2020-08-05 Meeting notes

Date

05 Aug 2020

Attendees

Discussion items

Time	Item	Who	Notes
5 min	New business/agenda items?	TC	Check on other potential agenda items.
10 min	ElasticSearch and Circulation from Shanghai	VBar	Updates on latest activities/discussions/progress ElasticSearch - active effort, blueprint agreed to - Shanghai is implementing towards that, Shanghai has a very tight deadline - by 20-September. Had a POC of their own - expanding that to fit the FOLIO wiki approach. Going to decouple systems by using Kafka. 3 Modules being proposed at this point: Module to present interface for querying/etc Module to pull messages off Kafka and push into the index (not using Pub/Sub for various reasons) Module to place messages on Kafka (a little unclear - being discussed) We are hoping that the overall project can adopt and/or build upon from the work Shanghai is doing (they are building in a separate repo - GitLabs). Note that this does not mean that the architecture for the search proposal has been accepted/adopted to be included in Folio. We also are going to need to evaluate the implementation choices that they make in terms of module and interface separation/independence. If we do bring that work in we could change some of their decisions. Circulation - Recall their requirements - extreme transaction rate, users and volume of material. Have decided to stay with current model and ended a POC that would have been a dramatic departure. Recent improvements in Okapi and Performance encouraged them to do this. Are adding developer(s) to work on the Core Functional team.
25 min	Process for external modules being introduced into FOLIO	Mike Gorrell	Lehigh has built a POC that has shown a lot of interest by other adopters who are waiting for FOLIO functionality. Likewise, Duke has built a module to integrate with an LTI system. These efforts raise the question as to how a module that was built by a lone developer/institution could make its way into the FOLIO build and deployment process. Key Questions: Do we want/need to allow these modules into the "FOLIO base-LSP "? If so, what are the expectations/requirements and gates/bars wrt testing, look and feel, etc - same as all other modules. Does it have to be approved? Right now the process if very loose. If not, what are the implications? Hosting organizations should have a mechanism for pulling "external" modules into their system. Are these well known and robust enough? This will imply/require the hosting provider to vet the module according to its standards and processes. Does Folio need a mechanism for making available and pulling in external modules into an installation? (App Store) If there are dependencies on this external module from platform-core modules, this module may no longer be considered "external". How does the community stay aware of modules that are available even though they may be external? Is FOLIO a "unified thing" or a system that allows you to assemble a "collection of things"? Next Steps/takeaways: Should formalize the approval process, which would start with the product needs and requirements and once validated then technical vetting (definition of done, dependencies, architecture, etc.). Determine whether we need better ways to allow external apps to be identified, accessed, etc... (the App Store).
20 min	Technical Debt discussion	TC	To be discussed more next week: We have the DEBT project and we also have a new IssueType of "TechDebt" that people have started to use in JIRA. What would we like to establish for a review process? Frequency Who (subset of TC, Tech Leads, Cap Planning, other)?