2024-11-11 Meeting notes

1 min

Scribe

All

@Marc Johnson is next, followed by @Ingolf Kuss

Reminder:  Please copy/paste the Zoom chat into the notes.  If you miss it, this is saved along with the meeting recording, but having it here has benefits.

5 min

Liaison Updates

@Maccabee Levine

@Tod Olson

@Jakub Skoczen

@Craig McNally

@Jenn Colt 

• CC: @Maccabee Levine

  • Lee Braginsky presented on dataset anonymization for Better Sample Data working group. Randomized data while maintaining patterns. Looking for 2 devs, mid-senior, familiar with FOLIO DBs, fluent with Java or Python and SQL, for a few months. A tool for libraries to run themselves to create the data set, test, make sure no sensitive data. (Or EBSCO can run.)

  • AWS Costs. $30K cap is widely communicated. ACRG is adding budget alarms to help with reporting. A cost chart from AWS console will be shared with ACRG twice monthly, for Shawn to report to CC. No plans for circuit breakers. Continued discussions of what is 'shared' infrastructure that community would fund, split conversations about this current FY budget vs planning for future FYs.

• PC: @Tod Olson: Rolling Out of Prioritization process for the SIGs. Where does the community as a whole see the priorities?

• RMS Group: @Jakub Skoczen: No updates.

• Security Team: @Craig McNally: No updates

• Tri-council Application Formalization:   @Jenn Colt

  • Group has paused while POs and App Interaction work on spreadsheet from Taras

  • Talking a bit outside of meetings about  sequencing,  can discuss during Eureka topic

1 min

Upcoming Meetings

All

• Nov 11, 2024 - Regular TC meeting

• Nov 13, 2024 - Dedicated discussion:  Topic TBD

• Nov 18, 2024 - Regular TC meeting

• Nov 20, 2024 - Dedicated discussion: Topic TBD

• Nov 25, 2024 - Regular TC meeting

• Nov 27, 2024 - Dedicated discussion: Cancel for American Thanksgiving?

5 min

TCR Board Review

All

mod-linked-data is new. No volunteers. @Craig McNally & @Jenn Colt will determine a reviewer from the rotation list.

@Julian Ladisch: mod-reporting is finished. Could be moved to "under review".

5-10 min

Technical Council Sub-Groups Updates

Static code analysis: No meeting

@Ingolf Kuss / @Julian Ladisch : The group will re-meet on Tuesday. We need experts for the Go Language. Julian points out, that there is a disagreement inside the group.

Developer documentation: Nothing new

30 min

GitHub RFCs

Wiki RFCs

All

• Discussion and vote on Go RFC

  • Discussion about voting rules. ... we are at full attendance now (11 members).

    • @Ingolf Kuss: current voting rules are strange. I can understand that we need 50%+ of attendance to make a decision at all. But if attendance is only 7 people, it does not feel right that all but one need to vote with "yes" for something to get accepted. I propose that in this case, Slack votes should fill up the rest of the votes (for those, who could not attend).  Alterantively, a simple majority could be considered sufficient (in case of a low attendance, this does not feel right, either, though).

  • Slack voting- acceptable?

    • are there objections to having Slack voting ? ...

    • @Jakub Skoczen: Let's discuss as part of a bigger process.

  • @Jeremy Huff votes for TAMU as Jason's proxy.

  • Vote: 7 yes, 2 no, 2 abstentions.

  • @Marc Johnson : Can we shedule a subgroup "new language adoption policy" ? We have no sufficient policy for mod-reporting or for any future Go module.

    • The static code analysis subgroup already deals with the Go-language.

    • @Jeremy Huff: Multi-language support. The CI-CD pipeline. That is a big point of discussion when it comes to multi-languages support. There could be polcies that retroactively will apply to other languages.

    • @Jakub Skoczen: Will we have a fixed policy that will not let us post an RFC (when a new language comes up) ?

    • @Marc Johnson: A minimum thing that has to be satisfied (e.g. build tools). That could be the entire scope of the group. Is there a limit to the number of languages that we want to take on ? 7 ? 10 ?? Could be we have to deprecate one of the existing languages if another language comes in. That will not anymore be sustainable for the community.

