CC Meeting Full 20070227
hengsin [n=chatzill@60.48.170.8] ha entrado en la sala.
vpj-cd_ [n=vpj-cd@200.93.230.45] ha entrado en la sala.
vpj-cd_ Hi
CarlosRuiz Hi Victor
vpj-cd_ Hi Low , Carlos
trifon [n=chatzill@home-213-240-227-143.megalan.bg] ha entrado en la sala.
Jaztek [n=juddm@87-194-66-231.bethere.co.uk] ha entrado en la sala.
_vclark [n=vclark@c-24-9-155-147.hsd1.co.comcast.net] ha entrado en la sala.
karsten-thiemann [n=karsten-@extern.schaeffer-ag.de] ha entrado en la sala.
CarlosRuiz Hi everybody, today the theme proposed by Heng Sin is just to check on 315 status
trifon Hi Carlos,
hengsin Hi trifon
trifon Hi Heng Sin
vpj-cd_ ok we can start?
hengsin sure
CarlosRuiz ok, I think these days are being very moved in trunk
CarlosRuiz that's very good, but the stabilization of trunk is compromised
vpj-cd_ yes
vpj-cd_ we can make sumarry what is main change to 315
trifon yes i agree with Calros
CarlosRuiz 1st question
CarlosRuiz are we going to stick with plan - freeze the trunk and release 3.1.5 in 28th?
CarlosRuiz I think yes, but want to hear if someone have different opinion
karsten-thiemann no I think we should stick to the plan
hengsin carlos, what is your definition of 'freeze the trunk' ?
CarlosRuiz freeze the trunk -> just allow bug fixings, nothing new
karsten-thiemann feature freeze
CarlosRuiz yes, sounds better feature freeze :-)
CarlosRuiz we're going to freeze feature until 3.2 released
hengsin shall it be only bug fixing for bugs of higher priority ?
karsten-thiemann no - every bug fix is a good bug fix :)
CarlosRuiz low priority bugs allowed too, the goal is to have the most stable version ever seen
CarlosRuiz :-D
hengsin yes but there always the risk of a bug fix that create more bug ...
vpj-cd_ yes
karsten-thiemann ok - but then we have only handful of bugs with priority higher than 5
vpj-cd_ I think we need test all new changes before release
trifon so we will free trunk for 2 weeks?
CarlosRuiz until march 20
trifon 3 weeks.
CarlosRuiz maybe we can allow all bug fixings until march 15, and just high priority until march 20
vpj-cd_ yes we can use 2 or 3 week to test
karsten-thiemann exactly what I was typing :)
hengsin I means when in customer project, when you are close to go-live schedule, you don't fixed every bug, you only fixed low risk and high priority bug.
CarlosRuiz karsten, are you going to integrate filesystem attachment?
karsten-thiemann yes - today
karsten-thiemann fixed a last bug and deathmeat is testing it right now
CarlosRuiz how is the mechanism? old behavior is allowed too?
karsten-thiemann sure
karsten-thiemann you choose method of storing at client level
karsten-thiemann I just add some methods to MAttachment that will store/load on file system
CarlosRuiz good
trifon on file system of each clinet?
trifon i mean each user in the local file system?
karsten-thiemann no you need a shared folder
trifon does it means that every user need to have access to this forlder? Samba?
hengsin karsten, I guess you mean the user have to ensure it is a share folder
karsten-thiemann you define a windowsRootPath and unixRootPath and all clients must have access
karsten-thiemann we use samba
trifon i understand it. thank's.
CarlosRuiz ok, we can put a disclaimer saying that this approach is not recommended for confidential files :-)
CarlosRuiz BTW, I think people installing in production must stick with 314
CarlosRuiz and we must encourage people testing to test with trunk (or nightly build - Gavin is working on that)
hengsin yes, for better security and reliability, a repository like webdav, svn or cvs is better.
karsten-thiemann but you can limit access to the files on org level
karsten-thiemann because they are stored in a tree clientNo/orgNo/tableNo/recordNo
vpj-cd_ I am think to leave in prouction
vpj-cd_ with 315
hengsin karsten, issue is it is a share the user can access outside of adempiere.
vpj-cd_ but this depend that change we will make in future
CarlosRuiz karsten
hengsin victor, I'm still awaiting your feedback on the VInoutGen and VInvoiceGen patch
vpj-cd_ we have a issue
vpj-cd_ yes I Low I need test I hpe make today
CarlosRuiz I'm planning to integrate today the system config table proposed by armen
vpj-cd_ the main issue
vpj-cd_ we need some way to pach old version
vpj-cd_ and new version
vpj-cd_ exist pach aplly old version ,,but also we have new functionality
vpj-cd_ the main problem ss if we use any version ie 324
vpj-cd_ the pach are only en trunk
vpj-cd_ we need some way to pach ol version
vpj-cd_ without dowload trunk
vpj-cd_ download
vpj-cd_ ie if somebody think leave in production with 3.1.4
CarlosRuiz I think you can test with Garden data
CarlosRuiz in trunk
vpj-cd_ we solve pach in trunk to 3.1.5
croo_ [n=croo@89.19.95.43] ha entrado en la sala.
vpj-cd_ but we need version stables and version beta to new features
vpj-cd_ my question is what is best procedure to this ?
trifon this is very hard question.
CarlosRuiz 3.2 will be our first stable
CarlosRuiz 3.3 will be the trunk again open to new features
hengsin 3.2 will be tag and maintenance branch can be created if needed from the tag
croo_ i missed the beginning ... but can't the current version be frozen as a 3.1.5 branch where only patches are applied ... mean while the trunk remain a development... but the 3.1.5 will be the branc to become 3.2
croo_ at the point of tagging 3.2 - it will be merged back to trubnk to ensure the fixes are included
karsten-thiemann hmm - but if we do a feature freeze the activity will concentrate on bug fixing
croo_ yes
vpj-cd_ yes Carlos
CarlosRuiz and testing 3.1.5
croo_ but you hav eto draw a link some time
vpj-cd_ but ie now in 3.1.5 we set document validation
vpj-cd_ I think this to 3.3
vpj-cd_ also we set the pach from compiere260b
vpj-cd_ we need validate this change
CarlosRuiz don't get the point
vpj-cd_ the I think small change to have version stable
croo_ yeah 315 is the QA testing so patches can be applied ... just not huge chunks off new functionality!
vpj-cd_ 3.2 and great change after 3.3
CarlosRuiz for me 3.1.5 is a release candidate
croo_ yes
vpj-cd_ yes but if we set new things we need test
croo_ but against 315 we only apply patches to bugs
vpj-cd_ new thing always may be to disstable code
vpj-cd_ yes I am agree only paches to bugs
vpj-cd_ I agree
croo_ but meanwhile if need be major code changes can be happening in the trunk .. which remains the development version
hengsin yes, so after the freeze only bug fixes to trunk
croo_ no only bug fixes to the 315 branch
CarlosRuiz Colin, I think as Karsten, better to concentrate people on testing and bug fixing
croo_ 315 will become the 3.2 stable
hengsin colin, it is simpler to branch for major changes ...
croo_ well the point is developers can develop and testers and bug fixers can do their bit!
hengsin rather than the other way round
CarlosRuiz we don't have enough resources yet to keep eyes in trunk and branch
vpj-cd_ yes if we can fix only bug
karsten-thiemann right - developer should develop the bug fixes
croo_ so developers can if they want to ... continue in trunk while the masses continue testing and fixing 315
vpj-cd_ we can test 3.15 via QA cicle
vpj-cd_ and when we fishish we can release 3.2
croo_ yes THEN we must merge the fixes made to 315 back to the trunk!
