PMC Functional Meeting 20100422
Date: 2010-04-22
Time: 9PM GMT
Venue: irc #adempiere-team
Support Spreadsheet: Adempiere PMC Functional
Chat times in GMT-5
(16:00:13) CarlosRuiz: Hi
(16:00:23) ar_howard: Good morning
(16:01:02) CarlosRuiz: let me check for Joel's availability ...
(16:01:20) ar_howard: Steven also told me yesterday he would be online
(16:04:51) CarlosRuiz: ok - if you want we can start and they will catch up later
(16:05:06) CarlosRuiz: please if you can ping steven also in skype
(16:06:30) CarlosRuiz: Proposed agenda for today:
(16:06:30) CarlosRuiz: review proposal trying to ease and improve the process to contribute
(16:07:23) ar_howard: I've sent Steven an SMS
(16:07:57) CarlosRuiz: ok - red1? trifon? jmpiloq? are you here?
(16:08:12) CarlosRuiz: agree with proposed agenda?
(16:08:12) CarlosRuiz: any other topic you want to discuss?
(16:08:49) ar_howard: We are a bit slow preparing an environment for Adaxa to work in (other business priorities). We are setting up a powerful server in our DR facilities to house a copy of our environment which is to be slimmed down to a template environment for uploading and review (ie. database and code)
(16:12:46) CarlosRuiz: Allan - if you want we can review a proposal - or we can wait for Steven
(16:15:23) ar_howard: sorry - on phone trying to help steven get connected
(16:15:28) trifon: yes, i'm here, but not full time.
(16:15:34) trifon: i have to put my daughter to sleep
(16:17:02) ar_howard: I have got steven onto #adempiere, but it will take a bit too lost time to get him onto adempiere-team : do you think we could switch to there for now?
(16:18:23) JoelS [~4b64bbbd@gateway/web/freenode/x-kqablvqqiujexrvf] entered the channel.
(16:18:54) ar_howard_ [~76d0128d@gateway/web/freenode/x-ialnclgoczlkzreq] entered the channel.
(16:19:23) steven__ [~3aa13191@gateway/web/freenode/x-smulppfcqdkkyant] entered the channel.
(16:19:30) ar_howard: Hi Steven
(16:19:49) CarlosRuiz: Hi Joel and Steven
(16:19:55) ar_howard: I'm off phone now and steven is in #adempiere-team
(16:20:26) ar_howard_ has left the channel.
(16:20:26) JoelS: Hi Steven, Allan, Carlos, everybody
(16:20:56) CarlosRuiz: next time - the easiest way is to navigate to this link
(16:20:56) CarlosRuiz: http://webchat.freenode.net/?channels=adempiere-team
(16:20:56) CarlosRuiz: and join
(16:21:48) CarlosRuiz: so, Proposed agenda for today:
(16:21:48) CarlosRuiz: review proposal trying to ease and improve the process to contribute
(16:23:05) ar_howard: We are currently working with Adaxa to provide a powerful DB machine which can be used to run some automated routines to shrink the size of our database for uploading with code.
(16:23:39) ar_howard: This will be on our DR machine, so it is conceivable we could open it up to remote access for trialing of the functionality
(16:23:42) JoelS: doing some automated cleanup?
(16:24:11) ar_howard: Steven - would you like to describe the recursive delete functionality and what it can achieve?
(16:24:24) steven___ [~3aa13191@gateway/web/freenode/x-osgegrqezrlpmggh] entered the channel.
(16:24:42) steven___ has left the channel (quit: Client Quit).
(16:26:17) steven__ has left the channel (quit: Quit: Page closed).
(16:26:34) steven__ [~3aa13191@gateway/web/freenode/x-jfwxfnikbqfdqyar] entered the channel.
(16:27:01) ar_howard: Well looks like Steven might still be having some trouble - I'll describe this the best way I can. The functionality allow us to target a particular tenant and table, and then the system works out how to delete that entity and any linked entities
(16:28:45) JoelS: so it leaves some sample data, or does a clean sweep?
(16:28:50) ar_howard_ [~76d0128d@gateway/web/freenode/x-yyxsujpirusbytly] entered the channel.
