont work on Debian
[18:49] <manusheel> kandarpk: Ok, great. bernie, do you have an answer on this?
[18:49] <dfarning> kandarpk, my packages probably would work on debian.
[18:50] <kandarpk> dfarning: Ok, yours would
[18:50] <dfarning> kandarpk, I just didn't cross all of the tee and dot all of the eyes so that they would be accepted.
[18:51] <kandarpk> but where is Debian so strict ?
[18:51] <kandarpk> the source codes ?
[18:51] <dfarning> kandarpk, my package were very rough drafts.  But you would not want to hand them in.
[18:51] <kandarpk> or the dependencies ?
[18:52] <dfarning> kandarpk, I'll try to answer that in an email this afternoon.
[18:52] <kandarpk> dfarning: that will be great
[18:52] <kandarpk> dfarning: thanks a lot
[18:53] <dfarning> The dynamics are confusing
[18:54] <bernie> kandarpk: I don't know the details, but ABI and API issues, probably.
[18:55] <bernie> kandarpk: Ubuntu and Debian diverged quite a lot in the past, and there doesn't seem to be a tendency to realign one or the other on the same versions of libraries and critical system daemons.
[18:55] <kandarpk> bernie: that means they differ on the dependencies ?
[18:55] <kandarpk> sources are almost same.
[18:56] <bernie> kandarpk: so, for example, ubuntu uses Upstart while Debian still uses sysvinit (unless I'm out of date). which means that daemons need to install their startup script in different places!
[18:57] <bernie> kandarpk: in general, debian is more conservative. they ship older versions of libraries and they tend not to jump on modern new subsystems as quick as ubuntu does
[18:57] <bernie> kandarpk: from the "plumbing" point of view, Ubuntu today resembles more Fedora than Debian: pulseaudio, udev, udisks, networkmanager, libvirt, packagekit, rsyslog, upstart, latest Xorg, etc.
[18:58] <kandarpk> dfarning, bernie : its a bit confusing right now
[18:58] <kandarpk> dfarning's mail will help
[18:59] <bernie> kandarpk: in other words, it's not really clear any more which project is downstream and which is upstream. Ubuntu now has newer stuff than debian almost at every level, so the flow has started to go backwards.
[18:59] <kandarpk> Ok.
[18:59] <bernie> kandarpk: that is, several packages get imported from ubuntu into debian nowadays.
[19:00] <kandarpk> bernie: yeah, I understand that
[19:01] <bernie> (which tends to piss off people who were accustomed to the opposite work-flow)
[19:01] <manusheel> bernie: David said that it was very difficult to get packages at Ubuntu.
[19:01] <kandarpk> *at Debian
[19:02] <bernie> manusheel: besides this gossip, I have to admit I don't have much hands-on experience getting packages accepted both in Debian and Ubuntu
[19:02] <manusheel> bernie: What should be the good strategy in general?
[19:02] <manusheel> bernie: Ok.
[19:02] <bernie> manusheel: historically, it used to be the case that becoming a debian maintainer was one of the hardest achievements of a hacker's career. it might have changed now :-)
[19:03] <manusheel> bernie: Ok, we'll have to get into the field and realize that.
[19:03] <manusheel> bernie: In any case, we are looking at both Sugar on Ubuntu and Sugar on Debian. So, we need to be proficient in both.
[19:03] <kandarpk> bernie: is that because Debian has become soft ?
[19:04] <manusheel> bernie: Which means, whether it is syncing packages from Debian to Ubuntu or the other way round.
[19:04] <bernie> manusheel: the debian maintainer for sugar is still jonas?
[19:04] <bernie> manusheel: who's the ubuntu maintainer? nobody?
[19:04] <lfaraone> bernie: becoming a Debian Developer, you mean.
[19:05] <manusheel> bernie: Right. He is the maintainer. We'll have to work with him in these regards.
[19:05] <kandarpk> lfaraone: Hi
[19:05] <bernie> kandarpk: lfaraone would know, he just became a debian maintainer
[19:05] <lfaraone> bernie: it still is a very rigorous process.
[19:05] <bernie> lfaraone: ah, are the two things different?
[19:05] <manusheel> bernie: And, are happy to do so.
[19:05] <lfaraone> bernie: Debian Developer > Debian Maintainer.
