Attendees: clsk, Cupu, J_K9

Note: All timestamps in UTC+1 (= CET, Central European Time).

---

[17:36] clsk: do you have an agenda of what we want to go through?

[17:36] j_k9: yes

[17:36] j_k9: in summary:

[17:37] j_k9: 1) which modules require the most attention

[17:37] j_k9: 2) what we *absolutely* need to accomplish for the 0.1 release

[17:37] j_k9: 3) discuss moving to SCons

[17:38] j_k9: 4) decide who would like to work on which module

[17:38] j_k9: is there anything else we should discuss?

[17:40] clsk: hm I think that should be it

[17:40] j_k9: wrt issue 2) I would like to break it down into a series of small 'tasks' so that we can create a path towards the release

[17:40] j_k9: but we can discuss that when we get there :)

[17:40] cupu: So you;re considering moving to Scons? I know I can't say anything major, but is there any posibility you could also throw CMake in the battle for a build tool?

[17:40] clsk: I think we already looked at CMake and decided for SCons a while ago

[17:41] j_k9: yes, I think so - I think we had a thread on the mailing list about this and decided that SCons was the better of the two for this project

[17:41] cupu: I should probably back-up my nominations so I'll just say I've used it a bit and it an excellent tool .. plus KDE 4 used CMake (while initially looking at Scons)

[17:41] cupu: Ok, just thought I'd say

[17:41] j_k9: cupu: do you know how good its support for Qt 4 is?

[17:42] clsk: and boost

[17:42] clsk: those are our main concerns

[17:42] cupu: I can find the KDE success story around the internet somewhere, the KDE guys really worked to push features in cmake

[17:42] j_k9: I ask because SCons' support for Qt 4 (or Boost) is not great, to say the least

[17:42] clsk: actually boost is more of an issue thant Qt 4 is

[17:43] cupu: I can't say anything about boost .. I'll have to look into it

[17:43] clsk: cupu: are you good with CMake?

[17:43] cupu: what kind of problems does scons rase with boost?

[17:43] j_k9: yes - I couldn't find a way to automatically find the version or location of Boost with SCons

[17:43] cupu: I've only played with it for a couple of weeks .. I can find my way around in theory

[17:43] cupu: *around it

[17:43] clsk: would you be willing to write the build system with CMake then?

[17:44] cupu: I'd be willing to write some CMake examples first this week (well the next actually .. the one that begins tomorrow), post them and see where we go from there, is that ok?

[17:44] cupu: Well .. not hello world examples .. but you get the idea

[17:45] j_k9: that sounds great. are there any plans to follow this through as well: https://bugs.launchpad.net/mira/+bug/226814 clsk?

[17:45] j_k9: (it's more of a feature request than a bug)

[17:45] cupu: thing is, CMake generates eclipse, visual studio, code::bloks and Kdevelop build files

[17:45] cupu: And makefiles of course

[17:46] clsk: oh cool

[17:46] clsk: I use code::blocks in windows

[17:46] clsk: j_k9: that would be part of the build system

[17:47] clsk: and yes, that's more of a feature request. Is there a way to post feature requests in launchpad?

[17:47] cupu: It's a good IDE, afaik CMake generates “build” files but not “project” files. In the way that you can build the project but not work on it, I have to check on this though .. maybe it's a bit more flexible still

[17:47] clsk: yea I know.

[17:48] j_k9: would the build system (i.e. CMake, in this case) need to be edited in order to compile new files, for example? or is your suggestion, clsk, to avoid having to do this manually?

[17:48] clsk: I don't need project files. I like to arrange my project files in a different way that a regular person would

[17:48] cupu: clsk agreed

[17:49] clsk: j_k9: it would have to be every time a person creates a new file that has to be compiled. But the changes should be minor. There's no way around that.

[17:49] clsk: actually there might be. That's not really a big concern though

[17:49] j_k9: thank you, just thought I should mention it in case it was something cupu should consider when writing the CMake scripts

[17:49] cupu: I've only manually added source files to CMake files, but it's a pretty powerfull system and it should be possible to automatize this

[17:50] clsk: if you can do it…by all means. But it's not a priority right now

[17:50] clsk: priority is to identify the version of boost that user has and where it is. And Qt.

[17:50] clsk: and to make sure it works in windows, *nix and Mac

[17:50] cupu: right so I'm supposed to make some Cmake files that show it off next week, right?