(16:29:04) steven__: test
(16:29:10) ar_howard_ has left the channel.
(16:29:26) steven__: Hi All, sorry for my dumbness on this tool
(16:29:40) ar_howard: This way we can remove a lot of operational data and give us an easier place to start creating a sample database
(16:30:19) ar_howard: Steven - I was just talking about how we are setting up a copy of our production system to slim down and de-identify data
(16:30:36) ar_howard: and Joel was asking about the automated cleanup
(16:31:31) ar_howard: I also mentioned that we may be able to look at giving others remote access, but of course we have never tested our installation against the webgui (e.g. call centre screens, etc)
(16:32:49) steven__: Hi Joel Paul created a process called 'delete entities' . when you run it you select an entity like say Invoice it looks through all the related things, lists them and then says do you want to delete..It would then delete all invoice tax invoicelines, invoices etc. for some related things it nulls them rather than delete
(16:34:00) JoelS: that's good
(16:34:03) steven__: We plan to contribute it but have not had time .. we have shared with others who had an interest.
(16:34:16) ar_howard_ [~ar_howard@ppp118-208-18-141.lns20.bne1.internode.on.net] entered the channel.
(16:34:52) McBoss [~ca59367a@gateway/web/freenode/x-sftlhhixanurjuxu] entered the channel.
(16:35:31) McBoss: Excuse the delay - customer call
(16:35:40) steven__: We developed it so that when we testing Allans imported data Allans people could just undo what they had done. In fact we used it to delete every tBP, Product and transaction but still leave an otherwise set up client with all lists, reports, config etc still in place
(16:36:16) JoelS: nice tool
(16:36:37) steven__: I will get Paul to put in contributions for moment
(16:36:57) JoelS: carlos, was there an aspect of 'easing contributions' that you wanted to focus on today?
(16:37:54) CarlosRuiz: yep - we would like to validate a proposal to ease contribution process - if you agree we can review that here
(16:38:05) JoelS: sounds good
(16:39:54) CarlosRuiz: Steven? Allan? Peter? do you agree - or you prefer to discuss a different topic?
(16:40:09) McBoss: My interest backs into this topic and relates to some comments from last week - just how to advertise what contributions and more what finesses are available - some functional descriptions
(16:40:16) ar_howard_: Agreed
(16:41:03) CarlosRuiz: ok - so, let me try to round up the idea
(16:41:47) CarlosRuiz: we think the current contribution process has some flaws that can be stoppers for implementors
(16:41:49) McBoss: I'm late to the meetinh and we should probably concentrate on teh things you've already got underway - so if easing contributions is a common concern I'd go with that.
(16:42:26) CarlosRuiz: * The current process is not helping the definition of new functionalities or improvements to existing functionalities.
(16:42:26) CarlosRuiz: * Communication via forums can be easily misunderstood.
(16:42:26) CarlosRuiz: * Decision making is almost impossible, one single comment can stop a good idea.
(16:42:26) CarlosRuiz: * The process is not structured - trying to cover all the edges of designing a solution.
(16:43:02) CarlosRuiz: so, Joel and I met yesterday to think how can we make this process different - more agile - and more prone to be attended
(16:44:22) CarlosRuiz: so - there is a proposal of three big points:
(16:44:22) CarlosRuiz: 1- we need to have a single place to document the suggestion (forum or wiki, not random at the whim of person)
(16:44:22) CarlosRuiz: 2- we need to have a simple format for expressing the key issues (what it does, why needed, what corollary consequences)
(16:44:22) CarlosRuiz: 3- we need a schedule- 1 week in forum, 1 week on Func agenda, announce the change, 1 week to code merge (expressed as minimum times, not as strict times)
(16:44:37) CarlosRuiz: and we have a more detailed suggestion for the first two at least
(16:45:10) CarlosRuiz: for the first
(16:45:10) CarlosRuiz: we propose to create a new tracker type called ASR - Adempiere Specification Request
(16:45:38) CarlosRuiz: same as a feature request with an URL field that must point to a page or document with the format
(16:45:38) CarlosRuiz: and different workflow of Status ->
(16:45:38) CarlosRuiz: Open
(16:45:38) CarlosRuiz: Documented (when the document is complete)
(16:45:38) CarlosRuiz: In Review by PMC (meaning that PMC has taken this tracker for review on meeting or any mechanism)
(16:45:38) CarlosRuiz: Improvements to the document requested
(16:45:38) CarlosRuiz: Approved
(16:45:38) CarlosRuiz: In development
(16:45:38) CarlosRuiz: Pending->Closed
(16:46:47) CarlosRuiz: do you want to discuss this now - or must I continue with the second part of the proposal and we discuss at the end?