[19:05]  * lfaraone is in town for a few minutes :)(
[19:05] <lfaraone> ?away
[19:06] <manusheel> lfaraone: Great.
[19:06] <manusheel> How has been your trip?
[19:06] <lfaraone> bernie: DMs are people who can upload certain packages that have been marked as "DM-Upload-Allowed: yes" and have the DM listed in "uploaders" or "maintainer"
[19:06] <lfaraone> manusheel: it's been pretty great.
[19:06] <manusheel> lfaraone: Wonderful.
[19:07] <lfaraone> bernie: DDs can technically upload any package they wish. DDs can also vote in the project, and are full Debian members.
[19:07] <manusheel> lfaraone: Kandarp had questions on packaging.
[19:07] <manusheel> lfaraone : That is, using git for packaging.
[19:07] <bernie> lfaraone: ah, it's the equivalent of Fedora's überpackager
[19:08] <lfaraone> bernie: To become a DM, you only need a DD to sign off, a signature on your key, and no objections. DD need an endorsement from an extant DD, and have to go through the confusingly named "new maintainer" process.
[19:08] <lfaraone> kandarpk: okay, what's confusing?
[19:08] <lfaraone> (re ease of getting packages in Ubuntu, it's not that hard to get a package in Ubuntu or Debian, we just prefer to get it into Debian so the results can flow downstream)
[19:08] <kandarpk> lfaraone: for packaging in Debian, what steps are involved after generating the control file ?
[19:08] <bernie> lfaraone: what' the typical workflow of git package maintenance for debian/ubuntu?
[19:09] <bernie> lfaraone: are there a bunch of command line tools to manage the vendor branch and the debian patches?
[19:09] <lfaraone> bernie: make changes, commit, push. prepare changelog for vendor version, release.
[19:09] <kandarpk> lfaraone: dipankar said he only generated control file before pushing
[19:09] <lfaraone> bernie: not as far as I know.
[19:10] <lfaraone> kandarpk: well, you don't need to regenerate the control file unless you've done something that would cause changes.
[19:10] <kandarpk> lfaraone: changes to the source you mean ?
[19:11] <bernie> lfaraone: the problem with getting packages in debian first is that the release cycles are now irremediably skewed. Going through debian would mean that ubuntu would always lag 1-2 releases behind Sugar's latest release.
[19:11] <lfaraone> kandarpk: well, a subset of that: if you change something that would cause the control file differently.
[19:11] <lfaraone> bernie: why do you think that? 0.88 is in Debian Unstable and Testing.
[19:11] <bernie> lfaraone: didn't you tell me some time ago that jonas was using some really fancy stuff to maintain the sugar packages in git?
[19:12] <lfaraone> bernie: well, git-buildpackage.
[19:12] <bernie> lfaraone: since when is 0.88 in debian unstable? would it have made it in time for lucid?
[19:12] <kandarpk> lfaraone: we need to generate the control.in ourself, right ?
[19:12] <lfaraone> bernie: it is in Lucid...
[19:12] <bernie> lfaraone: and, more importantly, would have it worked?
[19:12] <lfaraone> kandarpk: well, you need to write it, yes.
[19:12] <lfaraone> bernie: it does work right now. the issue is that Browse does not work.
[19:13] <lfaraone> bernie: that is due to a local Ubuntu problem where Canonical (AFAICT) does not wish to support programs which embed XULRunner.
[19:13] <bernie> lfaraone: my feeling is that the two systems have diverged too much to expect any non-trivial package set like sugar to work on both without some amount of testing and bugfixing
[19:14] <lfaraone> bernie: of course you'll need to test your work in Debian and Ubuntu.
[19:14] <kandarpk> lfaraone: so, after writing the control.in, we just need to push the package into debian repos ?
[19:14] <bernie> lfaraone: if there's just a minor issue, why is there a need to host so many ubuntu-specific packages in the Ubuntu Sugar Team PPA?
[19:14] <lfaraone> bernie: there is no need, it's not something I like that we do.
[19:14] <bernie> lfaraone: (not a rhetorical question, I really don't know for sure)
[19:15] <lfaraone> bernie: we provide point releases in the PPAs. Ideally, we can get these pushed as stable release updates if we had well-defined usecases and approval from the Ubuntu Release Managers to do so.
