'morning! How are people today? * wrobel (n=user@gentoo/developer/wrobel) has joined #horde-board morning everyone hello everyone We'll give people another minute or two and then take topic suggestions/questions up front, and try to have a small+short but productive meeting Hey guys I'm on a train with a bad connection so I'll probably just listen Okay. Let's get started. My topic suggestion is H4 releases - where are we with the current code, and how far do we want to go before releasing Anyone else have topics, or questions? Would it be possible to get an update on Ajax Kronolith? (the boss has been bugging me about that) Unfortunately Jan isn't here, and he's the best person to speak to it. Michael (mrubinsk) or Michael (Shpoon), do either of you know the status of that project? mrubinsk would know much better than me. My work has focused exclusively on imp/ I don't have any specific timelines I know that a good bit of the work is done, but there are still significant missing pieces I was going to bring up again the issue of gettext in packages, and how we are going to handle that, but that might be a bit too much code-specific for this forum do you have any ballpark, back of the envelope estimates? Within the year? * ShadowSpawn (n=marcus@horde/marcus) has joined #horde-board I would hope within the year, but again, we really need to ask Jan - as it is really his project. kevinK: within the year is likely from what I know about SAPO's timeline Cool - that'll be good enough for 'em ;) Shpoon: let's take that one up on the dev list I know the client funding kronolith is planning on rolling out the new version of dimp soon I haven't talked to carlos to see if kronolith is required to be part of that release cjh: ok Alright, I wanted to bring up H4 We've got a lot of new stuff in IMP, the new Kronolith coming a lot of libs ported over to git, but not most apps, and not the horde base there are also a lot of thing that I was originally thinking we'd do for the first H4 release, like using Horde_Routes, Horde_Controller, etc. Since those are far away, though, and we want to have more frequent releases, I'd like peoples' thoughts and opinions on what the first H4 releases should include #1 - moving all login function logic to H4. Luckily, this is what I was doing most of yesterday (I have rewritten Auth:: -> Horde_Auth:: already) would it be possible to convert at least horde base so that it supports Horde_{Route,COntroller,View}? I thought that horde base could already do much of that through the rampage.php script I think that as far as releases go, in addition to the framework libraries already moved, we should be able to release the basic groupware apps as well - though that means we will need to move a good deal more libraries over to git first * cjh has quit (Remote closed the connection) but, I agree with cjh, in that we don't need to wait until the apps are Routes capable * cjh (n=Adium@ip65-47-144-166.z144-47-65.customer.algx.net) has joined #horde-board I made it here! re (adium crash) welcome back, cjh :) what's the last that you saw? Hi protagonist Also, I knew there was talk to combine conf.php files and prefs.php files and maybe all the config files as well that may be a bit ambituous, but from a Prefs UI standpoint - it would be nice to implement the Simple/Advanced view, where prefs can be tagged with different priority levels I thought there was a clear separation with conf and prefs between things that are static/set-by-the-admin (conf) and things that are dynamic/set-by-the-user (prefs) but this distinction is fuzzy for some config items and, regardless, the config items can be marked as either private/public in the unified file ok fine by me then :) generally speaking, the sooner you can put some code in users' hands, the sooner you'll get useful feedback on it protagonist: agreed x1000 maybe some -pre release would be in order at least, I'll volunteer to install it in our dev instance and give you feedback well, gmail just came out of beta - maybe we can go into alpha :) so the apps would not be packaged as PEAR packages then? How are we going to handle the framework packages that are still in CVS? I'm assuming we need to port all that are used in a H4 app, like IMP, over to git before doing any sort of release? yes err, yes to which question? mrubinsk. (hit return to fast) too fast wrobel: it doesn't seem we're ready for that just yet IMO, that shouldn't be a requirement to get a release out - even if that is the eventual goal also, I'm all for combining prefs and config vars. the only real difference is whether users can change them; admins can decide whether to lock or unlock for most things mrubinsk: agreed btw, I missed anything between mrubinsk's comment about being able to release the groupware apps, and coming back you didn't miss anything then From an IMP standpoint, my remaining goals are 1.) move login to horde (REQUIRED) 2.) rewrite search code (REQUIRED) and 3.) add support for multiple accounts (NICE TO HAVE) only my comment: but, I agree with cjh, in that we don't need to wait until the apps are Routes capabe eh, capable I would also really like to see us get rid of Horde_Template and Horde_DataTree for good basically anything we know that we're phasing out should never appear in an h4 release, even alpha, in my opinion agreed - doubly so for DT :) i know i disagree strongly with cjh, but I don't see a rush to nuke Horde_Template getting rid of it is a good goal, but for stable apps like IMP it is proven and, most importantly, it is fast and as many users tell me - it is very easy to alter removing HT from H4 is completely on the bottom of the list for m e okay. for now we can agree to disagree on that one, and it's on me to prove that horde_view will be a fast and capable replacement i'm planning on rewriting gollem first as a less-complicated test case that would be great :) as far as DT goes, the only thing still using it, I believe is Trean mrubinsk: I believe you're correct also framesets... oh, yeah we need to nuke those. badly. get rid of them :) although, with a single login screen, framesets become less of an issue albeit still an issue Shpoon: with moving the login code - I'd *really* like to see a new login process be implemented with Horde_Controller so that we have a portable controller for login that can be theoretically dropped into any app i don't want login code in an app - period the login should be done entirely in horde Shpoon: think about things written with the framework code what i'm saying is that the horde base should include a login controller that horde apps use other non-horde but horde-framework apps should be able to use the same controller, or modify it also, implementing it with routes/controllers means not having to introduce more webroot .php files for it hmmm... i am worried that login stuff is too horde specific and i wasn't planning on adding more webroot.php files okay. i'll see what you do and we'll talk So, to wrap things up - we want to move towards H4 alpha releases still very in flux, but getting them out there in a form other than raw source control or snapshots that would be super, from my standpoint Any objections? Anything else to add in summary, or to discuss? not from me s'all good over here Alright. let's call it Since I crashed out in the middle, will someone else please send the minutes ? just cut-and-paste? I can do that yeah, of the log thanks! sure Alright - thanks all. I'm off to lunch before I fall over. :) ciao!