(16:47:33) McBoss: Keep going thedetails look pretty good - might want to consider timings ...
(16:47:52) ar_howard_: I'd suggest going to see big picture
(16:47:54) CarlosRuiz: ok - the second part of the proposal is about the format of the specification request
(16:48:47) CarlosRuiz: 2- we need to have a simple format for expressing the key issues (what it does, why needed, what corollary consequences)
(16:48:47) CarlosRuiz: conditions:
(16:48:47) CarlosRuiz: * making document simple to fill out
(16:48:47) CarlosRuiz: * easily editable by others (collaborative editing)
(16:48:47) CarlosRuiz: * it must allow to attach graphics, mindmaps, screenshots, flows
(16:48:47) CarlosRuiz: * user must not have to learn a new language (like mediawiki)
(16:48:47) CarlosRuiz: * strict format with fixed structure
(16:48:47) CarlosRuiz: candidates can be google docs or foswiki or a wysiwyg editor for mediawiki
(16:48:47) CarlosRuiz: my considerations:
(16:48:47) CarlosRuiz: - google docs have the problem that are not searchable in google - even if publicly published
(16:48:47) CarlosRuiz: - wysiwyg editor for mediawiki - I researched and it seems they are all in alpha or beta status
(16:48:47) CarlosRuiz: - so, my preference would be to research about foswiki - we'll need a test server on amazon or rackspace to check the factibility
(16:50:16) CarlosRuiz: and on the third part of the proposal we didn't have a clear agreement about timing - but we both think that these functional meeting must be exploited for decision making
(16:50:39) CarlosRuiz: as an example Joel proposed - we can call for a functional meeting stating ...
(16:50:57) CarlosRuiz: in the apr 29 functional meeting we're going to discuss cash payment requirement
(16:51:12) CarlosRuiz: and if possible take the necessary decisions on that meeting
(16:51:34) CarlosRuiz: those who don't attend can discuss in forums - but cannot stop the decisions made to integrate in release
(16:51:41) CarlosRuiz: something like that
(16:51:43) JoelS: carlos- you might check Google Sites as well
(16:51:52) JoelS: it is their version of wiki
(16:52:11) CarlosRuiz: good tip Joel - I'll do
(16:52:25) CarlosRuiz: so, that's the summary I have - did I miss something important Joel?
(16:52:26) JoelS: the timings to me are more like drawing a line in the sand so we can make progress
(16:52:32) JoelS: no- you did great
(16:53:02) steven__: can we ask sourceforge if they would deploy foswiki if they have nothing like it already?
(16:54:06) trifon: othe wiki which has wysiwyg editor is xwiki. java based.
(16:54:20) trifon: but google sites is good starting.
(16:54:26) CarlosRuiz: at this moment I don't see support for those in sf - just for mediawiki and trac
(16:54:26) trifon: free and immediately avaialble.
(16:55:11) CarlosRuiz: what I liked (and would like to research deeper) is that foswiki allow structured content
(16:55:14) CarlosRuiz: on the document
(16:55:28) JoelS: what does the group think of the general concept? Do you see potential for us to use the process to move faster?
(16:55:31) McBoss: Is their a sense in the functional description in stage two, of a "Core of Functionality" and some extensions from the core - Is it really possible for a good functional description to cover all requirements from everyone - I sometimes think that's what takes the time in the forum?