[19:16] <lfaraone> bernie: the main utility of the PPA is providing the "sugar-activities" package, which is a collection of precompiled activities taken from ASLO. It's probably illegal, since it's not GPL-compliant, but nobody really cares to sue us.
[19:17] <lfaraone> bernie: right now, OOTB, two sugar activities packaged in the Ubuntu repositories work without any PPAs. The other nonfunctional ones are locally-Ubuntu-packaged from what I can tell.
[19:17] <bernie> lfaraone: why not GPL compliant? there are no src debs?
[19:17] <lfaraone> bernie: the "source deb" is a colletion of XO bundles.
[19:18] <bernie> lfaraone: ah, they're not really being built from sources... I wonder what the fedora folks do... but who cares. we're not going to be sued, as you said
[19:18] <lfaraone> bernie: and those contain precompiled binaries.
[19:18] <lfaraone> bernie: well, as it is we can't include sugar-activities in Debian and Ubuntu.
[19:19] <lfaraone> bernie: the fedora folks package each activity individually, which is what I'm trying to teach us to do.
[19:19] <bernie> lfaraone: are you sure sugar should be working flawlessly *today* on ubuntu except for the Browse activity? I just spoke with a teacher trainer last week who tried to install it on Lucid and could not even get the emulator to staet.
[19:19] <bernie> start
[19:20] <lfaraone> bernie: please hold.
[19:20] <bernie> lfaraone: back to the GPL compliance thread: I think packaging from source is not that hard... most of them are just "python setup.py" projects.
[19:20] <bernie> lfaraone: ok
[19:20] <lfaraone> bernie: http://git.debian.org/?p=collab-maint/sugar-terminal-activity.git is an example of my work in packaging a certain activity. Once we are able to get a working package, we should be able to trivally able to apply it to related packages.
[19:21] <bernie> lfaraone: also, it should be easy to copy whatever the fedora spec files are doing.
[19:21] <lfaraone> bernie: it's not that hard, but there are some things (like writing debian/copyright) and wrangling CDBS that are hard to get right. debian/copyright will be unique, but CDBS should be the same once we get it right.
[19:21] <lfaraone> bernie: we have what they're doing, it's CDBS's python-sugar.mk.
[19:22] <bernie> oh, /usr/share/cdbs/1/class/python-sugar.mk does all the magic
[19:23] <lfaraone> bernie: yep. we just need to tell it what branches are supported.
[19:27] <kandarpk> lfaraone: can you please provide some overview of what all is pushed in the debian repository
[19:27] <kandarpk> complete package or only the patches
[19:30] <lfaraone> .bernie: bernie http://article.gmane.org/gmane.linux.ubuntu.devel.sugar/6
[19:30] <lfaraone> bernie: see http://article.gmane.org/gmane.linux.ubuntu.devel.sugar/6 and ed's followup: http://article.gmane.org/gmane.linux.ubuntu.devel.sugar/7
[19:31] <lfaraone> kandarpk: well, it pushes the current contents of the git repository to the remote repository.
[19:32] <kandarpk> Ok.
[19:32] <lfaraone> kandarpk: keep in mind that pushing to the remote repository is different than "uploading to Debian", the package is not published in Debian when you push up.
[19:32] <kandarpk> yes
[19:32] <bernie> lfaraone: reading
[19:32] <kandarpk> lfaraone: Maintainers need to approve those, right ?
[19:33] <lfaraone> kandarpk: yes.
[19:33] <bernie> lfaraone: does debian use dput to upload packages? just like ubuntu? I would expect so.
[19:33] <lfaraone> bernie: Ubuntu used dput to upload packages, just like Debian :)
[19:34] <lfaraone> bernie: git usage isn't mandatory, some use bzr, some use hg, etc, and some use nothing at all.
[19:34] <kandarpk> lfaraone: what does uploading to Debian mean ?
[19:34] <kandarpk> making packages available for download ?
[19:34] <lfaraone> kandarpk: it's like when you upload to the PPA. It means that all people who track debian's testing or unstable package repositories will automatically get the update.
[19:34] <kandarpk> hmmm
[19:34] <bernie> lfaraone: reading your post about sugar in ubuntu, I believe the formador was hitting the gdm bug
[19:35] <bernie> lfaraone: do you know why these packages aren't flowing to the ubuntu updates repo?
