# Decisions 9th March 2022
* Will make this meeting bi-weekly until we've worked our way through the issues listed here.
# 9th March 2022, 21:00-22:00 UTC+0
Attendees: Anastasia Klimchuk, Felix Singer, Thomas Heijligen, Stefan Reinauer, Edward O’Callaghan, Maciej Pijanowski, Miklos Marton, Martin Roth, Nico, Carl-Daniel, Piotr, Angel Pons
Agenda:
* [FelixS] First flashrom meeting. Let’s introduce ourselves and say what we are working on at flashrom. * [FelixS] flashrom has been accepted to GSoC * [GSoC project page](https://summerofcode.withgoogle.com/programs/2022/organizations/flashrom ) * [Official announcement](https://mail.coreboot.org/hyperkitty/list/flashrom@flashrom.org/thread/MGGC2... ) * Current tasks? What is left to do? * [aklm] Respond to GSoC candidates, help and guide them * [aklm] Help improve [Development Guidelines](https://www.flashrom.org/Development_Guidelines), [Easy Projects](https://www.flashrom.org/Easy_projects) pages. * If you want to help and don’t know what to do, contact Anastasia or Felix. * [Martinr] Who is the flashrom leadership and how is that decided? * [Nico] For the last years this was rather implicitly who was maintaining the code to be safe and sound. And helping users on IRC and the ML and such. * Is this how the project wants to continue? It seems somewhat disorganized. Just because someone is a good coder doesn't necessarily make them a good leader. * Who decides "safe and sound"? Are there any metrics for this? * The coreboot project has 3 people who actually make the decisions. This was set up when we joined the SFC. Also the 3 leaders need to be from different companies. * SFC has a requirement that organisation defines its leadership. What exactly the leadership is, it is up to organisation, but there should be some leadership". And that's how coreboot got leadership: coreboot wanted to join SFC, so needed to define leadership. * Which means, if flashrom wants to join SFC we need leadership too * Nico: In the past there wasn't much to decide, so there was no need for official leadership. * [Anastasia] We need someone to make decisions after a discussion on the mailing list if there isn't a clear direction. * [Edward] Some kind of election process, [Martin] organized the vote system for coreboot. * [Nico] There are currently two different projects in one repo, so if the leadership picks one, there might end up being two projects? * [Martinr] How are decisions made in the flashrom project? In this meeting, on the mailing list, something else? * [Nico] So far for smaller topics on IRC. For bigger topics, I suggest the mailing list. * [Thomas and Felix] Before deciding on leadership we need to talk more to each other. This is the first meeting we see each other and talk. We need to talk more about development, code, need to know more what we are doing. * Conclusion: let’s everyone think about it and continue discussion later. Topic added to Ongoing Items. * [Martin/Nico] Should we split flashrom into two separate projects? * Nico when I started, the project was for everyone, and stability was a priority. * Now we're trying to add the google fork, which has different objectives. * Edward: This isn't a different objective. * Quality is degrading. * Is code even tested? * Edward: yes, there's a strong emphasis on testing and improving tests. * Felix: what is tested? Is this unit testing? * Google tests any paths that they have available for testing. * Intel, AMD, Linux, mtd, Dediprog, raiden debug. * In the past, we kept all the code working for generic systems, so when support was added for a chip, it works for all of them, or bails out early. * New code is for specific boards, but not documented * [Martin] So let's focus on testing and make stability branches. * Nico: There's nobody to do that testing. * Felix: with the jenkins replacement, I'm thinking about how to add hardware tests as a part of the CI. * Miklos: There was someone who had boards with spi roms on boards with multiplexers for testing. \ Referred to this [document about the Chromium team’s hardware test setup.](https://docs.google.com/document/d/1IoBR7rHXJFiahC6dGnNKqaNFfg-q89_NMHuFyiJ7... ) * [Edward] Exactly this has come up between a discussion originally Martin and latter with Nikolai Artemiev to get CI coverage for writeprotect changes. * [Edward] We need to document testing what testing is done on the wiki. * Google is working on testing and making sure that the tests can be run by anyone. * [Edward]: First part of this is upstream-first unit-testing efforts, second part harder. * [Edward]: We need to get boards hooked for E2E tests, looking at what is happening for coreboot CI? * [Piotr] aklm on IRC just received commit rights without any previous discussion [[1]](https://matrix.to/#/!SAlNdlJiEjYGbpbVfu:libera.chat/$eKIApfMuImHqA9T0TowmUaT... ) * [Martinr] Anastasia has earned her commit rights with 125 commits over the past 2 years, more than anyone else. She is active in the community. She has done more than 82 reviews, more than anyone but Angel, Edward, and Nico. If she doesn't deserve her commit rights, Angel, Edward and Nico are the only ones who do. The discussion about commit rights is listed below, but as mentioned there, IMO, Thomas is the only other person who has earned commit rights. * [Nico] Edward’s history in the project is quite unique. He got the submit rights before he landed a single patch and with close to 0 prior activity in the project. So we can’t really put him above or below anyone in this matter. * We can absolutely look at commits and reviews that he's done since then. * [Nico] I don’t think we can decide this matter simply on the amount of patches / reviews. * [Martin] So what are your thoughts on how people should get +2? Is this also the "safe and sound" rule? You don't give any suggestions here. * [Piotr] How to get patch submitting rights in flashrom? We have 4 patches in +2 for some time: [[1]](https://review.coreboot.org/c/flashrom/+/55713/), [[2]](https://review.coreboot.org/c/flashrom/+/55714), [[3]](https://review.coreboot.org/c/flashrom/+/55714), [[4]](https://review.coreboot.org/c/flashrom/+/55848/) * Current Submitters * Anastasia Klimchuk * Angel Pons * Carl-Daniel Hailfinger * David Hendricks * Edward O'Callaghan * Idwer Vollering * Nico Huber * Patrick Georgi * Stefan T * Urja Rannikko * The only other person that has a significant number of commits in the last 5 years is Thomas Heijligen with 83 commits since July 2019. Thomas seems like a good candidate for getting submit rights. * What are the criteria for giving someone +2? When is someone ready, and who makes that decision? * coreboot has a relatively loose criteria for adding people, but decisions are made by the leadership. * Felix: Change the process about how patches are merged * Maybe require 2 +2s? * In case patches are merged too early, request that patches stay on gerrit for greater than 2 days or so? * [Carl-Daniel] We have different strategies, availability. Reviews are harder than writing code. We need to give people time to do reviews. Maybe we need to give 2 weeks time for reviewers. Unless some really bad bug which kills machines. * We have different modes of working on code, different goals. * Nico: 2 weeks is probably too long. * [Angel] too many patches sometimes, too little time. Very stressful. Patches pile up, not enough time to review and respond to everything. A feeling that if I miss something and do not respond in time then things will go wrong (e.g. errors missed during review). * [Anastasia] maybe if reviewer is busy, they can send a message to gerrit saying “sorry I am busy for next X weeks, I will get back to you after X weeks, thanks”, and this gives information to patch owner. Makes a big difference.
-
Felix Singer