(16:56:43) McBoss: That and negotiating function point by point without a design whole view
(16:57:12) CarlosRuiz: I think we can differentiate about implementing everybody's needs - and just taking into account and leaving some for future implementation
(16:57:47) CarlosRuiz: trying to implement everybody's needs for some topics is not feasible - but we need to hear, take into account, and at least prepare the scenario for further improvement
(16:58:03) CarlosRuiz: but I don't like minor requirements stopping big improvements - that happens a lot
(16:58:23) steven__: I have some trouble thinking about these things in abstract terms but at the moment I am seeking feedback and suggestions about using Payments and Carlos and Croo have responded
(16:58:56) ar_howard__ [~ar_howard@58.104.161.103] entered the channel.
(16:59:00) McBoss: I agree with minor holding up Major - will deciding the core be part of the PMC role?
(16:59:06) steven__: but even afeter a few responses I wont have a good sense of whether it is something everyone else wants in the core product
(16:59:33) steven__: will this proposal have a solution for that type of issue?
(16:59:57) CarlosRuiz: @McBoss -> yes - deciding which functionality go into core (and how) is in the scope of PMC role
(16:59:58) McBoss: ask steven and ar_howard they have wide financial experience too
(17:00:19) CarlosRuiz: Steven - precisely that's the point
(17:00:38) CarlosRuiz: if two opinions in forums are incompatible
(17:00:40) CarlosRuiz: or opposite
(17:00:44) CarlosRuiz: then the whole thing tends to stagnate
(17:01:20) CarlosRuiz: what we propose is to have a mechanism to try cover all the edges in an organized way
(17:01:20) CarlosRuiz: and a mechanism to take decisions (probably on meetings) when opinions collide
(17:02:28) ar_howard_ has left the channel (quit: Ping timeout: 252 seconds).
(17:02:31) ar_howard__: Forums are a bad vehicle - only the vocal are heard, a bit like minority rules
(17:02:49) McBoss: I don't know the payments problem well enough so can't contribute to specifics
(17:03:39) McBoss: Also typed forums even like this one don't provide enough colour to the conversation to point to a compromise ...
(17:04:20) steven__: we have stopped using the voting mechanism. I thought it was helpful even though I often did not get my "ticket"
(17:05:15) JoelS: the idea of providing the decision at a scheduled meeting is to help resolve those issues
(17:05:21) red1: its predawn here.. reading back..
(17:05:45) ar_howard__: Bnnmm
(17:05:51) CarlosRuiz: in functional and technical case - voting is just a measure of popularity - but not a measure of completeness of the proposal - or proper discussion
(17:06:05) steven__: scheduled meeting to resolve sounds good to me.
(17:06:06) JoelS: if there are two conflicting POV, then if both care enough to show up at the meeting, we can discuss it out
(17:06:37) red1: agrees with JoelS
(17:06:46) JoelS: if after an hour of mediated discussion, no agreement can be reached, then the idea is not baked well enough yet
(17:07:41) red1: thats right.. at least both sides presented and its printed in discussion
(17:07:44) JoelS: but mostly, functional issues can be talked through quickly in a chat
(17:07:56) McBoss: I disagree with JoelS when there is complex inter country (take tax) rules and variation on process it can take a bit of getting a resolution even with good will
(17:08:40) JoelS: McBoss- complexity like that needs to be handled with flexible, configurable architecture
(17:08:58) McBoss: However a scheduled meeting is the only vehicle right now
(17:09:08) JoelS: if someone proposes a hard-coded change that takes away flexibility, the answer is simply no
(17:09:25) JoelS: we wouldnt have to understand the complexity of all the countries
(17:09:30) steven__: I think we will all agree to this meeting dealing with major changes and we should proceed with a plan for that. How will we identify those small things that just get put in - usually without damage and for good reason?
(17:09:31) McBoss: I agree with that JoelS
(17:09:33) ar_howard__: Now mobile - I may drop out
(17:09:42) JoelS: to know whether a new idea adds options or takes them away
(17:10:15) JoelS: Carlos- do you want to describe the bug discussion for steven?