CarlosRuiz Colin, I would prefer not to open a 315 branch, simply freeze the trunk
CarlosRuiz the reason is to concentrate efforts of community
croo_ well we needto keep a branch from all stable releases
hengsin yes, that will need more effort and we are a bit short on resources.
croo_ not everyone will always upgrade so it is possible we will need to make
croo_ patches to different stable versions
croo_ critcial patches that is not all
CarlosRuiz yes, the plan is to tag 3.1.5
CarlosRuiz continue bug fixing in trunk
vpj-cd_ ok the we release 3.1.5
CarlosRuiz when we consider stable (hopefully march 20) we release 3.2.0
hengsin and tag 3.2 when it is ready
croo_ but that doesn't allow us to keep a branch from the stable release
vpj-cd_ and freeze trunk
croo_ if we find a big bug in 3.4 we might want our 3.2 users to get it without upgrading
vpj-cd_ then we can make branch to 3.31
CarlosRuiz and we can open a branch for bug fixings on 3.2.0 to become 3.2.1 ?
croo_ so we need a branch for that line
hengsin after that trunk will be open for new development and maintenance branch can be created against the tag 3.2 version
CarlosRuiz that's good Colin, just I think that we don't have enough resources still to keep several versions :-(
vpj-cd_ and applay in branch new changes 3.3
karsten-thiemann colin we can do the branch right before the release of 3.2
vpj-cd_ this new functionality is rigth we can move to trunk
hengsin colin, in svn, tag is more or less equivalent to branch
croo_ but in the scenario where developers want to continue development while QA is happening then they can't make any changes to the trubk
croo_ because only fixes are allowed
vpj-cd_ yes I think trunk is to stable version
vpj-cd_ and branch to unstable version
karsten-thiemann but they can do it local and post it later
vpj-cd_ so, any dowload trunk the know this stable and have the last pach
CarlosRuiz and we want to ENCOURAGE developers to focus on fixing
vpj-cd_ patch
hengsin yes, colin, that is the policy. Of course, if the developer can't wait, he can create a branch to do new developmnet
karsten-thiemann right CarlosRuiz
hengsin so there is actually no real blocking there.
vpj-cd_ if any want add new fuctionality then set in branch 3.3.1
croo_ ok today we do but it won't always be that way ... typically QA cycle is long ... so development work would continue on a project
vpj-cd_ until I can set libero in 3.3.1
croo_ this way no development can happen until QA is over
CarlosRuiz agree Colin, in future with more resources we can go your path
vpj-cd_ is easy trunk is to stable version and branch to change that do not bug
croo_ why change later ... why not do it one way today for good?
trifon even with more resources branch is big pain. i think that commercial companies support branches because they demand money from users for upgrades.
vpj-cd_ when a change in brannc is rigth the we only need merge the trunk with branch
hengsin colin, we talking the same thing in different term, it is either new development in trunk or new development in branch ... it is actually the same thing, parallel work can always happen.
trifon in our case we migrate for free.
vpj-cd_ to this change
CarlosRuiz yes, I mean resources in people and maybe resources in $$
vpj-cd_ but this only if the change is validate and rigth
croo_ companies do like to upgrade to every new release ... it's like a whol enew implementation
croo_ so expect many to stay on a stable release!
trifon yes i agree with Colin.
trifon in this case if companies want to stick with some version and reguire fix for it. they should support it.
CarlosRuiz agree
croo_ well they will probably be our customers so they will expect us to support it
trifon i think that we need to have some policy we can't do everything
CarlosRuiz another approach is not release new versions on 3.2, but release patches.jar
croo_ when I say we I mean as indivdual service providers
vpj-cd_ yes also in bisness aplication is best have version 3,2 and
vpj-cd_ maintenaice to 6 mothod
vpj-cd_ apply change
trifon yes :) this is true. Colin
croo_ no but maintaining the stable releases means only critical bugs
vpj-cd_ in ERP software the industry do not release each mouthd
trifon one general question how long do you think that one version must be supported?
CarlosRuiz Heng Sin told me something about the naming convention, even/odd numbers to mark unstable/stable is confuse
croo_ I think companes epect to go years without upgrade .... the attitude is if it works don't touch it!
trifon SAP support it 8 years
CarlosRuiz I think we can support while there are contributors to support it
CarlosRuiz contributors in people or money
trifon yes. Colin is right. some ERP systems work for 20 years.
karsten-thiemann main problem is that we'll have lots of versions
croo_ well again when I say we I mean if I have a customer who won't move from 3.2 it's up to me to support him and maintain it ...
karsten-thiemann because we need lots of enhancements
hengsin this is opensource project so I guess as long as they are people that want to maintain it ...
croo_ if I am doing it should have a place in the tree to store those changes so others can use them too
vpj-cd_ mmm MFGPRO use 2 year to support old version
CarlosRuiz yes Colin, but with a larger base costs will be lower
CarlosRuiz cost of maintaining a version of just one customer are big
vpj-cd_ but they release new version each 6 mouthd
vpj-cd_ in a year only set pach
vpj-cd_ and new version after yest have great change
trifon from my experience cost to maintain old version if bigger than cost to migrate to last.
trifon because i do big changes.
vpj-cd_ I think that ADempiere we can support 1 year
croo_ cost of implementation is mostly the services ... the software is very cheap .. here anyway ... the per seat might be 1500-1700 but you can expect to pay at least 1000 per day for the consultant
trifon but if changes are small then cost should be smaller
CarlosRuiz it depends on how the customizations were done, Trifon
hengsin eventually you have to upgrade because it will reach a point no one can maintain it :)
CarlosRuiz I mean -> if bad customizations - migration is a big project
karsten-thiemann right
vpj-cd_ yes I think that 2pack can are util
vpj-cd_ as tools to migration
karsten-thiemann but at least now the changes in db and code are transparent - so migration will be easier
CarlosRuiz anyway, this is open source, if a customer want or must stick to a version, they can maintain his own release
vpj-cd_ ok then we freeze the trunk?
vpj-cd_ and only apply pach to bug?
CarlosRuiz yes, tomorrow
karsten-thiemann ok
croo_ carlos, you will tell you customer ... if you don't upgrade you can support it yourself?
CarlosRuiz Victor, are you going to include vertical tabs and master/detail ?
vpj-cd_ I think that this is change but need are test
AFalcone [n=AFalcone@190.49.85.1] ha entrado en la sala.
trifon i think that feature like master/detail will destabilize trunk
vpj-cd_ the I think we can think nexet release
AFalcone: Hi
croo_ hi Alejandro
trifon Hi Alejandro
CarlosRuiz sorry about that, it's a BIG ENHANCEMENT
vpj-cd_ or we can release this as beta functionality
vpj-cd_ to ppl test
croo_ I never liked that ... no ned for beta in a stable/production release
vpj-cd_ yes I know we can set fi you want to as beta
croo_ people think they something they don't
trifon One open source project: Open for business do not have branches
CarlosRuiz Victor, changes are isolated? I mean, if you include the changes it won't stabilize current "only-master" windows?
croo_ yes a good project but very much a "hackers" project
trifon :)
croo_ I think something core like this we need to project stability and long term support (not 8 years now)
croo_ but you guys are the CC ... I've said my bit ... you can decide
trifon my opinion is more close to Colin. we need branches but support for them should not be main task of Adempiere team
vpj-cd_ my current issue is when i set collapse panel this donot take of display rule
CarlosRuiz it's not a decision of CC, it's a decision of $$
vpj-cd_ I need add this into the APanel
CarlosRuiz maybe a BDC decision
croo_ trifon, precisely ... those with customers can maintain
croo_ it might be us it might not
vpj-cd_ now APanel and Gridcontroll use only CPanel
CarlosRuiz ok Victor, let that enhancement for 3.3
croo_ the point is there is a place in the svn tree to maintain any patches that are made
CarlosRuiz yes Colin
trifon but i would like very much we can assure our customers that their versions will be supported in next 8-20 years.