[19:35] <lfaraone> bernie: yes, GDM segfaults because the sugar desktop file does not contain a comment. THis bug has been fixed in an update of GDM.
[19:36] <lfaraone> bernie: because nobody has decided to do so. We'd have to follow the SRU proceedure, and document all the changes. I am happy to do that if somebody wants me to, the changes meet the SRU criteria, and the release managers are not going to reject it.
[19:36] <bernie> lfaraone: well, I lack context but I guess dfarning or manusheel would know
[19:37] <bernie> lfaraone: also, if someone fixes Sugar bugs in the debian packages, how would these changes flow into the ubuntu updates repositories as well?
[19:37] <bernie> lfaraone: I suppose upstreaming a bug fix is a common thing to do...
[19:37] <lfaraone> bernie: somebody would have to request an SRU, and perform the fix.
[19:38] <lfaraone> bernie: they will automatically be synced over in the next release of Ubuntu.
[19:38] <lfaraone> bernie: but previous release require a SRU.
[19:39] <lfaraone> bernie: we probably should request https://wiki.ubuntu.com/StableReleaseUpdates/MicroReleaseExceptions
[19:39] <lfaraone> dfarning: would persuing this be useful?
[19:42] <bernie> lfaraone: is anyone doing bugfixing in debian now?
[19:43] <lfaraone> bernie: jonas is. I am.
[19:44] <lfaraone> bernie: when I become a DD, it will be much easier for me to fix bugs in Debian. As is, I need Jonas to first upload a package along with the authorization for me to make future changes to it before I can upload.
[19:44] <bernie> lfaraone: (otherwise, this would just introduce one additional level of indirection to the USR workflow without anything in return)
[19:44] <bernie> lfaraone: ah, I typed too fast. well, that's good to know
[19:45] <lfaraone> bernie: what we get in return is Jonas' continued packaging of new upstream releases of Sugar, which is an enormous amount of work we are very thankful for.
[19:45] <bernie> lfaraone: another question (sorry to bother you so much). are any of the changes in the Ubuntu PPA mergeable in debian? would changes that are neutral for debian but useful for ubuntu be welcomed?
[19:45]  * lfaraone will be back later.
[19:46] <lfaraone> bernie: yes. I have done so in the past, and there is no ill feeling towards doing such things as far as I can tell.
[19:51] <bernie> lfaraone: when you come back: do you know who would normally react to an SRU for the sugar packages? Are they responsive and reasonable people?
[20:06] <lfaraone> bernie: an sru would be filed by any MOTU, like me. any member of ubuntu-sru would process it like any other SRU request. If we wanted to make iur and their livez easier, we can get a continuous exception from the sru critwria if we establish that SL makes predictable, stable point releases.
[20:07] <lfaraone> im in the car , pardon my crappy iPhone ssh-fu.
[20:08] <lfaraone> ubuntu-sru are reasonable, if not overworked, people.
[20:09] <lfaraone> bernie: ^^. i'll be in and out, hilite as usual to grab my attn
[20:17] <dfarning> kandarpk, are you still around
[20:17] <kandarpk> dfarning, yes
[20:17] <dfarning> did lfaraone and bernie answer many of your questions?
[20:18] <kandarpk> dfarning: yes
[20:18] <dfarning> kandarpk, you were asking about the strictness of debian vs ubuntu.
[20:18] <dfarning> they are both equally strict at this point.
[20:19] <kandarpk> dfarning: no necessarily strictness
[20:19] <dfarning> the piece that I was forgeting to mention was that so far, we have been working in the sugar team ppa which is a like a personal sand box.
[20:20] <kandarpk> dfarning: true
[20:21] <dfarning> kandarpk, the activity packages ( and many of our bug fixes) would not have met the coding standard for debian or Ubuntu universe.
[20:21] <kandarpk> dfarning: what do they look at before rejecting a package
[20:22] <dfarning> I just set the ppa up as some where we could test and learn to package with running into coding standards.
[20:23] <dfarning> kandarpk, the most important thing is that someone else can easily review ones work.
