• consider uvicorn poetry plugin as an example
  • structurzr
  • slack inspector stale bundles daily

2025-07-31

  • flowable incident

    • after dropping all sorts of things to get the engine to start then had to recreate some
    • probably have database in a broken state and need urgently to migrate to either clean 6.7.2 or hopefully 7.x
  • flowable active data

    • probably the only thing we need to keep are Alife tasks
    • impersonate Hazel thru keycloak, go to tasks.knowprocess.com/dashboard
      • 180 tasks all instances of definition SimpleToDo:2:d72742d7-418a-11ed-82a8-8233c9b99355
    • all active instances total 1160: curl -X GET --header 'Accept: application/json' --header 'Authorization: Basic <redacted>' 'https://flowable.knowprocess.com/flowable-rest/service/runtime/process-instances?suspended=false'
      jq -r '.data[].processDefinitionId' all-proc-inst.json | sort | uniq
      96eaa67c-4b07-11f0-b7ea-8a6894413383 # count 29 (automated tests?)
      CreateContactAndAccount:14:f8d5bbc3-aab2-11ec-b141-1e970d1ca1ca
      MergeMemoTemplate:21:902dc7a0-c14b-11ec-b59a-9aa199c37e18 # 193, from 2022, real data from hazel, represents failures?
      MergeMemoTemplate:22:17ee15bf-6145-11ed-bf9a-860764942ce5 # 12, from 2022, real data
      RecordAndRespondToEnquiry:1:52db9c05-7545-11ec-9945-1a8958d6d2ce # count 36
      RecordAndRespondToEnquiry:3:cbac5934-b0e2-11ec-9358-1e970d1ca1ca # 42 (34 mine)
      # down to 5 that need to be cross refrenced with sdu notes for caulfield and teresa
      SelectAndSendResponse:1:52db9c04-7545-11ec-9945-1a8958d6d2ce
      SendMemo:19:abf5f2df-a3d0-11ec-9526-1e970d1ca1ca
      SimpleToDo:2:d72742d7-418a-11ed-82a8-8233c9b99355 # count 206, of which 180 hazel 4 me + ???
      
    • use admin ui to find 231 deadletters, close what relates to me testing and fixing
    • cleaned up all the deadletters, removed all the old miwg demos and a lot of tests
    • all that is left is simple todo, 5 sdu enquiries and some sub-processes
      • TODO would be nice to find a way to prove subs are orphaned
  • flowable plan

    • test apps under 7.x: enquiry, send memo, simple todo, others?
    • write a script to recreate the todos as new process instance on new install
    • make the switcheroo
  • local dev env

    • flowable-rest (on h2): docker run -p 8080:8080 flowable/flowable-rest:7.1.0

2025-07-30

  • continue modernising other jasmine specs
  • replace leaked flowable rest-admin creds
  • accidentally upgraded flowable database
    • attempt to downgrade revealed a fragmenting of the schema version in ge property table
    • trial and error removal of upgrade columns slow and so far unsuccessful
    • no idea if pushing fwd will work, evaluate?
    • matched ge prop values to clean 6.7.2 install

2025-07-29

  • jasmine intgn test modernisation
    • update to 5.9.0
    • cli and standalone (embedded) integrations
    • disable random order
      • copilot got that wrong when faced with spec file, suggesting single it function
      • prompting in context of package.json identified –random
      • finally got standalone working by reading docs

2025-07-28

  • Will
    • paddleboard
    • drop at residential

CRM

  • merge swagger fix

    • GHA blocked due to being so out of date
    • bumped all GHA versions to latest, merged and retried -> ok
  • check test coverage

    • already set to min of 30% and enforced
  • bump java vsns:

    • started with spring boot and framework -> ok
    • centralise commons-io vsn, had only applied vuln fix to single module -> ok
    • same for commons-csv
      • much trouble finding compatible vsn until finally check csv pom–doh–tests pass
      • but… test failure
         Error:    ContactTest.testHashCodeForStackOverflow:108 expected: link.omny.custmgmt.model.Contact@5b410e27<Contact(id=null, firstName=Fred, lastName=Flintstone, mainContact=false, title=null, jobTitle=null, email=fred@bedrockslateandgravel.com, emailConfirmed=false, emailOptIn=null, emailHash=null, uuid=null, phone1=null, phone2=null, phone3=null, address1=null, address2=null, town=null, countyOrCity=null, postCode=null, country=null, existingCustomer=false, stage=null, stageReason=null, stageDate=null, enquiryType=null, accountType=null, owner=null, doNotCall=false, doNotEmail=false, twitter=null, facebook=null, linkedIn=null, description=null, source=null, source2=null, medium=null, campaign=null, keyword=null, alerts=null, tags=null, firstContact=null, tenantId=omny, customFields=[], account=null, accountId=null, notes=[], activities=[], documents=[], now=Mon Jul 28 17:28:59 UTC 2025, customHeadings=null)> but was: link.omny.custmgmt.model.Contact@1c124240<Contact(id=null, firstName=Fred, lastName=Flintstone, mainContact=false, title=null, jobTitle=null, email=fred@bedrockslateandgravel.com, emailConfirmed=false, emailOptIn=null, emailHash=null, uuid=null, phone1=null, phone2=null, phone3=null, address1=null, address2=null, town=null, countyOrCity=null, postCode=null, country=null, existingCustomer=false, stage=null, stageReason=null, stageDate=null, enquiryType=null, accountType=null, owner=null, doNotCall=false, doNotEmail=false, twitter=null, facebook=null, linkedIn=null, description=null, source=null, source2=null, medium=null, campaign=null, keyword=null, alerts=null, tags=null, firstContact=null, tenantId=omny, customFields=[], account=null, accountId=null, notes=[], activities=[], documents=[], now=M
        
  • look at jasmine tests in tests/index.html

  • considered serving crm-ui from crm-server but hit issue even building so put that on hold

  • discarded

    • create robot framework tests
    • integrate robot tests into GHA

2025-07-25: driving back from Seaweed

2025-07-26

  • crm fix swagger ui

2025-07-24

  • couple of applications through jobserve
  • beach

2025-07-23

2025-07-22

  • Vinkatesh job in Bradford
    • reading about django
  • Raul / OpenLP
    • reading about hatch

Goal setting till end Aug

  • KP
    • major upgrade to all components
      • CRM API: patch, regression test suite, compiled deployment
      • Modeler: complete migration even if not perfect feature match
      • Flowable: migrate to 7, regression tests, possibly compiled deployment
      • CRM UI: last priority, rebuild in svelte with bootstrap 5, maybe accounts page only?
  • Job
    • login jobserve and Reed daily
  • CPD
    • Real world
    • Certified K8s?
    • AI: eg Bible substack
  • OSS
    • OpenLP: hatch

END ELSEVIER

2025-07-17

  • all remaining PRs in, including adopting an official ibuild release in all components.
  • multiple releases
  • closed the poetry GHA as no time and no benefit on ephemeral single-use build images.

2025-07-17

  • common error on captest image build
    Dockerfile:11
    --------------------
       9 |     # Install Zscaler cert to permit apk and pip to run on ELS network
      10 |     COPY certs/* /usr/local/share/ca-certificates/
      11 | >>> RUN update-ca-certificates
      12 |     # dev packages requires to build cffi, which is in turn required for pyproject.toml based installs
      13 |     RUN apk update && apk add aws-cli bash curl gcc jq libffi-dev libc-dev openssl python3-dev
    --------------------
    ERROR: failed to solve: process "/bin/sh -c update-ca-certificates" did not complete successfully: exit code: 255
    

2025-07-11

  • Inspector release
    • latest issues are with nonprod terraform: GHA does not have perms to change K8s RBAC

2025-07-09

2025-07-09

  • ceip-6828 fix inspector ui test that fails when run in gha
  • open ticket for git hub rate throttling

2025-07-08

  • docs review completed

2025-07-07

  • docs:
    • plantuml
  • chase PRs

2025-07-04

  • complete ibuild changes inc. boost Inspector test coverage
  • pm: w Thomas relaxing the kyverno policy for SD to enable the fluent bit lua filter

2025-07-03

  • Ashish sbom PR again
  • 1-2-1
  • ibuild
    • check / fix the inspector_ui terraform post changes

2025-07-02

  • React real world
  • Ashish sbom PR again
  • complete website test at least to next commit
  • MIWG
    • fix MIWG demo (feel namespace)
    • meeting

2025-07-01

  • Ashish sbom PR
    • probably too long looking at specific refactoring instead of advising
  • ibuild PR to fail build