trifon they will be more confident
CarlosRuiz from my point of view, all depends on support contracts and money
vpj-cd_ I think if we only have a stable version and developer version
vpj-cd_ is fine in opensoure software
trifon but how we could do it. free upgrade to new stable versions or big list with old branches.we should decide.
vpj-cd_ we apply patch to stable version and set new things in development version
croo_ but victor, say we have just two versions
croo_ you have a customer on 3.2 and so do I
CarlosRuiz Colin, is just a matter of opening a new branch or tag
croo_ the main version has moved to 3.6
CarlosRuiz and let someone to maintain it
vpj-cd_ when developmnet version get the stablish then we canr elease as stable
croo_ then if I find a serious bug I cannot chare it with you!
vpj-cd_ the 3.6 is stable version
CarlosRuiz Colin, is just a matter of opening a new branch or tag
vpj-cd_ 3.2 sorry
vpj-cd_ and 3.3 is to development version
vpj-cd_ 3.2 this to trunk here we apply pach only
croo_ yeah no arguments there
CarlosRuiz we're going to have a tagged release 3.2
CarlosRuiz it's a business decision if you want a 3.2.1, 3.2.2 and so on
vpj-cd_ and 3.3 is to new development and this .ie libero , new improve in gui
CarlosRuiz and we allow in svn to set up those new tags
vpj-cd_ then if you need leave in production you use last 3.2.x
vpj-cd_ and you can are sure have not problem
vpj-cd_ if you want test and dev we can use branch 3.3
CarlosRuiz ok, can we change theme?
vpj-cd_ ok
croo_ ok
CarlosRuiz given last movements in trunk
CarlosRuiz and overload of CC members
vpj-cd_ yes , the the trunk is freez ?
hengsin colin, once a version is tag, branch can be created from the tag anytime after that.
CarlosRuiz maybe we need a way to ensure that every change in trunk is reviewed but one CC member
CarlosRuiz at least by one CC member
CarlosRuiz more eyes better
CarlosRuiz currently this is not organized, I don't know if you're reviewing, and you don't know if I'm reviewing
vpj-cd_ mm ok we need a procedure easy
vpj-cd_ a devloper set new pach
vpj-cd_ the developer have to test new pach this should work
CarlosRuiz and this can be combined with the issue "ease people to contribute"
vpj-cd_ whet this pach is test to another development or tester
vpj-cd_ the this is integrate in trunk
CarlosRuiz people are not testing patches Victor
vpj-cd_ this is way use Low I think is good way
croo_ if all changes are due to a bug (or rfe) then the first person to take on the review can asign the bug/rfe in sf .. if someone else is already assign they must contact that person to co-ordinate
CarlosRuiz Heng Sin put lots of patches, not tested until integrated into trunk
AFalcone: if you use the Assign To when you submit a patch?
CarlosRuiz VCreateFrom is in trunk and is not tested yet
vpj-cd_ the ppl have not pach because now is very complex
vpj-cd_ install
vpj-cd_ a pach
trifon wait. how in trunk we get things which are not tested?
vpj-cd_ if they have way easy to apply pach then we come to test
vpj-cd_ we have more tester
hengsin it depends on the audience
CarlosRuiz Trifon, just basic testings
trifon Carlos with VCreateFrom is in trunk and is not tested yet
CarlosRuiz thorough testings are almost impossible
vpj-cd_ the problem now is only developer can test nrepach bug
CarlosRuiz I fully tested 2pack with oracle, and failed on postgres to Victor
vpj-cd_ pach bug
vpj-cd_ yes we need test in 2 database
CarlosRuiz developer ensure the thing compiles and make some basic testings
vpj-cd_ I can tester to postgres
croo_ perhaps on confirmation in sf we should list the test env it worked with
trifon so everything which go into trunk is started and working in the main use case.
vpj-cd_ but we need way easy to any functional consulting can download pach apply and tester
vpj-cd_ very ppl have wait new release to test your bugs
hengsin another issue is some bug reported without a detail use case making fixing and testing difficult to begin with
vpj-cd_ if a functional consultang need test now a pach
vpj-cd_ he need download trunk
vpj-cd_ and complie
CarlosRuiz ok, we have combined things
CarlosRuiz 1 - ease people to contribute
CarlosRuiz 2 - ensure that at least one CC member review the contribution
CarlosRuiz 3 - testing (too much for CC members)
vpj-cd_ very consultan functional or ordinal user
vpj-cd_ have knwnolodgnet to compile version
croo_ i think the CC should just give a basic code review ... did they create a bug report, did they follow any coding standards etc
croo_ and others can test functionality
vpj-cd_ I propuse use 2pack to pach the code
croo_ unless it's a technical issue
vpj-cd_ we cah join the pach to week and creat a 2pack with al patch
CarlosRuiz normally CC member have the skill to know if a patch broke other parts
vpj-cd_ all pach
croo_ i think 2pack to patch stable releases makes sense but it sounds like overkill in a development environment
CarlosRuiz the problem I see is that some contributions can pass without revision
vpj-cd_ I think new contribution
CarlosRuiz ok, how can we ensure that at least one CC member review the contribution ??
vpj-cd_ we need that new contribution to branch
CarlosRuiz Victor, that's against "1 - ease people to contribute"
vpj-cd_ yes but this same question have JJ
vpj-cd_ how we accept new contribution without distable the core
vpj-cd_ ?
hengsin carlos, for "ease people to contribute", we are a bit short on the documentation part.
CarlosRuiz linux way? as Colin proposed?
CarlosRuiz we have a stable branch, and don't worry on unstabilizing core
trifon ok, how can we ensure that at least one CC member review the contribution ??
trifon Carlos here i see another point.
trifon As we all give our free time we can push each other and give each other more work.
CarlosRuiz just taking an idea, not proposing that
vpj-cd_ I think each version should take you time to mature
trifon ok.
vpj-cd_ if one contribution is validate to 2 member the CC
vpj-cd_ then this is set into stable version
Jaztek ha salido de la sala (quit: "Leaving").
vpj-cd_ antheother way we always have unstable version
vpj-cd_ or a version that need test a contribution
CarlosRuiz Victor, this is unrealistic, I would prefer to have the tranquility that at least one reviewed
CarlosRuiz I'll review all that I can
trifon another point here is that this who review is in some extent responsible and users will most probably aks him in case of problems.
vpj-cd_ ok the i have a example
trifon I also review each commit.
CarlosRuiz ok, one question, who reviewed armen contribution on inventory?
vpj-cd_ we should set libero now ?
vpj-cd_ this is a contribution we can set libero?
CarlosRuiz armen contribution is in trunk, someone reviewed?
trifon but i must confes that it is general review and not sure in what extent i can be sure that commted code will not destabilize core.