[20:24] <dfarning> for example we make all of our changes inside the debian/ directory
[20:24] <dfarning> everything outside of the debain/ directory is often refered to as a pristean tarball.
[20:25] <dfarning> so if you what to review what a packager did to a package you can just look in the debian/patches and see.
[20:26] <kandarpk> hmmm.
[20:27] <dfarning> If you look in my sugar-activities package.  I included about 20 separate tar balls.
[20:27] <dfarning> that would be hard to defend.
[20:27] <dfarning> kandarpk, each of the tar-ball (activities) should be individually packaged.
[20:28] <dfarning> kandarpk, does that make sense?
[20:28] <kandarpk> doesn't having more tar balls lead to increased modularity
[20:28] <kandarpk> why should it be hard to defend ?
[20:29] <dfarning> each package should be a single (modular)  chunk of code
[20:30] <kandarpk> Ok.
[20:30] <dfarning> generally each tar ball is a very disinct upstream piece of software.
[20:30] <kandarpk> you mean it gets harder to manage these many sub-modules, right ?
[20:31] <dfarning> in our case the line gets kind of blurry becase al of the activites tarball come from the same upstream source.
[20:32] <dfarning> Yes, there is a philosophy in unix and linux the 'thing should do one thing, but do it well'
[20:33] <dfarning> hence one tarball per-package.
[20:33] <neeraj> dfarning, hi, thats y you were telling us that day to break sugar-artwork..
[20:33] <neeraj> ?
[20:34] <dfarning> manusheel, are you still awake.
[20:34] <kandarpk> dfarning: Ok.
[20:34] <dfarning> neeraj, I don't remember the exact context but probobly.
[20:35] <kandarpk> dfarning: It was about coding practices
[20:35] <kandarpk> does the source code too play iys part ?
[20:35] <kandarpk> *its
[20:36] <dfarning> kandarpk, I currently have three project going and get confused:)
[20:36] <dfarning> kandarpk, Can you explain what you are asking further?
[20:36] <kandarpk> the efficiency of the code,
[20:37] <kandarpk> I mean there must be some codes which are more likely to break
[20:37] <dfarning> kandarpk, I still don't understand.
[20:37] <neeraj> dfarning: my bad.. it was sugar-activity package whose size is quite large as comparison to other one..
[20:38] <dfarning> neeraj, exactly sugar activites is about 20 individual packages which need to be properly split into individaul packages before either debian or ubuntu would accept them.
[20:39] <kandarpk> dfarning: what causes packages to break/not function properly ?
[20:39] <neeraj> kandarpk sir are you asking that we *might* need to make some changes in source code to make it compatible/pass thorough debian?
[20:39] <kandarpk> neeraj: reght
[20:39] <neeraj> if that is the question then I don't think we do any such thing..
[20:39] <kandarpk> *right
[20:40] <kandarpk> Ok.
[20:40] <neeraj> dfarning, please tell us whether we do such things or not?
[20:42] <dfarning> As we have seen sometime there have been bugs in ubuntu but not in debain and vice-versa.
[20:43] <dfarning> the biggest problem as bernie refered to earlyer is when ubuntu and debian diverge and our packages need to interact with those divergent pieces.
[20:44] <dfarning> kandarpk, were you working on the data-time bug?
[20:44] <kandarpk> dfarning: that was automatically fixed on update
[20:45] <dfarning> where the corner was not being drawn correctly in ubuntu?
[20:45] <kandarpk> dfarning: I didn't know about this issue
[20:46] <neeraj> i guess dipankar was working on that..
[20:46] <dfarning> The (older) version of gtk ubuntu was using had a bug that they needed to fix before the code would work right.
[20:46] <neeraj> he had zeroed down some *_draw.c file(not sure )
[20:47] <neeraj> Ok..
[20:47] <dfarning> yet the code worked fine on the newer version of gtk that debian unstable used.
[20:47] <dfarning> those are the sorts of problems we are likely to encounter.
[20:47] <kandarpk> dfarning: diverging means use of different libraries by the two OS's ?
[20:48] <dfarning> kandarpk, yes.
[20:48] <manusheel> dfarning: Yes, I am awake.
[20:49] <kandarpk> dfarning: so, as the two diverge it is like making packages for two different OS's ?