(17:10:27) CarlosRuiz: the bug process? yes, that's different
(17:10:41) McBoss: I think Carlos is looking for a mechanism with as much talent and wise heads as he can assemble in one place to get a resolution out and development started
(17:11:09) CarlosRuiz: Process for bugs must be straight:
(17:11:09) CarlosRuiz: 1 - discussion in forums just if you're in doubt if it's a bug
(17:11:09) CarlosRuiz: 2 - if you're sure it's a bug - open a bug tracker
(17:11:09) CarlosRuiz: 3 - if you have a solution - upload it (commit it if you have permissions)
(17:11:09) CarlosRuiz: 4 - if you're really sure about the fix you can commit on release - otherwise you commit on trunk for peer review
(17:11:38) CarlosRuiz: is that Joel?
(17:11:54) steven__: well some new things dont take away but they can be argued to be an unnecessary duplication in the data model
(17:12:55) steven__: like for example the suggestion I made to add Bpartner to allocationhdr
(17:14:08) JoelS: that's mostly it Carlos, but I think we also said that a bug would be something where it 'doesnt work like it's supposed to'
(17:14:31) CarlosRuiz: ah - right
(17:14:33) JoelS: but if it's a bug like "it works like this, but it should work like that" then we need to bring it here
(17:14:53) JoelS: for discussion and agreement before promoting it to release
(17:15:33) JoelS: i'm hoping we can move things more quickly through a pipeline like that
(17:15:41) JoelS: than we can in forum discussion
(17:16:10) JoelS: Paul still probably has a dozen good things sitting in trackers that just sit there because there was a little back and forth with no resolution
(17:16:31) JoelS: if we grab those and move them through a process- even 3 weeks is better than nothing
(17:16:33) ar_howard__: Hi Joel, sounds correct but it is lack of documentation which gives opportunity to confusion on trying to decide how something should work
(17:17:51) CarlosRuiz: yes - that's why we need to combine the ASR format + forums discussion + meeting to take decisions if they were not agreed in forums
(17:18:07) CarlosRuiz: we don't pretend to drop the important discussion in forums
(17:18:22) CarlosRuiz: or, in this case, in a tracker called ASR
(17:20:20) JoelS: ar_howard-> that can be true- though I believe if this group is comprised of the right people we can discern that quickly... Steven, Carlos, myself, others that have multiple implementation experience can usually discern pretty quickly how disruptive something is
(17:20:49) JoelS: i know pretty quick if it 'smells good'
(17:21:22) steven__: I need to leave. thanks to Joel and Carlos for the suggestions. they smell good ;-) bye
(17:21:43) JoelS: no doubt there will be some tough issues like 'fixed assets'
(17:21:59) CarlosRuiz: thanks steven__ for attending
(17:22:00) JoelS: that are loved and hated because they are big or unclear
(17:22:05) JoelS: bye steven
(17:22:15) ar_howard__: Cya steven
(17:22:19) JoelS: but if we can expedite the 80% it will be a big win
(17:22:28) McBoss: Bye Steven
(17:22:57) ar_howard__: If we successfully move past tough issues, we have the right model
(17:23:19) CarlosRuiz: well - if Allan, Peter, Redhuan, Trifon also thinks that the proposal smells good - I would propose to write a wiki page with the proposal and call for forum discussion on this one
(17:23:32) McBoss: Are we going to give the new process a try on a nominated problem - Payments? See if it resolves and refine from that?
(17:23:54) CarlosRuiz: I think many people is interested in "Cash Payments"
(17:24:03) ar_howard__: Agreed carlos
(17:24:35) CarlosRuiz: so, we could use it to test the process
(17:24:36) trifon: +1 to move.
(17:24:38) McBoss: I don't mean to hold anything up just a clearing round down th eprocess in parallel to casting in stone
(17:24:59) McBoss: +1 to move
(17:25:41) JoelS: +1
(17:26:36) CarlosRuiz: ok - let's move this to wiki and forums
(17:26:43) JoelS: Carlos, I suggest you post the procedure as ratified, and that functional group will begin implementing it, but there is room for input if someone has ideas for improving
(17:27:54) CarlosRuiz: I'll write the wiki page and we review it Joel
(17:28:17) JoelS: ok, I can do the post if you want
(17:28:32) CarlosRuiz: sure
(17:28:32) CarlosRuiz: good outcome - anything else for the meeting?