CarlosRuiz me too, Trifon
vpj-cd_ is practical
hengsin functional patch can't be review by just looking at the code
CarlosRuiz and pretend that CC can review every contribution to keep core stabilized is not achievable
vpj-cd_ when you detect a bug you set
hengsin it need a use case or test case to verify.
vpj-cd_ and generaly I fixed too
vpj-cd_ I think the filter to test the contribution need QA cicle
hengsin we need to first determine the nature of the patch
croo_ i think CC's review should be to ensure standards ... actual testing must envolve the whoel project otherwise there is no point to having the bazaar!?
vpj-cd_ exist some modet to follow
hengsin for technical patch, a review by one of the cc memeber could be sufficient, for functional patch, it must have a test case/use case for verification.
vpj-cd_ ?
vpj-cd_ ubuntu ie
vpj-cd_ somebody some model?
croo_ i think commit to a development trunk is not so strict as a commit to the QA branch
croo_ many can commit to trubnk ... but few to the QA
CarlosRuiz Colin, we make the task as a contribution, but not as our core function
croo_ yes ... that why I say only a review to ensure standards are maintained
croo_ it's not reasonable to expect you the CC to test every change made!
croo_ it would take you all your time
CarlosRuiz exactly, that's the point
hengsin sure we can't do that :)
croo_ so you check that croo followed the rules when commiting and commented his code or whatever the rules you have
CarlosRuiz we can't ensure to keep stabilized trunk, it's a good effort, but not ensured
croo_ but it's up to the bazaar to test
CarlosRuiz ok, we're testing a little as first part of the bazaar :-)
croo_ yes .. but I think we a QA branch or tag or whatever it will be .... all patches should be submitted as patches and tested by someone before commited
croo_ and by someone I don't mean the CC ... but just a second person
croo_ not just the person making the submission
CarlosRuiz and into development trunk? also via patches?
croo_ i think it would be a good idea too to encourage people to use the severity field on bug/rfe & patches
CarlosRuiz Colin
CarlosRuiz the problem is that people is not reviewing patches
CarlosRuiz even is difficult currently that people test trunk
croo_ no - i think but you are the CC - but I think the development trunk should be moe relaxed
CarlosRuiz we don't have a large community yet
croo_ well there are a lot of people in the forums & irc that I think would help but are not sure how
croo_ perhaps I should rescue my QA docs for the 315 release with info on how people can test
CarlosRuiz nightly build is a must !
croo_ but you (the CC) must decide how you wish the process to be
croo_ yes joel offer to automate didn't he ... I think that would be very useful
vpj-cd_ Carlos and if we solve this issue
vpj-cd_ to ppl can be test easy?
croo_ yes
hengsin nightly build will help alot.
croo_ means more eyes
vpj-cd_ we can kit pach release via 2pack
vpj-cd_ the only need set import and ready :-)
vpj-cd_ the problem the nightly build is the time to download
CarlosRuiz 2pack is not for that purpose, maybe patches.jar, but is more difficult
vpj-cd_ we need a simple zip smaill
vpj-cd_ becasue not
CarlosRuiz currently with 2pack you can't install a jar
hengsin most of the time, just replace Adempiere.jar and run_setup again will work.
vpj-cd_ 2pack is exelet to have AD change
CarlosRuiz but nightly build is changes to lot of core classe
vpj-cd_ and we can set a script have take the current jar complie the class pach
vpj-cd_ and integrate the current jar
croo_ does it export AD changes? because I have some I wanted to submit but I dont now how - or at least the best way ... perhaps at the end someone could advise me?
trifon yes 2pack export AD changes.
vpj-cd_ Carlos but 2pack ca execute a script
vpj-cd_ one simple script take the current jar and compile source java class this are set in build directory
vpj-cd_ and we use jar to replace the old java.class into Adempiere.jar
vpj-cd_ and ready
CarlosRuiz Victor, but who is going to construct installer every night?
vpj-cd_ the user have way fast to apply pach
vpj-cd_ ie example
vpj-cd_ we can new cahnge in AD
vpj-cd_ change in AD
CarlosRuiz Victor, do we have a resource to do that?
vpj-cd_ and have 10 java source code to pach
hengsin we can have nightly build that output 2 types of file, the full tarball and also the jars file. For user that have existing installation, most of the time just replacing the jar file in lib folder and run setup again will work.
vpj-cd_ ent via ant we build new surce code usa binary jar
vpj-cd_ we get new class complie and use jar to add in patch.jar
vpj-cd_ and ready
vpj-cd_ :-)
vpj-cd_ the problem with opetion Low is need very time to matenaice
hengsin ??
CarlosRuiz Victor, do we have a resource to implement your idea?
vpj-cd_ we need set in 2pack only a button
vpj-cd_ you and create new pach.zip
vpj-cd_ we et in sf and ready
vpj-cd_ yes Carlos I can commit with one resurece of techinical consultan
CarlosRuiz ok, we deviate again from the theme
CarlosRuiz can we ensure that at least one CC member review every commit?
CarlosRuiz ready Victor -> please implement the idea, good!
vpj-cd_ but I need exist commit to to support and help to test
vpj-cd_ this oprion
vpj-cd_ option
vpj-cd_ ok
CarlosRuiz again, how can we ensure that at least one CC member review every commit?
CarlosRuiz can we divide the project in modules and assign modules to each CC member (we're 5)
vpj-cd_ the I have to Fernando to begin work to make 2pack to system to pach adempiere
CarlosRuiz ok
vpj-cd_ Carlos I think is easy
karsten-thiemann have more CC members ;)
trifon Carlos, i'm not sure if we can do this. load of every CC is different.
trifon we have developers
vpj-cd_ we can divide to bisness process
vpj-cd_ but QA should wiil make to partner to adempiere
trifon last weeks my load is big and my commits are few.
vpj-cd_ I will set 3 my consultan to test somes process
croo_ i suggested before that the CC split the project and each takes control of each section...
croo_ then that person is allow delegate to a perosn chosen by them
croo_ eventually as the project grows perhaps you might even delegate the power to delegate
vpj-cd_ I think that Vince, Jhoe, Croo , Ramiro can set people to make QA cicle
vpj-cd_ the comunnity current have way to validate QA cicle
vpj-cd_ very ppl do not know the system
vpj-cd_ and we have not a manual they can follow
karsten-thiemann perhaps it would be ok also if 2 other developers are checking the code
vpj-cd_ the the QA cicle need are make to parnter
croo_ well people not knowing the system can be good testers too .. they do things wrong and break it ... which is what we want to find out!
CarlosRuiz testing is different from CC review
vpj-cd_ yes croo but the real test come the partner that know adempiere
croo_ yes
CarlosRuiz I'm trying to know if we can ensure the CC review at first
vpj-cd_ ie I know the issue about the project
CarlosRuiz for testing we need a different approach
vpj-cd_ beacuse I need to implement
trifon i have the feeling that we started thinking like commercial team. we want to split and delegate but we do not have the right to push other CC memeber to do it.
croo_ yes victor, I'm just saying that it takes all sorts and everyone can help
trifon i think that we must push community to test.
CarlosRuiz ok, I don't want (and I can't) to delegate :-) but again, how can we ensure that at least one CC member review every commit?
CarlosRuiz any ideas?
vpj-cd_ yes trifon we can make this but they need a guide
vpj-cd_ a manul
CarlosRuiz CC review, not testing
vpj-cd_ we donot have manual
AFalcone: i think that Carlos is talking about other issue. Not for testing
trifon we can announce that we will releaee stable version in 3 weeks and it depends on thier effort to report and support stabilization.
croo_ know you can delegate though if someone volunteers ... now only the CC has rights correct... but if you trust someone you can they delegate if they agree
AFalcone: testing can be done for the community
AFalcone: but review code not
croo_ i'm not suggested we allocate work to people who don't ask
hengsin carlos, only way is make patches a must, and auto assign to CC by category
CarlosRuiz ufff, category listing is too big
vpj-cd_ Carlos the code is review to CC
croo_ yes it is way to big
CarlosRuiz ok, the question is again: how can we ensure that at least one CC member review every commit?
vpj-cd_ but the version and QA is focus to bisness process
CarlosRuiz I would prefer to have a mechanism to ensure that
croo_ needs only small number of things ... UI persistence ... maybe the major ERP modules
CarlosRuiz if not, what's the purpose of being CC member?
trifon every commit must have number of patch. patch must be assigned to CC memeber.
vpj-cd_ Carlos the rule is basicly if 2 member review code then this is apply
CarlosRuiz anybody can review commits
croo_ all submissions of QA patchs should be submitted as a patch first.
vpj-cd_ we now use this Low, Carlos and me
CarlosRuiz anybody can be included in e-mail list to review commits
vpj-cd_ and is easy
croo_ only when okayed can it be commited
AFalcone: yes Victor, the problem is ensure that
CarlosRuiz but as CC members we must be sure that someone of CC review
croo_ if someone commits with follow rules you remove their commit access
CarlosRuiz and as you said, I can't delegate if you don't volunteer
trifon in this case CC memebr must write a post in tracker that it is reviewed
vpj-cd_ Carlos the pach need will be test 2 member CC
vpj-cd_ the is set in trunk
vpj-cd_ now we make this is work fine
CarlosRuiz Victor, patches are even more tricky to review than commiots
CarlosRuiz the problem is also time as Trifon pointed
croo_ i think one CC is enough unless a CC is submitting?
