[02:21] btw, i wont be on lter [04:17] dfarning: hi [13:12] dfarning: good morning. [13:12] kandarpk, good morning .... we lost you for a while:) [13:12] dfarning: fitness problem :( [13:14] dfarning: the latest tarball of s-speak-a isn't available at http://download.sugarlabs.org/sources/honey/Speak/ [13:14] any other place I can look for it ? [13:14] kandarpk, I recommend that each documenter set up there own local jhbuild and builddoc set up. [13:15] dfarning: Ok, that is something which can yield long term benefits [13:15] kandarpk, did you look on aslo for the development information? it might have a another page listed there. [13:15] dfarning: the model is very similar to git.debian.org [13:15] 'morning [13:16] bernie: Hi :) [13:16] good morning [13:16] kandarpk, it take me about and hour to prepare a patch:( But working with the developers here it takes them a couple seconds:) [13:17] dfarning: Great. [13:19] kandarpk, yes the usage of git with the sugar packages is same as with git.debian.org.... That is why we spend all that time learning git. [13:19] bernie, are you around? [13:20] dfarning: I meant the model we are thinking of, for documentation [13:21] kandarpk, yes when creating the documentation you will be working with the core sure packages and adding docstrings to those packages. [13:26] dfarning: yep [13:51] dfarning: do you have anything for me to work on ? [14:00] kandarpk, sorry my computer stopped beeping at me:( [14:00] hello dfarning , a very good morning [14:00] dipankar, good morning. [14:00] dfarning: np. [14:02] dfarning: just that I am logged into windows now. [14:02] kandarpk, I guess a high priority would be setting up your local document build system. And figure out how to modify the docstrings and create patches [14:02] dipankar, good morning [14:02] dfarning, how are you? [14:03] at Paraguay? [14:03] dfarning: when we push changes, only patches are submitted to git, isn't it ? [14:04] kandarpk, hi [14:04] kandarpk, I don't think so [14:04] dipankar: Ok. [14:04] kandarpk, I think we need to submit the whole modified code too. [14:05] dipankar: by the way, how are you doing now ? [14:06] kandarpk, yes. technically a commit is just a diff(patch) [14:06] kandarpk, I am doing good. ^^ Here's your answer [14:06] I was wrong I guess :) [14:06] dfarning, but we upload all the files. oops, thats the mistake. [14:07] http://www.newartisans.com/2008/04/git-from-the-bottom-up.html [14:07] dipankar, things are going well. [14:07] ohk, when we use git push, it sends only the changes up to the repo. That must be it then. [14:07] dipankar, kandarpk: there's a pdf linked in that page. it's a very good reading about how git "thinks" [14:08] bernie: thanks [14:10] dfarning: users can clone the source, make changes and push it back. [14:10] bernie, thanks a lot. [14:10] kandarpk, Yes. [14:10] dipankar: yup. push sends one or more commits (~= patches) to a remote repo... no merges are possible on the remote side, so the stuff you send needs to be based directly on top of what the remote end has. [14:11] dfarning: so they only require to know how to test patches they apply [14:11] bernie, Thanks again. How are you doing btw? How's your project coming along? [14:12] kandarpk, who are the they in you above sentance. [14:13] dfarning: the users who want to add patches [14:13] dipankar: dfarning arrived here a few days ago... the uruguayans arrived a little earlier... we're all working together on completing Dextrose (which is the new name of F11-0.88) [14:14] dipankar: well, admittedly, we've been a little unfocused lately... meeting new people, show them the deployment, discuss strategy... we need to do more technical work next week [14:15] bernie, I didn't know you and dfarning are meeting :). I am sure you guys can come over these problems easily [14:16] kandarpk, yes the developer needs to know how to clone a repo(code), modify the code, test the modification, and either push or make a patch. [14:18] dipankar: we're nearing release, so only the hard problems are left :-) [14:19] bernie, Best of luck with that :) [14:26] kandarpk, I need