IDempiere/FullMeeting20121128

From WikiQSS

Table of Contents | Full Meeting Minutes | Full Meeting 2012-11-28

CarlosRuiz: Good morning
Deepak: Hello Everybody
a42niem: hi all
nmicoud: Bonjour
nmicoud: CarlosRuiz, will you have time to review tickets in queue (with a patch attached) ?
CarlosRuiz: yep
CarlosRuiz: about IDEMPIERE-480 - did you see the approach from Adaxa?
nmicoud: The approach seems fine.
nmicoud: But, speaking with my accountant, he prefers to allow multiple dates on bank statement
nmicoud: Thus, i would recommend to integrate the patch of Adaxa
nmicoud: and i will modify it for our installs in order to make accountant happy
CarlosRuiz: ok
nmicoud: Seems that it would be safer to allow a single date for every document
a42niem: concerning Manufacturing Light
a42niem: it is indeed problematic when there are already entries in table m_production
CarlosRuiz: ah yes - Javier reported that - the migration script needs to be fixed
a42niem: well, the additional columns are mandatory but for the already existing lines ther is no correct "default" for them
a42niem: the best fix would be to empty m_production first and then apply migration :-D
CarlosRuiz: thinking if we better make those columns nullables - and enable the old production window in read-only mode for these records - and on the new window we can filter out the records with null product
a42niem: yep, good idea
a42niem: maybe we can use mandatory logic to make them mandatory in the new window
CarlosRuiz: yes, overwrite mandatory
a42niem: how to proceed? change the existing scripts? or "deprecate" them and create new ones?
CarlosRuiz: interesting :-)
CarlosRuiz: I think we need both
a42niem: hm, "bug fix" scripts to amend them won't work i guess, they come too late
CarlosRuiz: change the existing scripts to drop the "NOT NULL" condition
CarlosRuiz: hmmm - thinking
CarlosRuiz: what if we write an additional script for people coming from 361 -> 862a_ManufacturingLight.sql.361
CarlosRuiz: and we advice people to apply 862a_ManufacturingLight.sql if they have old production records and want to preserve that date
CarlosRuiz: data
CarlosRuiz: the only difference would be to change the NOT NULL by null on lines 72 and 81
CarlosRuiz: :-) or we can make an annotation on the migration script for people coming from 361
Deepak: Annotation for executing failled script manually with suggested changes is good idea. We should consider ant script as it may try to run both script in case of keeping two version
CarlosRuiz: and we create a wiki page for "Backward Compatibility Notes" and/or "Migrating from 361 Notes"
CarlosRuiz: WDYT a42niem ?
a42niem: good plan, i create a ticket to document it
a42niem: hm, bug or improvement?
CarlosRuiz: I think is a migration bug
a42niem: ok
Edwin_Ang: hi
CarlosRuiz: Hi Edwin_Ang
Edwin_Ang: hi carlos
nmicoud: hi
Edwin_Ang: what r being discussed?
Edwin_Ang: hi nmicoud
CarlosRuiz: not specific discussion at this moment
CarlosRuiz: I'm checking peer review queue - and making some MStorage concurrency tests
a42niem: CarlosRuiz: created IDEMPIERE-521, please review
Edwin_Ang: well then, i would like ask about the web services work done by Deepak
Edwin_Ang: can i access the work?
CarlosRuiz: Thanks Dirk - looks good
a42niem: .
CarlosRuiz: ah Edwin_Ang - Deepak just left - maybe is too late there (he's at Ahmedabad)
CarlosRuiz: he's still working on it and I think is going to commit soon
Edwin_Ang: ic
Edwin_Ang: ok no prob
nmicoud: Is there a way, in a M class, to ask something to user, in swing and webUI.
nmicoud: Let's take an example : you want to delete a record ; if some conditions are verified, a 2nd confirmation message should appear
CarlosRuiz: I read your thread on red1 forums
nmicoud: But i can only code it using ADialog of FDialog ; is there a common way to do it ?
nmicoud: ah ok :)
CarlosRuiz: on 361 you could that trick - because you added the code to customization and that had access to both zk and swing
CarlosRuiz: but on idempiere I think that's not so easy - cos zk and swing are independent bundles
nmicoud: i know it is possible in callout, using mTab.fireDataStatusEEvent("",alert,false); but not tested on idempiere
nmicoud: and i would prefer a solution that could be used on idempiere (but if it is also ok for 361, i'll take it)
CarlosRuiz: if callout can do the trick - then I would say do it on callout
CarlosRuiz: cos is not a strict rule - the user is still allowed to delete on certain conditions - right?
buildmaster: Project iDempiere build #527: SUCCESS in 10 min: http://jenkins.idempiere.com/job/iDempiere/527/
buildmaster: globalqss: IDEMPIERE-385 Resolve M_Storage locking and data consistency / lock storage record being updated
nmicoud: Yes, but the message should be something like : "Are you really really sure that you want to delete it ?"
nmicoud: He could say "yes" or "no" ; and using a callout, i can only say something, user cannot say yes or no
CarlosRuiz: I think I solved a similar situation somewhere forcing the user to check a flag on the window
CarlosRuiz: another option is to create a button that deletes the record
nmicoud: to force user to used this button when condition are verified ?
CarlosRuiz: yes - on the process you can verify whatever conditions - and force a double check - two flags on parameters :-)
CarlosRuiz: or - you can check on delete callout or validator - that the user checked a flag on the window saying "I'm sure to delete this" - kind of that
nmicoud: Yep, i will try this, thanks for the tip
CarlosRuiz: nmicoud, I need to move out - thanks a lot for attending these meetings - I appreciate it a lot - you can bring always good things to discuss
CarlosRuiz: c u later
nmicoud: bye bye