CarlosRuiz but I must ask, what's the purpose of being CC member then?
trifon to be CC memebr?? not sure...
croo_ to ensure standards are maintained? and hold ultimate control... only you can decied you can commit
croo_ only CC can decide to un commit
karsten-thiemann to make really shure that CC has reviewed the code just remove commit rights from all others
trifon :)
croo_ sorry that was only CC decide who can commit
vpj-cd_ ye sCarlos the problem is time
croo_ you decide the coding standards
vpj-cd_ but not exist another way
croo_ and framework changes to include
vpj-cd_ if we want have a trunk unbreack
CarlosRuiz ok, I asked -> does someone reviewed armen contribution on inventory that is into trunk currently?
CarlosRuiz nobody answered, so I think that nobody reviewed
trifon do oyu know feature number?
CarlosRuiz at least code review, not testings
trifon i reviwed all emails with commits, but they are so many theat i do not rememebr this one
karsten-thiemann ok but last weeks showed that most contributions are ok - at least they don't crash the system
karsten-thiemann and if they do they are corrected fast
AFalcone: ok karsten, but not QA cycle present there
karsten-thiemann so perhaps there is a way (with svn) to restrict committing to a certain branch
croo_ i think in a development release we can be more flexible ... you can even review after the fact
croo_ but QA id different
CarlosRuiz yes Colin, but again, how can we ensure?
CarlosRuiz what if everybody trust that someone else in CC reviewed?
karsten-thiemann ok but why not remove the commit rights for the QA time?
karsten-thiemann and go the SF patch way
CarlosRuiz I don't have problems, people are committing good contributions, and mostly in the right way
cjerome [n=chatzill@ADSL-TPLUS-100-42.telecomplus.net] ha entrado en la sala.
CarlosRuiz just want to know if every commit is being reviewed by CC
croo_ well we said all submission to a QA release as a patch .... and when a second person has tested the patch it can be commited
karsten-thiemann and only CC members can contribute the patches
vpj-cd_ ok is easy we need confirm the 2 CC member
vpj-cd_ via email
croo_ i don't think so
CarlosRuiz not agree
vpj-cd_ :-)
CarlosRuiz we don't have ghe time
croo_ I submitte all my mods as patches but they sat for months until hengsin say just commit them
CarlosRuiz Victor, I haven't achieved to have at least one review, and you want 2 :-)
croo_ I was surprised I could commit to be honest ... I didn't know it was possible!
hengsin carlos, don't that is realistic now -> every commit is being reviewed by CC
CarlosRuiz and we don't have the time to download patch, install, review and commit
croo_ no no by two I mean two people
vpj-cd_ ok the the partner need make your work :-)
croo_ ot just the CC ... you guys can give it te quick once over to ensure standards
croo_ but two people must test functionalty
vpj-cd_ we can set if a partner want are into adempierebussnes they need give time to test pach :-)
AFalcone: the review must be done for who know the code
hengsin I guess we can only review from time to time and if things break, regroup and fixed it :)
croo_ if good functioanlity is sitting in the patch area people will review it to ensure it goes into the code!
karsten-thiemann now for the development trunk it is more like commit it and it will be reviewed later by the fact that it is there
vpj-cd_ we have in community do not only development
CarlosRuiz so I ask again: what's the meaning of being a CC member?
croo_ but I think core stuff in the framework like persistence ... well only a few can really test that kind of stuff
vpj-cd_ Ramiro,Alejandro,Croo, Idalica , Vince , Moyses , etc
hengsin victor, good suggestion, partner should contribute
vpj-cd_ the can hep to validate
CarlosRuiz everybody (not just CC member) can subscribe to the commit list and review
vpj-cd_ the partner need have commit with project
vpj-cd_ they have to give time
AFalcone: hengsin, but to review you need to know the code
croo_ define standards, control commit list (who), rollback a commit, revoke commit access
croo_ lots of things
vpj-cd_ no Alejandro si en el description bug set wha is fixe de bug
vpj-cd_ the consultant only need review that pach make to tell
CarlosRuiz yes Colin, but I can't ensure that someone is reviewing a single commit
vpj-cd_ to review code need make to CC member
vpj-cd_ this is fact
AFalcone: ok, but this is a functionality testing, not code review Victo
CarlosRuiz my conclusion can be: CC members does the best effort to review every commit according to their time
vpj-cd_ we can not to review to ppl that we is into project
vpj-cd_ we need all the help to review code
croo_ carlos, you can ensure the commit comment include the bug# or patch# and ensure that it was tested by someone other than the submiter
CarlosRuiz Colin, I can't ensure that every commit is being reviewed
croo_ carlos, if someone commits without that (repeatedly) you can revoke their access
CarlosRuiz I try to do the best effort, and all CC members too, but I can't ensure ALL
vpj-cd_ now only Carlos, Low, Teo and me review code
vpj-cd_ but we need help the all CC commite
vpj-cd_ this the proce to are into CC Member :-)
vpj-cd_ proe = price
CarlosRuiz that's my only point, but I think that we (CC members) are doing a good work
AFalcone: yes, I'm sure Carlos!
croo_ then you need more people ... and to delegate to them ... this is why I suggest the CC split the project somehow ... you say I'll be responsible for PO then you get others YOU trust to help you do that
vpj-cd_ yes Carlos
CarlosRuiz just that as in the best bazaar way, we're subject to volunteer contribution
CarlosRuiz it's ok for me, CC members are filtering a lot of possible problems, and even we're trying to make first testing
croo_ if there are not enough people to test then simply do not include the functionality .. I'm sure more oeple will volunteer if a fix or new functionality they want is ready but needs testing!
AFalcone: I agree Colin
hengsin For functional patch, I think it is important to set the rule that all functional patches must have a valid user case to verify.
croo_ hengsin, yes I totally agree
croo_ it should be a confirmed bug report first ...
AFalcone: yes hengsin, very good point
vpj-cd_ yes Low shiping is a good example
croo_ and the bug report should include a use case ... if not then we set the severity to the lowest level
trifon i need to go now.
croo_ maybe if someone has time they will do some more work with it
hengsin bye trifon/
trifon bye
AFalcone: bye trifon
croo_ again, if it's important people WILL do it
CarlosRuiz bye, thanks
trifon ha salido de la sala (quit: "Chatzilla 0.9.75 [Firefox 1.5.0.10/2007021601]").
croo_ bye trifon
CarlosRuiz Colin, why haven't you contributed your patches? you said before that you have some things to contribute but don't know how to export it
croo_ i did
croo_ oh that
croo_ thaht's the fix you end me the link for
CarlosRuiz I want to balance that
croo_ to restrict info quer access
CarlosRuiz I want to have a review for every commit, but also want to easy the way to contribute
karsten-thiemann thats the problem...
croo_ I have it working here ... but I have added new fields to the AD_ROLE and Role window and I don't know how do I get them out of the AD
CarlosRuiz yes, the problem there is that you don't know how to contribute dictionary changes
croo_ yes
CarlosRuiz yes, it must be the problem of lots of people
karsten-thiemann oh - for that we need some kind of id pool too
croo_ I only mad little/simple fixes up unyil now :)
hengsin carlos, like I mention earlier, documentation is lacking
CarlosRuiz yep
karsten-thiemann so that you can say I need Ids 50100-50105 of AD_Message or something like that and everybody knows
croo_ well I have all the coding done and it works .... just the AD stuff ...