your help with debian/rules file [14:26] kandarpk, my log activity didn't work on 0.84 as you tested. [14:26] yeah [14:27] so now I need to change the debian/rules accordingly. [http://lists.alioth.debian.org/pipermail/debian-olpc-devel/2010-July/002600.html] [14:27] but it seems the rules file was very different. [14:28] just a sec I will link the debian/rules [14:31] kandarpk, here it is: http://git.debian.org/?p=collab-maint/sugar-log-activity.git;a=blob;f=debian/rules;h=ea5e9a66489c8ad5064a7de8fd6fb4fc1dcb6535;hb=HEAD [14:31] kandarpk, except for line no. 5 there is nowhere else mentioned about DEB_SUGAR_BRANCHES in the file, pretty weird [14:31] Oh [14:32] could just help with that? [14:32] I am stuck with this portion only [14:34] dipankar: trying to see if I can help [14:34] kandarpk, thanks very much [14:42] dipankar: I think after restoring debian/rules to the original state, you need to rename the activity by adding -0.86 [14:42] in rules and control.in [14:43] dipankar: that is sugar-log-activity-0.86 [14:43] ohk [14:44] kandarpk, but here it mentions something else I guess (I may be wrong though :)) [14:44] If it works in 0.86 and above, for example, restore your debian/rules [14:44] file to it's original state. Change "DEB_PYTHON_SUGAR_PACKAGES" adding [14:44] "-0.86" to the end. Change the binary package name in debian/control.in [14:44] adding "-0.86" to the end. [14:45] Luke has mentioned to add -0.86 in the DEB_PYTHON_SUGAR_PACKAGES [14:45] DEB_PYTHON_SUGAR_PACKAGES = sugar-log-activity [14:45] oops [14:45] ohk :) got it [14:45] your rules file says ^^ [14:45] *thats sickness has raptured my attention. [14:46] sorry to trouble you kandarpk :( [14:46] dipankar: I often do that when asking for help. [14:46] dipankar: np [14:48] * dipankar is taking a 5 min break. [14:48] * dipankar is away: Be Right Back [14:53] kandarpk: Hi Kandarp. [14:53] Around? [14:53] manusheel sir: hello sir [14:55] kandarpk: How is Speak activity packaging coming along? [14:56] manusheel: sir, its latest tarball isn't available. [14:57] latest version is 16 and tarball available has version 11 if I remember correctly [14:57] kandarpk: Ok, that is unfortunate. [14:58] manusheel: sir, I was waiting for Luke to know if we can generate the orig tarball ourself [14:58] manusheel: but I don't think thats possible [14:58] kandarpk: Let us send an update on not been able to package Speak activity, and possible suggestions on how we could expedite the generation of tarball. [14:58] else he would have told before [14:59] If not by ourself, then by the Sugar activity team. [14:59] ok. [15:02] * dipankar is back (gone 00:13:47) [15:03] Kandarpk: Did you look at the 4 steps suggested by David? [15:04] manusheel sir: yes sir, we discussed about it a little while ago [15:05] kandarpk: Can you try to do that process today? [15:06] manusheel sir: I am not very clear with the scheme. [15:06] kandarpk: What all things are we not clear with? [15:06] Let us ask David about it. [15:06] manusheel sir: can't we just clone jhbuild, make changes and push it back [15:08] dfarning: Around ? [15:08] kandarpk: I am not sure we can do it like this. [15:09] manusheel sir: are some permissions required to push into git repo of jhbuild ? [15:10] kandarpk: I suspect. [15:11] manusheel sir: Ok, then we'll need to follow the steps suggested by David. [15:12] kandarpk: Yes, we should follow the steps. [15:12] Let us try and see if there are any errors. [15:12] Or issues. [15:14] manusheel sir: I don't have much knowledge of git-format-patch. Rest all is quite clear. [15:15] kandarpk: I would like you to search and read on it. Let me send you the resource. [15:18] * lfaraone waves. [15:20] kandarpk: well, you can do the same thing with speak as we did with other activities for which there are no upstream tarballs, use git snapshots. [15:20] hey lfaraone good morning. [15:20] and ask alsroot nicely to make tarballs. :) [15:21] lfaraone: better late than never. [15:22] lfaraone, I finally am through with the checking of log-activity. Its working for 0.86 and 0.88 only. here is the repo : http://git.debian.org/?p=collab-maint/sugar-log-activity.git;a=summary [15:22] dipankar: cool. [15:23] * alsroot can release Speak tarball, if there is such need [15:23] alsroot: Please do so. [15:23] alsroot: Please release [15:23] Thank you. [15:26] alsroot: that'd be awesome, thanks. [15:27] lfaraone, I would like you to review that : sugar-log-activity. [15:27] dipankar: okay, mark it in the wikipage. [15:28] lfaraone, I think I have rectified all things. [15:28] ohk [15:28] done, http://download.sugarlabs.org/sources/honey/Speak/Speak-16.tar.bz2, it has bundled "toolkit" dependency but better to leave it as-is, toolkit was rewritten in vala and will be accessible later as polyol [15:29] kandarpk: Please have a look at http://www.kernel.org/pub/software/scm/git/docs/git-format-patch.html [15:29] alsroot: thatks a lot. [15:29] *thanks [15:30] np [15:36] kandarpk: Did you get a chance to look at git-format-patch? [15:37] The link that I had send. [15:37] manusheel sir: sir, will do that in 5 min. [15:40] kandarpk: Sure. [15:44] lfaraone, done [15:45] lfaraone, I have changed the wiki page: http://wiki.debian.org/Sugar/tasks [15:45] *modfied [15:49] manusheel: sir, I think I understand what needs to be done [15:50] regarding git format-patch [15:51] *switching to USR. [15:57] dipankar: in log, you found that it didn't work in 0.84, right? [15:57] dipankar: if so, please talk to neeraj about renaming the actiivty to sugar-log-activity-0.86. [15:57] lfaraone, actually I sent it over to Kandarp for checking. [15:58] and he reported me that it is not working on 0.84 [15:58] lfaraone, did I do something wrong? [15:58] I am not getting 'where' to rename activity? [15:58] dipankar: well, if an activity does not work in 0.84, the activity shoul d have -0.86 appended. If it didn't work in 0.86 or 0.84, it should have -0.88 appended to the name. for ex, see what neeraj did with write. [15:59] ohk. [16:01] * lfaraone is looking at etoys. [16:10] lfaraone: git init isn't creating branch master [16:11] kandarpk: what do you mean? "git init" will initialize a git repository with a master branch. [16:11] lfaraone: git branch gives nothing [16:11] lfaraone, I think I have done that step in renaming the package in debian/rules [16:11] lfaraone: some package missing ? [16:12] kandarpk: because you're on master by default. [16:12] lfaraone: how do I list the branches ? [16:13] lfaraone, here is the debian/rules. line no. 6 http://git.debian.org/?p=collab-maint/sugar-log-activity.git;a=blob;f=debian/rules;h=5151f39fdd41e9f31d82a1c4dff01ec79a90bb21;hb=89893508c370cf5c27a6eb69fecf4bd72920f3e3 [16:14] bernie: around ? [16:14] kandarpk: "git branch", but you won't see anything until you commit. [16:14] lfaraone: Ok. [16:23] kandarpk: yup? [16:23] kandarpk: what lfaraone said [16:23] kandarpk: "git branch -a" to see also the remote ones [16:23] bernie: thanks [16:24] kandarpk: normally, when you clone a remote repo called foo, you get many branches foo/bar, foo/baz... [16:24] kandarpk: that's the meaning of "origin/master". it's the "master" branch on the "origin" remote repo. [16:27] bernie: can we straightaway push changes to git repo of sugar-jhbuild ? [16:28] kandarpk: you first need to replace the remote url in .git/config [16:28] kandarpk: there's a command-line tool "git remote" to edit remotes, but I find it easier to directly edit the config file. [16:29] kandarpk: the remote url you'll find is of rhe git://... kind [16:29] kandarpk: that's the read-only git protocol which can't be used for pushing [16:29] kandarpk: if you go to the repository page on git.sugarlabs.org, you'll find what the correct push url is supposed to be [16:30] kandarpk: you'll see the push url *only* if you are logged in and you happen to be one of the authorized committers. [16:30] kandarpk: what repo is this? [16:30] bernie: Ok. [16:30] bernie: git://git.sugarlabs.org/sugar-jhbuild/mainline.git [16:30] bernie: it contains the source codes for various sugar modules [16:31] and more [16:31] kandarpk: then you could ask silbe for access... but normally one does not directly commit to a repository until