    • @Craig McNally: Usually the team supports the language. We should not make it easier to introcude new languages.

    • @Tod Olson: What Marc outlined is the bare minimum of consideration. Most compelling in this instance were AWS costs.

    • @Jeremy Huff: Would containerization (of the modue) make the concerns less ? You need to have a technology to build containers.

    • @Jason Root : A spring language proposal of TAMU was turned down. It popped back up later.

    • @Jenn Colt: We need a Wednesday discussion on this policy.

• Discussion of Eureka RFC

  • @Marc Johnson: We need to switch policies if we adapt Eureka for the Sunflower release.

  • @Julian Ladisch: The community wants to have claritiy. I am in favor of adopting Eureka for Sunflower.

  • @Marc Johnson: Let's make a special case. Ditch the RFC process, ditch the module reviews if we adapt for Sunflower. Just make a decision in the Tri-Council / the community.

  • @Marc Johnson: Release process for Sunflower starts in February. When is the last responsible moment for this decision ? We run out of time to make this decision. There is a lot of work to do to adapt it in Sunflower.

  • @Ingolf Kuss: I find it more relevant to discuss when will be the end-of-support time for Okapi, than when will Eureka officially be supported. It is clear that one can run Eureka already now, if one wants to. The question of official support is not so crucial for those who want to change to Eureka as quickly as possible. But it is is crucial for those who have limited capacities to change their enviroment ti Eureka.

  • @Craig McNally: We are talking about two releases of backword support. That means Okapi support will end with Trilium.

  • @Ingolf Kuss: Not only the community reference environment teams have to be ready to adopt it in Sunflower, but also the Sys Ops and Implementers have to be ready to change their production environments at a certain deadline. I have not heard any readiness from this side.

  • @Jenn Colt: Should we make an extraordinary procedure for the Eurerka adoption ? We should discuss in a Wednesday meeting.

10 min

Additional Eureka discussion

• What to do about Eureka TCRs?

1 min

Decision Log

All

Nothing new here.

5 min

Officially Supported Technologies (OST)

All

Check Recurring Calendar...

5-10 min

TC member changes

All

• Nominations are open

NA

Zoom Chat

17:09:14 Von Craig McNally an Alle:



relevant deadline for sunflower: Jan 24th



17:16:11 Von Marc Johnson an Alle:



If the TC has agreed, then I don’t need a meeting, only a clarification of the documented rules



17:16:50 Von Julian Ladisch an Alle:



Can we update the distributed configuration RFC with the outcome? https://github.com/folio-org/rfcs/pull/40/files



17:36:24 Von Jenn Colt an Alle:



I always assume it shows in order but maybe it doesn’t…



17:36:50 Von Maccabee Levine an Alle:



Replying to "I always assume it s..."







It does, but sometimes people put their hand down and back up.  Didn't see what happened here.



17:36:59 Von Jakub an Alle:



Thanks Marc, adding guidelines makes a lot sense to me and the process around Go could inform it.  I would be reluctant to set some hard limits on e.g the number of languages. I don’t think we will ever face having to apply those limits but they feel artificial to me.



17:38:51 Von Jakub an Alle:



Yes, for a high bar for adding nee languages ans new technologies in general. Strict but clear criteria.



17:38:53 Von Craig McNally an Alle:



To be clear, it's not a problem with Go, my objection is more about expanding the set of languages Folio supports for module development.



17:39:03 Von Marc Johnson an Alle:



I also believe we should do this for tech and other architecture aspects







However I think that’s scope for a different conversation



17:40:29 Von Marc Johnson an Alle:



Where I’m coming from is organisations like Spotify,







Initially they went down a polyglot route. Later they decided to default to Java and teams have to pass a high bar to deviate from that



17:40:54 Von Marc Johnson an Alle:



Yes, it’s me that states that Docker has never been officially adopted



17:42:12 Von Marc Johnson an Alle:



Containers only partially reduce the impact of some of the adoption



17:42:25 Von Craig McNally an Alle:



Reacted to "Containers only part..." with