(17:29:15) JoelS: just to clarify- as soon as we post the procedure, we'll also start Cash Payment into the queue
(17:29:50) JoelS: so maybe that will be the agenda two weeks from now?
(17:29:56) CarlosRuiz: yes - one last question
(17:29:58) McBoss: What tool have you decided on for specification Etc
(17:30:03) CarlosRuiz: I'll try to create the ASR tracker
(17:30:28) CarlosRuiz: is it better to create a new mail list for notifications on this?
(17:30:44) CarlosRuiz: I'm concerned that many key people are not following forums and trackers because of the heavy traffic
(17:31:07) CarlosRuiz: or I can keep simply using the current mail list
(17:31:14) CarlosRuiz: WDYT?
(17:32:15) McBoss: I think that keeps the process in front of everyone
(17:32:33) ar_howard__ has left the channel (quit: Quit: Colloquy for iPhone - http://colloquy.mobi).
(17:32:35) red1: i think ppl like to use a maillist where they can reply to.. the one that ppl used to use traditionally
(17:32:57) CarlosRuiz: adempiere-trackers is followed today by 53 persons
(17:33:14) red1: a more public mail list.. if we want ALL to know
(17:33:24) red1: 53 does not sound that many
(17:33:40) red1: (have to go.. morning meeting)
(17:33:42) CarlosRuiz: cvslog is followed by 87
(17:33:47) CarlosRuiz: citizens is followed by 65
(17:34:36) McBoss: would citizens be a choice?
(17:34:38) CarlosRuiz: My opinion is that the heavy traffic of forums and trackers made some key people to unsubscribe from trackers mail list
(17:35:27) CarlosRuiz: so - I would propose to create a new mail list called adempiere-a-s-r and let the people interested subscribe there (the traffic will be lower than forums so I would expect more people's attention)
(17:36:12) red1: would rather see a reuse of present lists.. i.e. citizens.. (gtg)
(17:37:02) CarlosRuiz: to avoid the hassle of people resubscribing - I could do that also - subscribe the current people from trackers on a-s-r
(17:37:59) McBoss: that would cover the "significant" members
(17:38:20) CarlosRuiz: I'm checking and some really key people are not subscribed
(17:38:23) McBoss: those that might contibute further ideas
(17:38:36) CarlosRuiz: so - I could also invite them
(17:38:46) red1 has left the channel (quit: Quit: red1).
(17:38:51) McBoss: are you at liberty to add some in?
(17:39:18) CarlosRuiz: yes - but that won't be kind :-)
(17:39:46) McBoss: could we ask if they'
(17:40:02) CarlosRuiz: yes - I can sent a subscription - or an invitation to subscribe
(17:40:26) CarlosRuiz: I would propose to subscribe all the 53 currently subscribed to trackers
(17:40:26) CarlosRuiz: and invite those key persons that are not subscribed
(17:41:47) McBoss: Sounds fine
(17:42:46) CarlosRuiz: ok - sorry the meeting extended beyond planned
(17:43:07) CarlosRuiz: thanks a lot for attending the meeting
(17:43:25) McBoss: Carlos - I only subscribe to the General Discussion tracker - will that include me or do I have to do something further?
(17:43:56) McBoss: excuse me - forum
(17:44:43) CarlosRuiz: ah - in such case you must subscribe to the new ASR
(17:44:58) McBoss: How is that done?
(17:45:24) CarlosRuiz: how are you subscribed to General Discussion forum?
(17:45:45) CarlosRuiz: clicking on Monitor I guess
(17:45:48) McBoss: just through Sourceforge
(17:46:03) McBoss: yes
(17:46:27) CarlosRuiz: trackers don't have such functionality - so to workaround that we opened a mail list and send all the news of trackers to the mail list
(17:47:02) CarlosRuiz: I can add instructions for that
(17:48:45) McBoss: Thanks - will hear from you during the week - I've got to go to another meeting. Sorry for the late start - bye
(17:48:51) CarlosRuiz: bye