he/she becomes a maintainer of the project. [16:31] kandarpk: that's a different mindset from cvs and svn... [16:31] bernie: I only wanted to confirm this [16:31] bernie: that ordinary users don't push their changes [16:32] bernie: thanks [16:35] kandarpk: yep [16:35] kandarpk: there's a BUT though... [16:39] bernie: ? [16:39] * dipankar says good bye to all. [16:49] kandarpk: Ok, great. [16:49] What needs to be done? [16:49] Can you share your approach? [16:52] manusheel sir: Sir, creating patches is quite easy [16:52] manusheel sir; 1) clone the repo [16:53] kandarpk: Ok. Did you read about git-format-patch, the link that I had send? [16:53] 2) copy master into some new branch [16:53] kandarpk: back [16:53] 3) make changes in new branch [16:53] 4) git format-patch master [16:54] kandarpk: Neat. [16:54] manusheel sir: it will create a patch of changes b/t master and new [16:54] manusheel sir: http://paste.ubuntu.com/468492/ [16:55] changes in activityservice.py and activity.py [16:55] Yes, I saw that. [16:56] kandarpk: that's a good workflow [16:56] kandarpk: So, have we submitted the patch? [16:56] bernie: thanks [16:57] kandarpk: you can also _not_ do a separate branch until you find out you need it [16:57] manusheel sir: No. [16:57] bernie: we need other users to submit changes [16:57] kandarpk: you get a local branch for free when you clone. "origin/master" gets implicitly branched into the local "master" branch [16:58] bernie: Ok. [16:58] kandarpk: btw, I mentioned an alternative to push earlier... [16:58] kandarpk: here it is: [16:58] kandarpk: gitorious lets *any* logged in user create clones of any repository [16:59] bernie: yes. [16:59] kandarpk: once you've created your own clone (jhbuild-kandarkpk, jhbuild-seeta or jhbuild-ubuntu), you can push to it [16:59] kandarpk: then, you can request merge with mainline (or any other clone) [16:59] kandarpk: that's an interesting workflow promoted by these new distributed forges such as gitorious and github. [17:00] kandarpk: I like more the workflow of sending patches with git-format-patch + git-send-email and get them reviewed on the mailing lists. [17:00] kandarpk: because all these inter-branch merges are likely to go unnoticed by the majority of the community [17:01] kandarpk: and one important factor in open source development is to make everyone aware of what the others are doing so they can learn from each other. [17:01] bernie: Nice. [17:01] kandarpk: for the same reason, reviews done in bug trackers are anti-community. [17:01] bernie: Ok. Great. [17:02] * bernie gets into the shower [17:02] :-) [17:02] bernie: thanks a lot. [17:02] bernie: still there ? [17:05] manusheel sir: the patch doesn't have anything substantial in it. [17:08] kandarpk: Ok. [17:08] kanarpk: Can you document this workflow with a sample example? [17:09] kandarpk: It is important for us to recognize that this workflow works. [17:11] manusheel sir: sure. [17:11] kandarpk: Great. Please share the expected output at every step. [17:12] kandarpk: Again, prepare the steps, expected output and commands from a perspective of encouraging patch submissions from the community. [17:13] Try to make it for a user with a basic familiarity with Linux. This strategy will be helpful for us towards getting maximum patch submissions committed. [17:15] manusheel sir: Ok. [17:16] kandarpk: Thanks Kandarp. [17:16] manusheel sir: I'll need to include steps for working on sphinx as well ? [17:17] that will be required for testing changes locally [17:17] kandarpk: Yes. Please do. You can take references from the document prepared earlier. [17:17] That would expedite process. [17:17] manusheel sir: One problem. [17:17] kandarpk: Ok, please go ahead. [17:18] manusheel sir: how will users be able to get the rst files prepared by us. [17:18] for testing locally they'll need that to create something similar to nline version [17:19] *online [17:20] but there is an online resource available to see how the docstring will look in HTML format after conversion. [17:20] manusheel sir: ^^ [17:20] kandarpk: Ok, interesting question. [17:21] manusheel