Skip to end of metadata
Go to start of metadata



Date and Time: Every Wednesday, 12:00-1:00pm ET / 11:00-12:00pm CT / 9:00-10:00am PT

  • Video Bridge 220488, Dial 812-856-7060 to connect via phone (conference number 220488#, See PolyCom Directory for connection info)
  • Xmeeting(Mac) :
  • Ekiga(Linux) :

Mailing List:


  1. Roll Call
  2. Shared api project discussed at KCW - Larry Symms
  3. Kuali Stackexchange
  4. Javascript documentation and Testing repositories - Erik Meade
  • KULRICE-9355 - Data cannot be retrieved due to an unexpected error
  • KULRICE-9356 - Data cannot be retrieved due to an unexpected error
  1. KULRICE-9358: Side effect of disabling the SessionDocumentService, the WorkflowDocument is no longer in the session - gilesp


Bryan H.
James S.


Shared api project discussed at KCW - Larry Symms

  • Changes in KFS api, caused significant breakage in Kuali Coeus
  • Discussion evolved from there
  • Need to start thinking about managing the shared apis across the projects so far and having better governance
  • One thing that we came up with was putting shared apis in external repository where there would be a governing body managing all of those changes and making sure all of those changes were still intact
    • communication could then float from there
    • managing the changes as well
  • That's where the discussion ended up at KCW
  • Eric - sounds similar to the work done in Rice on version compatiblity for service APIs
    • some challenges with that because of patch versions
  • Peter - some of the challenges relate to versioning
  • KFS calls KC and KC calls KFS
  • One thing that's been helping us quite a bit is that we get a WSDl from KFS and we generate the classes from KFS
    • also know that in Rice we know you use annotations
  • Are you using DTO?
    • yes, but we have two options
    • one is to make a direct web service call to KC from KFS, or to go via the KSB
  • James - when you generate the wsdl for KC and you need an account object, how do you get it?
    • we generate it from the WSDL
  • Should there be another group that talks about this?
    • building on Gayathri's questions, we talked about a shared test drive with everything running
  • Who would be interested in participating in that
    • Gayathri
    • Bryan
    • Tom Coppetto (services lead from KS) or one of his delegates
    • Eric Westfall
    • Jeremy Hanson
  • Automatic code review in Fisheye for api changes
    • Peter will present at KTI next week
  • Eric will take this as an action item to spin up the working group through TRC and convene the group

Kuali Stackexchange

  • need to propose enough questions and get enough votes in order to move to the next phase
  • Jonathan - what we are supposed to be doing right now is filling the proposal site for the types of questions that would be on a Kuali stack exchange
  • This is really just discussion over the site, but we need 60 followers and 40 questions
  • Jerry - we're hoping that by moving to something like this than just the mailing list that the community will help out more with answering questions
    • mailing list has been quite active and it can be fairly hard for the team to keep up
  • Also, people don't tend to search the google groups, this will make answers on the internet more persistent and searchable
    • definately see this on various
  • To clarify, this is not meant as a replacement for mailing lists
  • If you have people at your instituations who are using Kuali, encourage them to follow this site
    • and also have them add some questions!
    • also, if you are upvoting a question, each person has a limit of 5 upvotes to prevent a single person from doing all the work
  • Claus - on the proposals for the Kuali Rice, basically how that works is that currently it's in the definition phase
    • goals is to create questions that are on topic and figure out whether or not they are on topic
    • next phase is committement phase, at that point questions are locked in place and people are supposed to sign up for committment
    • from there we go into beta phase, we can ask and answer questions, just like on the regular site
      • that shows that we can create enough interest in the community
    • there is a two year limit, if you don't make it in two years than the proposal is cleared off
  • would also be good to have some functional people in there
  • David - was planning to bring this up at the next TRC meeting, in 2 weeks at 4 pm eastern

Javascript documentation and Testing repositories - Erik Meade

  • jsdoctk for javascript documentation, only maven plugin we've been able to find
  • other one is for saucelabs connection
    • allows for running automated tests across different platforms, different browsers, etc.
  • all of the libraries in these repositories are Apache 2.0

KULRICE-9358: Side effect of disabling the SessionDocumentService, the WorkflowDocument is no longer in the session - gilesp

  • came out of disabling the KNS session document service
  • replaced with a "noop" service with empty methods or methods that just returned null
  • one of the things going on in original implementation is that there was some code that stashed WorkflowDocument in the user session
  • when we disabled it, that behavior went away and it had an effect on the dev version of KFS because they were making changes to the application document status on the workflow document and not persisting it manually via the KEW WorkflowDocument service
    • was dependent on WorkflowDocument sticking around in session so change would be there on the next request
    • this changed behavior in some documents that were examining application document status to determine if fields were readable/writable
  • to complicate things, in Rice 2.1 disabled KNS but not KRAD session document service
  • Trying to figure out best approach
    • are there concerns with this not being stashed in the session?
    • KFS in this particular case was able to work around this by saving the workflow document on each request where it was changed
      • wonder if that side effect might bite
  • Eric - my concern would be the cost/performance of loading the WorkflowDocument each time
  • Jerry - one other question we had, this session had special handling for WorkflowDocument because it wasn't serializable
    • wondered why, because the WorkflowDocument doesn't implement serializable, any reason it can't be
    • Peter - i had a theory about that, i think that what it probably is is that the form is being serialized and kept around with the document, so could be reloaded on a new session
      • the workflow document could have gone through transitions while the serialized body was sitting out there in the database, so if you reload that you could have invalid state in your workflow document
  • Jerry - i think we should just be able to modify the document so that the WorkflowDocument is not marked transient any longer and is Serializable so we can remove the transient and the need to refetch the WorkflowDocument
    • Bryan - do worry mainly because of the timing of the change, at this point we're pretty much done making changes, so I would be concerned with any changed behavior
      • if we look at doing it later on down the road, that would be better
      • so maybe the right way forward is to do it with the utility class instead of the session document service in 2.1.6. Then maybe in the next 2.2 version remove transient from the field on the document header
  • that sounds reasonable to everyone

Open KULRICE JIRAs Marked for KTI Review

List below does not include JIRAs that are status: In Progress

key summary assignee reporter fixVersion priority created updated Received fatal alert: handshake_failure

View these issues in JIRA
  • No labels