Hi Shivansh, I think based on comments for others, we should hold it for now.
On Sat, Nov 14, 2026 at 1:46 AM James Dailey via dev < [email protected]> wrote: > +1 > > Vishwas - agree 100%. > Less code to maintain here is better. > > > > On Fri, Nov 13, 2026 at 12:09 PM Vishwas Babu <[email protected]> > wrote: > >> Hi Shivaansh, Aman, >> >> When we originally built out this functionality for Group Loans or >> Savings which needed Individual monitoring, there was a need for the same. >> I am not sure if the same is true today as well. Are there any implementers >> or MFI's on the list who have asked for the same ? If not, I >> wouldn't bother adding code to the back-office to show such niche >> functionalities . This extends to stuff like collection sheets etc and >> anything else that does not make sense with respect to hyper personalized >> individual lending. The aim of the backoffice should not be to have feature >> parity with an interface maintained by a sister open source community like >> Mifos or to show all possible functionality exposed by the core. >> >> My suggestions would be to reduce the scope to have tight >> representations of the things which make sense to people right now, like >> being able to express all the functionality exposed by progressive loans >> and working capital. The less code you have the easier for people to >> maintain going forward, people value tokens and nobody would want to burn >> more than they have to. >> >> Regards, >> Vishwas >> >> On Thu, 12 Nov 2026 at 21:21, Shivaansh Pandey <[email protected]> >> wrote: >> >>> Hi Aman, >>> >>> Thanks, will do. Along with the unit and Feign integration tests, I'll >>> add Cucumber E2E scenarios for GLIM loan creation, with and without >>> totalLoan. If the E2E part turns out large, I'll say so on FINERACT-3228 >>> before splitting anything. >>> >>> Thanks, >>> Shivaansh Pandey >>> >>> On Fri, Nov 13, 2026 at 10:48 AM Aman Mittal < >>> [email protected]> wrote: >>> >>>> Make sure to include relevant tests including e2e cucumber tests for >>>> this. I see that the code lacks sufficient coverage for this. >>>> >>>> Regards >>>> Aman >>>> >>>> On Fri, Nov 13, 2026 at 10:34 AM Shivaansh Pandey < >>>> [email protected]> wrote: >>>> >>>>> Hi all, >>>>> >>>>> Following up on my mail of 13 Oct, sorry this took longer than the >>>>> date I gave. I didn't hear otherwise, so I've filed the GLIM part as >>>>> FINERACT-3228: >>>>> https://issues.apache.org/jira/browse/FINERACT-3228 >>>>> >>>>> I'm working on the fix (validator rule, unit test and Feign >>>>> integration test) and will link the PR on the ticket. I've only taken the >>>>> GLIM part; GSIM and the collection sheet stay with Tofunmi. >>>>> >>>>> Thanks, >>>>> Shivaansh Pandey >>>>> GitHub: shivaansh0610-LUFFY >>>>> >>>>> >>>>> On Tue, Oct 13, 2026 at 11:47 AM Shivaansh Pandey < >>>>> [email protected]> wrote: >>>>> >>>>>> Hi Tofunmi, James, all, >>>>>> >>>>>> I joined the list after Tofunmi's mail of September 12, so I can't >>>>>> reply to it directly. The thread is here: >>>>>> https://lists.apache.org/thread.html/3oy7fxpc7cjdn2osc6ppnr5ttss327g8 >>>>>> >>>>>> Thanks for the write-up. I reproduced the GLIM part (#1) on current >>>>>> develop (f9c2fcdc1c, 1.16.0-SNAPSHOT, PostgreSQL), and it is narrower >>>>>> than "no working request shape": >>>>>> >>>>>> - POST /v1/loans with loanType "glim", isParentAccount true and >>>>>> totalLoan 10000 returns 200 with glimId 1. >>>>>> - The identical request without totalLoan returns 403 >>>>>> error.msg.unknown.data.integrity.issue, and the backend log shows the >>>>>> not-null violation on glim_accounts.principal_amount. >>>>>> - GroupLoanIntegrationTest's GLIM tests (which send totalLoan) pass on >>>>>> PostgreSQL in CI at the same commit. >>>>>> >>>>>> So this looks like a missing validation rather than a Postgres >>>>>> incompatibility: the absent totalLoan reaches the glim_accounts insert >>>>>> as null, and the database error comes back as a generic 403. I have >>>>>> not tested MySQL or MariaDB. >>>>>> >>>>>> I searched Jira for GLIM, totalLoan, isParentAccount and glim_accounts >>>>>> and found no ticket for this (FINERACT-1279 is an unrelated 2020 >>>>>> charge / linked-savings case). >>>>>> >>>>>> Proposal: I will take just the GLIM piece—implementing a rule in >>>>>> LoanApplicationValidator that returns a 400 specifying the missing >>>>>> `totalLoan`, along with tests. Tofunmi, you mentioned filing JIRA >>>>>> first: would you like to file the GLIM ticket, or shall I file it and >>>>>> link #376? If I don't hear otherwise by Friday, October 16, I will >>>>>> file it and link back. GSIM and the collection sheet stay with you. >>>>>> >>>>>> Thanks, >>>>>> Shivaansh Pandey >>>>>> GitHub: shivaansh0610-LUFFY >>>>>> >>>>>