sir: http://www.tele3.cz/jbar/rest/rest.html [17:22] manusheel sir: I think the user here would only like to see if his docstring will be correctly converted into HTML or not [17:22] creating a complete website look-alike wont be required [17:22] kandarpk: I don't think users will be able to get the rst files prepared by us. Yes, they'll need to use this tool to preview the correct conversion of their docstrings to HTML. [17:23] kandarpk: Yes, that is more than good for them and us. [17:23] They should be able to see their docstrings getting converted to HTML. [17:26] kandarpk: I think it should be easy for us to see their submissions, and revisions, wherever possible. [17:32] manusheel sir: Ok. [17:32] manusheel sir: so, we should isolate the user from the process of using sphinx ? === neeraj is now known as neeraj_afk [17:53] lfaraone: for how long will you be around ? [17:53] kandarpk: about 5 minutes. [17:54] lfaraone: Oh, not much time left then. [17:54] kandarpk: what can I do for you? [17:55] lfaraone: Jonas wanted one of us to add README.source to sugar wiki [17:55] lfaraone: so I thought if I could send it to you first for review [17:55] kandarpk: right, I can do that, since it'll be hard to track down the right version. [17:55] kandarpk: sure. I'll reply to you by monday on tha. [17:57] lfaraone: do I need to reply on the mailing list ? [17:57] kandarpk: well, you might as well post it to the wiki, we can always edit it later. [17:58] lfaraone: Ok. [17:59] *will be back in 2min. [18:02] lfaraone: sugar-emulator is crashing on my system as well [18:02] on installing python-gi [18:02] kandarpk: cool, thanks. [18:04] lfaraone: I am using README.source from core sugar [18:04] kandarpk: sure. [18:10] kandarpk: by the way, when you finish working on a bug and uplaod a new patch (like in https://bugs.edge.launchpad.net/ubuntu/+source/sugar-connect-activity/+bug/609249) you should set it back to "confirmed" or "new" from "incomplete" and unassign yourself. [18:11] lfaraone: Ok. [18:14] lfaraone: reset-upstream-to-v87 [18:14] what should I replace v87 with something general ? [18:16] kandarpk: not sure what you mean, but I have to go. [18:16] it's j ust a branch name, so it doesn't matter what you call it. [18:17] lfaraone: I mean, writing something like v.VERSION_NUMBER [18:17] sure. [18:18] lfaraone: thanks. [18:19] kandarpk: Yes, the user might not be interested in Sphinx like us. [18:19] manusheel sir: Ok. [18:19] kandarpk: However, we should keep in mind that we should have an option for that too. [18:19] Never know, who want to indulge in that :-) [18:20] manusheel sir: for that all our files should be accessible online. [18:27] kandarpk; Ok, then let us take that part later. [18:27] First, let us concentrate on the workflow, where they are not online. [18:27] However, please add a small section on this part too. [18:28] manusheel sir: Fine. [18:31] kandarpk: Great. === neeraj_afk is now known as neeraj [18:32] So, let us try to generate a sample implementation using the 4 steps. [18:32] Hope it works out well. [18:54] kandarpk: Please take screenshots, and copy paste outputs wherever possible. [18:54] We'll put it at the wiki. [18:54] manusheel sir: of the HTML pages ? [19:00] kandarpk: Output per step. [19:00] manusheel sir: Ok. [19:01] That is we write a command at the terminal, and we get this "...." as the output. [19:01] kandarpk: Great work. [19:01] manusheel sir: as suggested by Jonas, I've added README.sources at : [19:01] http://wiki.debian.org/Sugar/README [19:02] kandarpk: Neat. [19:03] Yes, can you send an update e-mail to the team about this. [19:03] That should be very helpful. [19:03] We should also reply on Debian olpc mailing list about it. [19:06] kandarp, this is very neat. [19:07] Do send an e-mail on olpc debian mailing list. [19:08] kandarpk, neeraj: Did you get a chance to review jigsawpuzzle? [19:09] manusheel sir: no sir. Shall I review it now ? [19:10] *seeing it now [19:11] kandarpk: after, we complete the sample implementation. [19:11] neeraj: you can review it now. [19:11] manusheel sir, doing that [19:15] manusheel