CarlosRuiz and the scripts for insert dictionary are tricky
hengsin many of the info in wiki is not correct or outdated, I try to fixed some but it is just too time consuming to go through that.
croo_ well I just did the scripts for the new fields in AD_ROLE by hand in the end
karsten-thiemann we could provide some standard scripts for inserting field, column, element, message
CarlosRuiz or even a database procedure to do that
karsten-thiemann because those standard changes are not too complicated to understand
CarlosRuiz or have a tool to construct the insert's
vpj-cd_ yes but all this issue are solve with 2pack
CarlosRuiz partly
AFalcone: yes, the problem is the ID range that you will use
karsten-thiemann right but not for insertions into trunk
vpj-cd_ a normal user do not set the sql to fixed the AD
croo_ I've just been loking at 2pack as we chatted ... I can
vpj-cd_ the do not know
croo_ sorry ... pressed return by mistake
CarlosRuiz 2pack is not a migration tool
vpj-cd_ I tlak with Frnando
CarlosRuiz for migration we need the scripts
CarlosRuiz but 2pack can be the way to contribute easily
vpj-cd_ now is dedicate to set 2pack support to export and import and apply pach
vpj-cd_ Carlos because 2pach do not can migration tools
croo_ when I add fields to the table I need to also System_Elements and when I added the field to the window I added a Field Group .. does 2Pack automatically ahndle all those too?
vpj-cd_ we should user ADXML to migrate
vpj-cd_ and not SQL
CarlosRuiz not in the way I like Colin
vpj-cd_ I need migrate also postgresql installation
CarlosRuiz you can't export just a column, you need to export the entire table
croo_ ok just wondering
vpj-cd_ the tier to solve this issue is ADXML
croo_ well I think we can seperate optional changes from core changes
vpj-cd_ we can have a 2pack file to migrate 313 to 314
croo_ oh yeah.. that's sounds neat!
vpj-cd_ the pleople only need add the file in 2pack and import an your migration is ready
CarlosRuiz I don't trust in 2pack for such task
vpj-cd_ Carlos you are worj to inprove the ddl use via class java
croo_ well I made a test 2pack with my changes and I can't even find the exported file :D
AFalcone: hehe
CarlosRuiz I don't trust in 2pack for such task - YET - but can be improved obviously
croo_ bt it certainly looks promising
vpj-cd_ we can focus force in have 2pack ready to export and import and migrate tools
croo_ but I was hoping to get this mod in for the 3.1.5 ... it's small and pretty useful I think
croo_ so I've spent all day doing it but now I can't get it in .... ;(
vpj-cd_ this is original idea the marco
CarlosRuiz the question is: how can Colin make a development and then contribute back using system sequences?
vpj-cd_ and how maked 2pack
vpj-cd_ ok I ame this with libero
CarlosRuiz or just let Colin export the patches, and let CC members to organize the sequences?
vpj-cd_ ame =make
vpj-cd_ somebody can have charge the control sequence in AD
vpj-cd_ ie I have a pach or modify AD
CarlosRuiz it's a loop :-(
vpj-cd_ I export my 2pack file with Customer user
CarlosRuiz it's too late in Malaysia, we can think on how to ease contributions on AD and talk about that later (or in next CC)
vpj-cd_ somebody import
karsten-thiemann why not just use a simple wiki page for that - id reservation
vpj-cd_ and export with sequence rigth
vpj-cd_ and this oficial 2pack file to all comunnity
croo_ I can export the relevant data from the AD_TABLE and AD_WINDOW etc but I'd probably be missing some related data like the "Field Groups" etc
CarlosRuiz and the point is that 2pack don't export single columns
vpj-cd_ ye you export the people to charge
CarlosRuiz or single fields
vpj-cd_ import in your oficial seed
vpj-cd_ and export with AD sequences
vpj-cd_ and this new 2pack file is oficial
croo_ yeah when I started I had thought that AD2XML was what I used
croo_ but it seems not
vpj-cd_ with sequences reserverd
karsten-thiemann just a small question - deathmeat has developed a jasper integration to use jasper as print engine and he wants to contribute it - it is tested and works and won't affect normal admpiere behaviour
CarlosRuiz ok, can we conclude this meeting?
CarlosRuiz pending task -> we must think on how to ease contributions on AD using system sequences
karsten-thiemann shall he just commit it?
CarlosRuiz how is that Karsten? to use jasper as print engine ??
karsten-thiemann you can define a jasper template for each document
karsten-thiemann if you do so and press print button it will open jasper viewer to print
croo_ well I will submits an RFE and attach the file changes and a description of what I did in the AD ... that's about all I think I can do for now
karsten-thiemann if you dont have a jasper template defined it works just like it does now
croo_ yeah that sounds neat!
hengsin karsten, by document type ?
croo_ he has been working hard (along with karsten) on getting it ready for 315 .. it seems a pity not to squeeze it in
CarlosRuiz via definition of process (report) ? using same field than jasper integration?
karsten-thiemann wait a second
DeathMeat [n=DeathMea@adempiere.reisinger.de] ha entrado en la sala.
DeathMeat hi
vpj-cd_ karsten what is the problem?
CarlosRuiz Hi Johannes
karsten-thiemann I just phoned DeathMeat
karsten-thiemann to join us
DeathMeat on the phone
vpj-cd_ what you want to do?
DeathMeat ;)
CarlosRuiz I was asking about your "jasper as print engine for documents"
DeathMeat ok, my integration of JP includes the following functionality
CarlosRuiz is it integarted via definition of process (report) ? using same field than jasper integration?
DeathMeat on a document type level you define (beside the normal printformat) your jasper process
DeathMeat if jasper process is set to a valid one and you press the print button in say SO you get the jasper viewer with the correct report
DeathMeat i extended our jasper viewer classes to make the jasper report viewer export able as well as email functionality (which we already have) is avilable in JP viewer
DeathMeat if you do not define a jasper process on a per document type level the standard printengine is used
DeathMeat i also extended the reportstarter class to be highly compatible with subreports of all kinds
CarlosRuiz then you added a jasperprocess in C_DocType ?