[17:51] cupu: noted

[17:51] j_k9: by the way, if you're interested in reading later: KDE CMake success story (http://lwn.net/Articles/188693/) and comparison with other build systems (http://ubuntuforums.org/showthread.php?t=692692) - CMake looks quite promising

[17:51] j_k9: cupu: yes, that'd be great

[17:52] cupu: oh also … I'm surprised it took me so long to notice … but great thing moving to QT4

[17:53] j_k9: thanks to Thurston :) it looks and works better

[17:54] j_k9: ok, we've decided that cupu is going to look into CMake. shall we discuss point 1)?

[17:54] clsk: hm sure

[17:55] j_k9: clsk you're the most familiar with the code: what do you think needs the most attention at the moment?

[17:56] clsk: well I need to add some stuff to the network layer

[17:56] clsk: that's easy

[17:56] clsk: I'd say the directory layer is what needs the most attention right now

[17:56] j_k9: anything specific about the directory layer?

[17:57] clsk: just in general

[17:57] clsk: We might even need to discuss its design again

[17:57] j_k9: ok - should we do that on the list?

[17:57] clsk: but code-wise. Not feature-wise, like we did last time.

[17:58] j_k9: right

[17:58] j_k9: would you like me to email the list or would you like to kick of the thread with some technical suggestions of your own?

[17:58] j_k9: *kick off

[17:59] clsk: I can start it

[17:59] clsk: I just need to look at the old code again

[17:59] j_k9: ok thanks

[18:00] j_k9: you also suggested we might need to work on the GUI. do you mean tying it into the rest of the codebase so that the GUI elements actually work?

[18:00] clsk: yes

[18:01] j_k9: now that Thurston is unable to contribute I think we're going to need to find someone else to work on this. cupu do you have any experience with Qt?

[18:02] cupu: next to none actually, I know the concepts but never done more than a window or a few sliders or the like

[18:02] j_k9: ok, I'll email the list about this then. if not, I'll try to recruit a developer with Qt 4 experience to work specifically on this.

[18:03] j_k9: cupu, after working on the build system is there any part of the project your would like to work on in particular?

[18:03] cupu: I don't mind learning Qt, just that I can't take a task on it right now

[18:05] cupu: I'd really work on anything right now, just so that I might get to write some code, get in the pace and all that jazz, mira-wise I have tons to catch up to

[18:05] cupu: Sorry if it's not clear, sometimes I forget someone else must understand my english

[18:05] clsk: for now you should probably just concentrate on the build system

[18:05] clsk: and then pick up the GUI layer

[18:06] cupu: okey

[18:06] j_k9: cool

[18:06] j_k9: clsk, are you going to keep working on the network layer or are you going to work on the directory layer?

[18:07] j_k9: cupu: your english is better than you think..

[18:08] clsk: j_k9: I'll probably be working on both.

[18:08] clsk: I want to finish the base for the network layer first though

[18:08] j_k9: ok. do you need any help with the directory layer?

[18:08] clsk: we could always use some help :)

[18:08] clsk: there just isn't anyone else

[18:09] j_k9: well, there is Justin. I just mean whether he should help you or work on another module and, if so, which module.

[18:09] clsk: well he could probably start working with me on the directory layer

[18:09] clsk: then

[18:09] clsk: that's if he actually ends up writing any code.

[18:11] j_k9: he seems quite eager to contribute so I think he will. I ask him if he'd like to help you on the directory layer, and if he says yes then I suppose he'll reply to your thread on the list

[18:11] j_k9: *I will ask him

[18:11] clsk: ok

[18:11] clsk: moving on?

[18:12] j_k9: yes, point 2) - let's discuss the requirements for Mira Client 0.1 and Server 0.1

[18:12] j_k9: *especially* outstanding requirements

[18:12] clsk: what were our requirements last time we discussed this? Do you remember?

[18:13] j_k9: they are on the wiki: http://miragroupware.org/wiki/doku.php/development/roadmap

[18:13] j_k9: we said we would include a semi-functional Files Utility for the 0.1 release. is this reasonable?

[18:14] clsk: We'd have to work on the utilities layer

[18:14] clsk: which hasn't been started yet at all.

[18:14] j_k9: right, and in any case that would require the file system backend in order to work (which I don't think has been started either)

[18:15] j_k9: shall we make this a requirement for 0.2 instead?

[18:15] clsk: oh that would be easy

[18:15] j_k9: ah

[18:15] clsk: hm.. sure

[18:15] clsk: it's a pre-alpha anyways

[18:15] j_k9: exactly

[18:16] clsk: so requirements stay the same

[18:16] clsk: except for no File utility

[18:16] clsk: and no utility framework at all

[18:16] clsk: utilities should be main focus for 0.2

[18:16] j_k9: so in terms of the Client's requirements for 0.1, it needs to authenticate with the server (which it does). should 0.1 also allow users to communicate using IM? weren't you working on this?