[20:50] <dfarning> ideally, in our work flow, we will fix bugs we find with local tempoary patches in either debian or ubuntu....  while at the same time we will be working with upstream to fix the problem there so the the fix will trickle back down so our tempory patch is no longer needed.
[20:50] <neeraj> in such case can we use debian as an upstream for all packages?If yes then how?
[20:51] <dfarning> kandarpk, exactly the farther they diverge the more work it is for us.
[20:51] <kandarpk> hmmm
[20:51] <dfarning> kandarpk, but there tends to ebbs and flows in divergances as things branch and merge.
[20:52] <dfarning> neeraj, ideally debian will be upstream for us.
[20:52] <kandarpk> dfarning: and your point about activity package
[20:53] <kandarpk> it had 20 tarballs, so it should be packaged as 20 ( or somewhat like that )
[20:53] <kandarpk> packages,
[20:53] <dfarning> kandarpk, exactly
[20:53] <kandarpk> Ok.
[20:54] <dfarning> neeraj, I hope Ubuntu and debian can work very closesly around sugar so we don't have significtant upstream downstream issues.
[20:55] <kandarpk> dfarning: initially, the only difference b/w Ubuntu and Debian was of the packages they came with
[20:55] <kandarpk> and not of standard libraries ?
[20:57] <dfarning> kandarpk, Yes, initally they were nearly identical.  If i recall corectly the only differences were cosmetic and a few setup wizards to make it easier to install and confgure ubuntu.
[20:57] <kandarpk> Ok.
[20:57] <dfarning> kandarpk, but alot has happened over the last 12 or so releases.
[20:59] <kandarpk> dfarning: yes. Ubuntu has been changing radically
[20:59] <kandarpk> over time
[20:59] <dfarning> for some reason there is often alot of bickering between debain and ubuntu.... it is just like two brothers fighting.
[20:59] <dfarning> kandarpk, and debain has changed just as much
[21:00] <kandarpk> dfarning, Ok !
[21:01] <dfarning> all distros try to cherry pick the best ideas from each other. so some time ubuntu leads, some times, debian leads and some times redhat/fedora lead:)
[21:01] <dfarning> it is the cherry picking the causes the continual evolutionaly progress.
[21:02] <dfarning> At the end of the day the disputes usually boil down to different people cheering for their home team:)
[21:03] <dfarning> kandarpk, does that help?
[21:04] <kandarpk> dfarning: alot
[21:04] <kandarpk> dfarning: thanks for such an elaborate explaination.
[21:04] <kandarpk> that was quite exhaustive
[21:05] <dfarning> kandarpk, it kind of needs to be exhaustive because of the history and interactions between debain we sometimes do silly things for silly reasons.
[21:06] <dfarning> so as develops sometime we close our eyes and not worry about the silly bits or what cause us to do something that way.
[21:06] <kandarpk> hmmm
[21:08] <kandarpk> dfarning: all the modifications made in Debian repositories are passed over to Ubuntu in the next release ?
[21:08] <dfarning> opps the above section should have been 'between debian and ubutu.' neither is to blame
[21:10] <dfarning> kandarpk, yes, (for the most part) please see https://wiki.ubuntu.com/MaverickReleaseSchedule
[21:10] <dfarning> debian tends to have a release cycle of about 2 years while ubuntu has a release cycle of exactly 6 months
[21:12] <dfarning> so for the first couple of week of a new ubuntu release.  most of the developer work is spent syncing the future ubuntu release with debain unstable.
[21:13] <dfarning> that period goes until june 24,
[21:13] <dfarning> then as it gets closer to the ubuntu release effort shift towards polish and bug fixing.
[21:13] <kandarpk> right
[21:14] <kandarpk> we have our primary target as Debian for this reason ?
[21:16] <dfarning> kandarpk, exactly
[21:17] <kandarpk> dfarning: once we start using git, what would be our next step ?
[21:18] <dfarning> work from ubuntu does not automaticly 'flow'into debian.  Thus if we use ubuntu as our primary target and are not very careful we can introduce divergence.
[21:19] <dfarning> ok for this we have to step back a bit.
[21:20] <kandarpk> dfarning: got your point about divergence
[21:20] <dfarning> if we package a ubuntu package it can live in our ppa for as long as we like.... but if we want it to get our packages promoted into universe someone has to review them.