sir, acco to lfaraone review on tasks http://lists.alioth.debian.org/pipermail/debian-olpc-devel/2010-July/002682.html, Ankur has made the changes.. [19:16] manusheel sir, As far as I can see he has made changes correctly. There might be just one redundant line in the last of debian/copyright, o/w all changes seems to be fine.. [19:17] neeraj: ok, great. [19:18] kandarpk sir, please confirm > acco to luke review on tasks http://lists.alioth.debian.org/pipermail/debian-olpc-devel/2010-July/002682.html, Ankur has made the changes.. [19:19] neeraj: just 5 min. [19:20] kandarpk: np.. take ur time [19:39] manusheel sir: git clone git://git.sugarlabs.org/sugar-jhbuild/mainline.git [19:39] isn't getting the source [19:40] kandarpk: Interesting. [19:40] What is the error? [19:41] manusheel sir: some folders are missing [19:42] manusheel sir: till now, I was working on the sugar-jhbuild folder I had used to install sugar [19:43] kandarpk: Ok. [19:43] Can you copy paste the logs? [19:43] kandarpk: We should touch base with alsroot and bernie about it. [19:44] manusheel sir: [19:44] ls [19:44] config debian sbuildbot scrils [19:44] config debian sbuildbot scripts sjhbuild sugar-jhbuild sugar.jhbuildrc [19:44] pts sjhbuild sugar-jhbuild sugar.jhbuildrc [19:44] * alsroot is not using jhbuild, so can't help with it [19:44] wait [19:44] -------------------------------- [19:45] ls [19:45] config, debian, sbuildbot, scripts, sjhbuild, sugar-jhbuild, sugar.jhbuildrc. [19:45] this is what I got on cloning [19:47] manusheel sir: contents of jhbuild-folder I already had [19:47] ls [19:47] config, debian, install, jhbuild, sbuildbot, scripts, sjhbuild, source, sugar-jhbuild, sugar.jhbuildrc. [19:48] kandarpk: Ok. [19:49] alsroot: Thank you for letting us know. [19:49] Appreciate it. [19:50] kandarpk: I see. [19:50] Lot of folders are missing. [19:50] * alsroot uses gentoo instead of jhbuild :) [19:50] manusheel sir: the size has reduced from 426 MB to 2MB [19:51] alsroot: Ok :-) Whom can we touch base on jhbuild. [19:51] kandarpk: That is a massive decrease. [19:52] manusheel: silbe on #sugar is a jhbuild maint, but looks like he is offline these days [19:53] alsroot: Kandarp does not get all the folders on cloning jhbuild from the mainline git. It seems some issues. Ok, will touch base with Sascha Silbe. Thank you for the pointer. [19:54] kandarpk: Let us write an e-mail to Sascha and Bernie on this issue. [19:54] They should be able to help here. [19:54] kandarpk; In the meantime, let us focus on 2 MB clone. [19:54] Move with the next steps to see the things in action. [19:56] manusheel sir: the source is missing [19:57] kandarpk; Ok. [19:57] We cannot proceed ahead then. [19:57] manusheel sir: there isn't much to do [19:58] kandarpk: Let us see if Bernie is around, and can help us on this. [19:58] bernie: Hi Bernie. Around? [19:58] manusheel sir: I've run the other step on the jhbuild folder I already had [19:59] kandarpk: Ok, how did it come along? [20:00] manusheel sir: sir, the patch file I generated was using the steps we had discussed [20:03] kandarpk: Great. That is neat. [20:03] kandarpk: Can we complete the next 2 steps using that patch? [20:04] kandarpk: Wish to see the missing gaps in the 4 steps. [20:05] manusheel sir: I didn't get what steps are missing ? [20:05] manusheel sir: we only need to submit the patch, isn't it ? [20:08] kandarpk: We are also the reviewers of the patch, and see their conversion to HTML. [20:08] I.e. docstrings to HTML. [20:11] manusheel sir: we'll need to generate HTML using patch file ? [20:19] manusheel sir: I'll leave now. [20:20] kandarpk: Sure. [20:20] Please take rest. [20:20] We have a meeting tomorrow. [20:20] manusheel sir, neeraj : good night [20:20] kandarpk: Please send an e-mail To Bernie. [20:20] And Sascha. [20:21] "Sascha Silbe" , [20:21] manusheel: Ok. [20:21] kandarpk sir, good night [20:21] kandarpk: I'll send you his sugarlabs id. [20:22] Sascha Silbe [20:22] kandarpk: This is the e-mail id for sugarlabs. [20:23] manusheel sir: Ok. [20:25] kandarpk: Great. Please send them an e-mail on jhbuild issue. [20:25]