[18:17] j_k9: agreed

[18:17] clsk: hm

[18:17] clsk: this shouldn't be hard

[18:17] j_k9: it only involves the GUI and Communication layers

[18:17] j_k9: because the IM element isn't actually a Utility

[18:17] clsk: so yea, it can be done

[18:17] j_k9: great, I'll add that.

[18:18] j_k9: we've made good progress. is there anything else to discuss/any questions?

[18:19] clsk: hm I think that's about it

[18:19] clsk: cupu: do you have anything to add?

[18:19] j_k9: for the record: I'm going to be contacting every Core developer (except the ones who have replied to the list) individually to see if they are still active. if not, I'll remove them from launchpad developer membership, sourceforge etc

[18:20] clsk: yep

[18:20] clsk: also from the wiki

[18:20] clsk: ah one more thing

[18:20] clsk: the webpage

[18:20] j_k9: yes?

[18:20] clsk: when I open miragroupware.org the news article (Hello World) is in some weird language

[18:20] cupu: I've been meaning to suggest the ACE framework as “food for thought” but it's learning curve is preatty steep and well I have to see first if there's a place where boost strugles. So yeah .. nothing from me now :)

[18:21] cupu: Latin stubs ;)

[18:21] cupu: that was supposed to be an ”:)”

[18:21] cupu: I'm not sure if I'm part of the launchpad project, I'll check now

[18:21] j_k9: clsk: I'm going to remove that article today and replace it with a custom welcome message in any case.

[18:21] clsk: ok

[18:21] clsk: I think we used to have some articles up there. What happened to those?

[18:22] j_k9: cupu: let me know if you're not and I will add you. I think you need to be in order to have commit privileges in the bzr branch

[18:22] j_k9: clsk: they were deleted by my hosting company, which is rather annoying…

[18:22] clsk: cupu: We've already discussed ACE a while ago. We went for boost because it's more modern C++

[18:22] j_k9: actually, I'm going to look on archive.org and see if they're up there. thank you for reminding me.

[18:22] cupu: I don;t think I'm on launchpad

[18:22] j_k9: cupu: what's your launchpad name?

[18:23] cupu: clsk: No doubt about the “modern c++” part, I agree

[18:23] cupu: j_k9: “Cupu” email: stefan.constantin AT gmail.com

[18:23] j_k9: cupu: thanks, I'll add you asap

[18:23] cupu: I just have a soft spot for ACE … wipe it from the record :)

[18:23] clsk: cupu: We left it open for implementation at a later time though. That's why the network code is somewhat modularized.

[18:24] clsk: It'd be easy to write a plugin to use another network framework.

[18:24] j_k9: but asio is working well so far, isn't it?

[18:24] clsk: yep

[18:24] cupu: cool

[18:24] clsk: asio is working nicely so far.

[18:24] j_k9: excellent

[18:25] cupu: the cool thing about ACE is the way it uses some software patterns and it's portability, but just to be clear I didn't want it to sound about “something vs. boost”; if something works (and boost *works!*) it's all good

[18:25] j_k9: by the way, would either of you like an account for the homepage? it'd allow you to write articles on your latest progress etc. entirely optional, just in case you want one

[18:26] clsk: I think I used to have one

[18:26] cupu: I don't think I need one now

[18:26] clsk: I probably won't be writing articles though

[18:26] j_k9: ok

[18:27] j_k9: no luck with archive.org, btw - I'll write some new articles every week or so to start building some interest

[18:27] clsk: ok

[18:27] clsk: we need developers too ;)

[18:27] clsk: If we could recruit a few, that'd be great

[18:28] j_k9: yes, I'm going to go on another SourceForge recruitment run ;) hopefully this time we can get the developers straight into the code, if not they get bored and quickly lose interest

[18:28] clsk: yep

[18:28] clsk: Also I think we should hang out here on irc. That way new people get interested… being able to talk to us

[18:29] clsk: like whenever we're on

[18:29] j_k9: sounds good

[18:29] clsk: just open up an IRC client

[18:29] cupu: cool

[18:30] j_k9: great. anything else?

[18:30] clsk: nop

[18:30] clsk: I think that's it

[18:30] cupu: nothing

[18:30] j_k9: ok - I'll send a summary of this meeting to the list and post the transcript to the wiki. thanks for attending :)

[18:30] cupu: great talking to you guys again, cheers!

[18:31] clsk: sounds good

[18:31] clsk: bye now

[18:31] cupu: good bye

[18:31] j_k9: bye all!