[21:21] <dfarning> and all of the sugar package are pretty complicated. so it is likely the our packages will sit in the reveu queue for a long time.  Most of the ubuntu developers are volunteers.
[21:23] <dfarning> but, if we get our packages accepted into debian and reviewed by a 'debian developer' it is very easy to get them synced from debian to ubuntu.
[21:24] <dfarning> and since we have Jonas who is willing and able to review our package in debian it is more expedient to go that route.
[21:24] <kandarpk> dfarning: does promotion into ubuntu universe too requires review ?
[21:26] <dfarning> yes, exactly, that is what takes so long:( the packages can sit in the queue to be reviewed until some one (usually a MOTU) reviews them.
[21:26] <kandarpk> Ok.
[21:28] <kandarpk> that means we need to push our packages using git to be reviewed by some DD to get listed in universe and then be synced into Ubuntu ?
[21:28] <dfarning> so, we are very luck to have lfaraone.  he is a 'debian maintainer', 'ubutu MOTU', and will so be a 'debian developer' so I hired he for the summer to mentor us and review packages as needed.
[21:29] <dfarning> exactly, and jonas is a DD.
[21:29] <kandarpk> great.
[21:30] <dfarning> over the next couple of months I hope that one of more seeta developer become 'debian maintainers' for the sugar packages and ubuntu developers with authority to commit sugar packages on ubuntu.
[21:32] <kandarpk> how do they select people to be debian maintainers ?
[21:34] <dfarning> kandarpk, basically one needs to become expert in a subset of packages. and the they are approved.  Luke would know more of the details.
[21:35] <kandarpk> hmmm
[21:36] <kandarpk> dfarning: that's a long way to go.
[21:37] <kandarpk> whats our primary task for now ?
[21:41] <dfarning> kandarpk, the most important thing is for everyone to become proficient at packing with git.
[21:42] <kandarpk> Ok.
[21:43] <kandarpk> I think if we put some effort, it can be done in a day or two
[21:44] <dfarning> kandarpk, now that everyone has learned basic package skills working in our ppa 'sandbox' we are ready to climb the next rung of packaging via git and requesting a review via jonas of luke.  with in a month or so of hard work some of us (not me) will then be able to commit directly.
[21:45] <kandarpk> dfarning: "not me" ?
[21:47] <dfarning> kandarpk, I have three project going USR, Sugar.88 on Fedora 11, and developing javascript based content.  hence the need for expert developers like you and the rest of the seeta team:)
[21:47] <kandarpk> hmmm
[21:48] <dfarning> bernie, is leading the Sugar.88 on Fedora 11 project, and Manu and I are trying to get the content piece going.
[21:49] <dfarning> so it is interesting:)
[21:50] <dfarning> kandarpk, do you feel more comfortable about how all the pieces fit together?
[21:50] <kandarpk> definitely
[21:50] <kandarpk> dfarning: great help from your side
[21:50] <dfarning> great.
[21:52] <dfarning> kandarpk, keep asking questions as you saw with bernie everyone is willing to stop and teach..... because that is how we learned.
[21:52] <manusheel> dfarning: Yes, very helpful indeed.
[21:53] <dfarning> I have to run and make dinner.  thanks for everything.
[21:54] <kandarpk> dfarning: bye
[22:39] <bernie> dfarning: you're a good mentor
[22:40] <dfarning> bernie, thanks
[22:42] <dfarning> ubuntu/debian relationships can be tricky.  We are the first project to keep our ubuntu package source as branches in Aloith:)
[22:44] <bernie> dfarning: tch is also learning his way through Sugar Labs' bureaucracy
[22:45] <bernie> dfarning: he's been pissed off several times by negative comments, but now he understands to respond constructively, without attacking the reviewer
[22:45] <dfarning> bernie, yes, I have been watching that.
[22:47] <dfarning> bernie, tch is an a particularity good position.  spending time winning an argument does not help him:(  Getting patches does:)  I expect that we cause and net shift in our upstrea,
[23:14] <bernie> dfarning: it is already happening
[23:14] <bernie> dfarning: sugar development seems to have resumed thanks to our push
[23:16] <dfarning> bernie, +1,  I am going though USR looking for bugs for tomorrows bug squashing session... i think I might cry:(