DeathMeat i added a table validation dropdown based on AD_Process in C_DocType table and window
DeathMeat right
DeathMeat the jasperreport name is still defined on a per process level
vpj-cd_ waw
CarlosRuiz why don't use the same jasperreport column in ad_process?
vpj-cd_ DeathMeat
DeathMeat you even have the ability to pass print_discount to the jasper engine
vpj-cd_ you have special bogject in jasper report
vpj-cd_ to aempiere
vpj-cd_ adempiere
DeathMeat SO report is done for demo purposes
vpj-cd_ the main problem to use jasper is
vpj-cd_ you need build all the subselet to get the real data
vpj-cd_ in adempiere is more easy
DeathMeat right, but adempiere engine is far less flexible
vpj-cd_ yes
DeathMeat no charts, no barcodes, no graphs, bad layouting
vpj-cd_ we will be 2 features
vpj-cd_ :-)
karsten-thiemann and its just additional to the existing engine
vpj-cd_ varcode is support current
vpj-cd_ :-)
DeathMeat dozens of reasons why a _opitional_ JR integration is good
DeathMeat oh ok
DeathMeat my bad
DeathMeat ;)
vpj-cd_ barcode is support
vpj-cd_ but exist way to build new componet into jasper to support adempiere
vpj-cd_ TableDir
DeathMeat so, the normal user will never recognize JR integration in this way
vpj-cd_ Table
CarlosRuiz I agree to give more options to implementors, implementors now can choose how to set up the reports, it's ok
karsten-thiemann and it's easy to share reports
CarlosRuiz just I have a question
DeathMeat as long as they don't explicitely define a Jasper Process behind their say Sales Order document txype
CarlosRuiz why to create a new column, why don't use the same jasperreport column in ad_process?
vpj-cd_ the problem in the report is time you use to buil a report
DeathMeat because you need to associate the document type (say a binding proposal or a non binding one) to a jasper process
DeathMeat the process itself defines what report file to use
vpj-cd_ in engine of adempiere is easy because you do need know very sql
DeathMeat but you cannot distinguish on a process level for what printout it is used
vpj-cd_ but jasper is easy to set in griid the compones
DeathMeat yeah right victor, you need to have SQL knowledge
CarlosRuiz let me understand, /me revieweing C_Doctype ...
CarlosRuiz or set the proper views
DeathMeat once the queries are written in the jasper report data query window you can easly drop fields in and out to your report
vpj-cd_ then in future if we ad componet to jasper
vpj-cd_ as tabledir
vpj-cd_ location
DeathMeat right
karsten-thiemann victor - the good thing is that somebody who knows sql and can do nice reports can share them with the community
vpj-cd_ Table
CarlosRuiz you can do that without the column in C_DocType
karsten-thiemann because you just have to drop the file
vpj-cd_ etc
vpj-cd_ will could very nice only take and set the componets to adempiere
vpj-cd_ in jaset without need the build sql :-)
CarlosRuiz not yet ... still reviewing C_DocType
AFalcone: Yes Vicor, but the JR integration is done!
karsten-thiemann I must confess that I have committed the attachment store patch only being reviewed by johannes - it was some kind of pair programming...
karsten-thiemann different story :)
vpj-cd_ yes Alex but I take same time to set x and y coordetates in adempiere engine
vpj-cd_ same time to build sql queery :-)
AFalcone: jajaja
karsten-thiemann yes victor but you can share the jasper report
vpj-cd_ also with adempiere
AFalcone: Victor, but with JR you can use subreports :)
vpj-cd_ you can export report format with 2pacl I test with libero and I can import all my report from libero old
CarlosRuiz I see now Karsten and DeathMeat
karsten-thiemann hmm - but only when 2Pack allows to export/import print formats
CarlosRuiz currently document is defined in tab
DeathMeat ha salido de la sala (quit: niven.freenode.net irc.freenode.net).
CarlosRuiz you defined it in Document type
DeathMeat [n=DeathMea@adempiere.reisinger.de] ha entrado en la sala.
vpj-cd_ yes Karsten 2pack let export adempiere report :-)
DeathMeat netsplit
karsten-thiemann CarlosRuiz> you defined it in Document type
AFalcone: yes Carlos, right
DeathMeat i guess i missed something
CarlosRuiz and that is good !!!
vpj-cd_ but the user or development have desition
CarlosRuiz "I see now Karsten and DeathMeat - currently document is defined in tab - you defined it in Document type"
DeathMeat right
vpj-cd_ use x and y of adempiere or are expert in build query :-) if he want jasper :-)
CarlosRuiz I have had requirements of making documents by doc type
vpj-cd_ the time is same :-)
CarlosRuiz or making different document (different logo) for different companies
DeathMeat we did it this way (my emyployer=
DeathMeat )
AFalcone: hehe
vpj-cd_ ok then now you development wotk same way the document in adempiere
CarlosRuiz DeathMeat, did your employeer now that this is being contributed as GPL?
CarlosRuiz I mean, if you make a program as employee, your boss must agree to contribute as GPL
CarlosRuiz not your boss, the owner of the company
DeathMeat he knows and i have his full support in integration in adempiere of any kind of development
CarlosRuiz goods
CarlosRuiz ok, I like your add-on
CarlosRuiz what do you think, Victor?
AFalcone: My vote too
CarlosRuiz hengsin?
vpj-cd_ yes I am agree
hengsin ok
AFalcone: Victor set the X and Y coordinates :)
vpj-cd_ I need also ie now Fernando with 2pack and pach sistem :-)
CarlosRuiz DeathMeat
vpj-cd_ is true Alejandro
AFalcone: hehe
karsten-thiemann he seems to have some timing problem
hengsin Just a note, in future we should build custom jrdatasource to read from Adempiere model instead of direct from db
CarlosRuiz I think the name on C_DocType must be just "reportname" or "ProcessName", not jasperreport
vpj-cd_ if you want build a report or tkae time to set x y or use time
vpj-cd_ build each query to name customer , name product , or any things
vpj-cd_ the time is same
vpj-cd_ :-)
DeathMeat ha salido de la sala (quit: niven.freenode.net irc.freenode.net).
CarlosRuiz sql is one time only - via views
CarlosRuiz karsten, did you get my naming convention?
vpj-cd_ yes but with jasper you need buil any query to retrive some data logic
vpj-cd_ :-)
vpj-cd_ you have only ID :-)
DeathMeat [n=DeathMea@adempiere.reisinger.de] ha entrado en la sala.
CarlosRuiz but only once, Victor
CarlosRuiz I think the name on C_DocType must be just "reportname" or "ProcessName", not jasperreport
CarlosRuiz and I think that we can change adempiere to use that even without jasper
DeathMeat i hate netsplits on irc!
vpj-cd_ if you want go to name product you need a query
DeathMeat :(
vpj-cd_ fi you want go to name partner you need query
vpj-cd_ if you need addres you need a query
vpj-cd_ to any thing in jasper you need a query
karsten-thiemann <CarlosRuiz> sql is one time only - via views
karsten-thiemann <CarlosRuiz> karsten, did you get my naming convention?
karsten-thiemann <vpj-cd_> yes but with jasper you need buil any query to retrive some data logic
karsten-thiemann <vpj-cd_> :-)
karsten-thiemann <vpj-cd_> you have only ID :-)
karsten-thiemann * DeathMeat (n=DeathMea@adempiere.reisinger.de) hat #adempiere-team betreten
karsten-thiemann <CarlosRuiz> but only once, Victor
karsten-thiemann <CarlosRuiz> I think the name on C_DocType must be just "reportname" or "ProcessName", not jasperreport
karsten-thiemann <CarlosRuiz> and I think that we can change adempiere to use that even without jasper
vpj-cd_ in adempiere report I only need set the field and ready :-)
CarlosRuiz DeathMeat, karsten-thiemann, what do you think on naming convention? and let adempiere use the process in C_DocType too (if defined) ?
vpj-cd_ but Jasper is very nice
CarlosRuiz I mean, if defined a process in C_DocType use it instead of the defined in print form
vpj-cd_ the my subggestion to DeathMeat
karsten-thiemann you mean for another engine?
DeathMeat ehm carlos
croo_ vpj-cd, but victor this is an addition .... jasper can make some very complex reports that connot be made using the default print engine!
DeathMeat you mixed something up
vpj-cd_ is build components to adempiere into jasper
CarlosRuiz can we discuss the issue
CarlosRuiz please
vpj-cd_ that solve the problem the id
DeathMeat jasperreport is not intended to be used in C_DocType
croo_ if you want a quick report using the old engine you can still do that!
vpj-cd_ the Jasper is the fatastic :-)
DeathMeat AD_Process has a jasperreport column
DeathMeat C_DocType is planned to get a JasperProcess column
CarlosRuiz that's the point
vpj-cd_ yes croo
CarlosRuiz I would prefer not to have a JasperProcess, but simply a Process
vpj-cd_ I only tell the time to make a report via Adempiere or JAsper is same
CarlosRuiz and we must discuss one issue, let me see the code, please
vpj-cd_ in adempiere you use x and y
croo_ anyway I'm off ... I submitted the changed I spoke of carlos as an RFE 1670185
CarlosRuiz Victor, this is not the discussion
vpj-cd_ in jasper you need know sql
DeathMeat oh ok, but the column is only used by the jasper engine since, nothing else will every invoke the process column on a per doctype level
CarlosRuiz the discussion is if we allow and how ?
croo_ ha salido de la sala (quit: "Leaving").
CarlosRuiz that's the point
vpj-cd_ yes
CarlosRuiz give me 30 seconds
DeathMeat sure i can show you the code
vpj-cd_ ok
vpj-cd_ I sorry
CarlosRuiz where is defined the priority of report format?
DeathMeat but i am late right now since i have a quick appointment, just at the moment, so i have to leave and be back in about 30 minutes
vpj-cd_ DeathMeat
DeathMeat yeah?
CarlosRuiz we can discuss with karsten and he can pass you the info
vpj-cd_ the current you define a jasper report use format report via dempiere?
DeathMeat ok sounds great
DeathMeat yes
karsten-thiemann ähm - sorry for my late question - but I already have commited the store on file system patch without waiting for a CC member response..
vpj-cd_ ok the this working same form that format report
karsten-thiemann I hope that this was ok...
vpj-cd_ you can set in windows and document
CarlosRuiz how is currently the priority of choosing a print format?
CarlosRuiz there is first bpartner, then printform, then system default, I think?
karsten-thiemann if you defined jasper - it will open jasper viewer
CarlosRuiz yes, but the point is that we must keep some priority
CarlosRuiz if bpartner has a print format defined
CarlosRuiz ??
vpj-cd_ if this work same form format report
vpj-cd_ the I can set in windows , BP, Document Type
vpj-cd_ etc
karsten-thiemann it is defined in ReportCtl.java
vpj-cd_ ok fine
vpj-cd_ the is nice functionality
hengsin karsten, you mean if jasperreport is define, it will take top priority ?
karsten-thiemann oh I see what you mean
vpj-cd_ what priority?
vpj-cd_ if you have a report in doc type use this
vpj-cd_ if you have the document in pb the tkae this
vpj-cd_ this is same form what print format
vpj-cd_ is rigth karsten?
CarlosRuiz look in ReportEngine.java line 1116
CarlosRuiz // Prio: 1. BPartner 2. DocType, 3. PrintFormat (Org) // see InvoicePrint
karsten-thiemann not sure and johannis is away for a minute... but as far as I know it is defined in ReportCtl ->start
karsten-thiemann but you are right with your prios
karsten-thiemann ok I see it
CarlosRuiz maybe it will be better to overwrite the process in printformat, but this looks like a loop
CarlosRuiz report have defined a printformat, then printformat can have defined a "Jasper Report"
karsten-thiemann until now we just use the reportCtl class and jasper is only used when you press the print button
CarlosRuiz the priority is just for invoices, other print formats are taken from other tables
CarlosRuiz so, it will pass over the priority of bpartner
karsten-thiemann not sure - but I guess that you are right
vpj-cd_ mmm
vpj-cd_ I think that jasper report should working same form format print
CarlosRuiz yes
vpj-cd_ this way you have need make some special
vpj-cd_ you only need set soruce report
vpj-cd_ if use adempeire engine or jasper
vpj-cd_ but all in the system shoul work same way
karsten-thiemann but it wont affect people who are just using the existing engine - but yes it should work the same way
hengsin what victor describe is the approach use in Posterita to support slip/receipt printer
CarlosRuiz are multi-lingual documents supported with this Jasper approach?
vpj-cd_ yes this another issue :-)
karsten-thiemann it will - just to make it clear - it is a first step - release early, update often
CarlosRuiz Karsten
CarlosRuiz I think that it can be easy if we define the JasperReport URL in print format
vpj-cd_ ok I am agree if the way is same that print format
CarlosRuiz you can look in startDocumentPrint if the printformat is jasper and call jasper engine instead of adempiere engine
karsten-thiemann ok so you mean you define a print document with just the jasper process
CarlosRuiz a printformat
karsten-thiemann right
karsten-thiemann and use all the adempiere functionality and in startDocumentPrint take the jasper way
karsten-thiemann sounds good to me
CarlosRuiz instead of adding column in C_DocType it can be added in PrintFormat
hengsin sounds more coherent
karsten-thiemann yes I think you are right
CarlosRuiz it could be the same in startCheckPrint
CarlosRuiz and even in startFinReport
karsten-thiemann would be a better integration
CarlosRuiz and to complete in startStandardReport :-)
karsten-thiemann :)
AFalcone: yes Carlos, with this approach you can use for all
CarlosRuiz one more thing
CarlosRuiz those print formats can't be selected in report viewer
CarlosRuiz I mean jasper print formats can't be selected in the listbox on report viewer
hengsin how about archiving ?
CarlosRuiz and printconfirm ?
hengsin is archiving supported if jasperreport is used ?
karsten-thiemann good point - jasper can generate pdf - so we can store it in db as well
hengsin also, numbers of copy
karsten-thiemann this is supported
hengsin direct print ?
hengsin I means without preview, print direct to printer
CarlosRuiz Victor
CarlosRuiz just reviewing your add-on to set IsPrinted
CarlosRuiz //vpj-cd e-evolution 15022007
karsten-thiemann wait a second - the question for me is - is it useful to have a jasper integration with limited features and enhance it or must we have all features in befor we can integrate it?
karsten-thiemann I know that DeathMeat is working hard on it and he will add those functions wihtin a short time
CarlosRuiz I think we can have a jasper integration with limited features (enoughly published) and open feature requests for the missed functionalities
hengsin it can be in with limited features, but what is missing, tracker items should be ceated
karsten-thiemann thats what I ment
karsten-thiemann the bazaar way
CarlosRuiz Victor
karsten-thiemann and let other developers help
CarlosRuiz IsPrinted is set when DirectPrint after confirmation
CarlosRuiz but with your change IsPrinted is fixed all the times
CarlosRuiz is it ok?
CarlosRuiz ReportCtl.java line 174
CarlosRuiz 7 lines above in printConfirm is updated when user indicates that was printed
karsten-thiemann sorry CarlosRuiz - I don't have the code here
CarlosRuiz np, the question is for vpj-cd_
karsten-thiemann ah ok
CarlosRuiz ok, we can talk about that later Victor, can we finish the meeting now?
karsten-thiemann so is it okay to commit the code?
CarlosRuiz but in PrintFormat Karsten, not in C_DocType
CarlosRuiz is it ok?
AFalcone ha salido de la sala (quit: Read error: 104 (Connection reset by peer)).
karsten-thiemann after the changes you mentioned: you can look in startDocumentPrint if the printformat is jasper and call jasper engine instead of adempiere engine
karsten-thiemann ok
karsten-thiemann I believe it's the better way :)
hengsin carlos, agree to end now
karsten-thiemann so thank you and good night :)
CarlosRuiz thank you very much to all attendants
CarlosRuiz bye, everybody
hengsin bye all
karsten-thiemann bye
hengsin ha salido de la sala.
karsten-thiemann ha salido de la sala ("Verlassend").
vpj-cd_ ha salido de la sala ("Abandonando").
AFalcone [n=AFalcone@190.49.77.121] ha entrado en la sala.
cjerome ha salido de la sala.
DeathMeat hm
_vclark ha salido de la sala (quit: Remote closed the connection).
