# Merge Team Call - 2020-07-13 - 2021-03-08

# ***THIS IS AN OLD DOCUMENT*** — New one is here: https://hackmd.io/r9WOQntNR7iyVwn97q-QYw

- Older meetings:
- 0000-00-00 to 2020-07-06: https://hackmd.io/MofecNp_TZyexeWJy4NfTA

- Newer meetings:
- 2021-03-15 - ...: https://hackmd.io/r9WOQntNR7iyVwn97q-QYw

Where: https://us02web.zoom.us/j/88060393605?pwd=b0RhZ292Y2pnbFlpbWVERk1uNmQvQT09

- [MR status tracker - google sheet](https://docs.google.com/spreadsheets/d/1Vh05penfFX8dai9HTO-Su3OsCfbzmSd3J30VTUSugLY/edit)
Use this [script](https://gitlab.com/arvidnl/tezos_merge_team_scripts) to export all current open MR into a csv

- Helping scripts: https://gitlab.com/arvidnl/tezos_merge_team_scripts

- Current merge members: (Arthur), Arvid, Clément, Ilias, (Jun), Klakplok, Mehdi, Raphaël C., Raphaël P., Seb, Vbot, Pirbo, Romain, Yann, Marco
- Past merge members: Adrian, Sylvain, Tomáš
- merge dispatcher:
- 01 -> 08 feb: Arvid
- 08 -> 15 feb: YRG
- 15 -> 22 feb: Vincent
- 22 -> 01 mars: Pirbo
- 01 -> 08 mar: Raphaël Proust
- 08 -> 15 mar: Ilias
- 15 -> 22 mar: Clément
- 22 -> 29 mar: Mehdi
- 29 mar -> 5 avril : Raphaël Cauderlier

- Merge dispatcher role:
- Self-assign to https://gitlab.com/tezos/tezos/-/issues/1062
- Assign people to incoming MRs.
- Label MRs
- Ping people who have something to do (use “MR status tracker” or Arvid2000).
- Assign people to MRs that don't have an Assignee.
- Click "merge" on easy/reviewed ones.
- Most work should be just after the meeting (so that pinged people have time to react before the meeting).
- Prepare the meeting (fill-up this document with new agenda).
- Also prepare a list of MRs for reviewer outside the merge team (thursday's review parties)
- Good fit are medium-sized MRs, should be reviewed in less than 2h
- A way to build this list is to use the "ready for review party 🎉" label

## Future H-OOO-lidays

# ***THIS IS AN OLD DOCUMENT*** — New one is here: https://hackmd.io/r9WOQntNR7iyVwn97q-QYw

## 2021-03-08

- Merge-team membership: how much clean-up do we need? how much do we want? (Adrian is the owner of this here document, There are people still in the slack channel, still in the default approver group, etc.)
- We have 3 distinct python environments (test, doc, snoop), do we need to do anything about it? (Raphael P ping people)

MRs:

- Needs input from the merge team (or at least worth discussing)
- https://gitlab.com/tezos/tezos/-/merge_requests/2646 CI, Tezts → needed for other MRs' CIs

- Ready for merging (has approvals, or at least one approval and seems mostly good to go)
- https://gitlab.com/tezos/tezos/-/merge_requests/2651 (consensus heuristic, probably important enough to warrant a second review, Vincent?)
- https://gitlab.com/tezos/tezos/-/merge_requests/2661 (openAPI documentation of Edo entry-points and some missing shell entry-points)

- Needs review but otherwise looks good (with suggestion for reviewer)
- https://gitlab.com/tezos/tezos/-/merge_requests/2637 (worker-based scheduling of callbacks in chain-validator instead of `dont_wait`, Raphael P)
- https://gitlab.com/tezos/tezos/-/merge_requests/2026 (Swap lwtreslib and error-monad dependency, Yann)
- https://gitlab.com/tezos/tezos/-/merge_requests/2539 (use rust-libs instead of vendored ocaml-bls)
- https://gitlab.com/tezos/tezos/-/merge_requests/2667 (documentation improvements for how-to-get, Nic)

- Needs work from author, contributor, someone
- https://gitlab.com/tezos/tezos/-/merge_requests/2622 (documentation typo fixes)
- https://gitlab.com/tezos/tezos/-/merge_requests/2585 (refactoring in python tests, has threads open for Raphael C and Arvid)
- https://gitlab.com/tezos/tezos/-/merge_requests/1595 (bloom filter greylists, has threads open between Ilias and Yann)
- https://gitlab.com/tezos/tezos/-/merge_requests/2629 (snoop improvements)
- https://gitlab.com/tezos/tezos/-/merge_requests/1857 (generic multisig support in client)
- https://gitlab.com/tezos/tezos/-/merge_requests/2508 (RPC to dynamically reconfigure logging)
- https://gitlab.com/tezos/tezos/-/merge_requests/2666 (tezt parallelisation)
- https://gitlab.com/tezos/tezos/-/merge_requests/2670 (tezt removing old protocol tests)

A lot of logging work (NOTE: this needs to actually be reviewed, it does more than just swapping old for new logging mechanism, introduces new code, etc.):
- https://gitlab.com/tezos/tezos/-/merge_requests/2634
- https://gitlab.com/tezos/tezos/-/merge_requests/2672
- https://gitlab.com/tezos/tezos/-/merge_requests/2650
- https://gitlab.com/tezos/tezos/-/merge_requests/2663

## 2021-03-01

- `dune utop` magic patch: https://tezos-dev.slack.com/archives/GB0UR34N8/p1614261026060100
- metastate MR 2356: https://gitlab.com/tezos/tezos/-/merge_requests/2356/ "update migration test and test the migration between 007 and 008" -> keep or throw? split?
- discussion about the diversity of the merge team

MRs:

- Needs input from the merge team (or at least worth discussing)

- Ready for merging (has approvals, or at least one approval and seems mostly good to go)
- https://gitlab.com/tezos/tezos/-/merge_requests/2522 (logging, already approved by Arvid)
- https://gitlab.com/tezos/tezos/-/merge_requests/2641 (new `with_file` functions in stdlib_unix, approved by Raphael P, has thread open with Vincent)
- https://gitlab.com/tezos/tezos/-/merge_requests/2649 (Avoid wrapping when printing from get-contract-code)
- https://gitlab.com/tezos/tezos/-/merge_requests/2621 (fix storage of predecessor, already approved by Pierre)
- https://gitlab.com/tezos/tezos/-/merge_requests/2526 (preserve file for multiple tezt tests, already approved by Clement)

- Needs review but otherwise looks good (with suggestion for reviewer)
- https://gitlab.com/tezos/tezos/-/merge_requests/2624 (mockup support for all proto constants)
- https://gitlab.com/tezos/tezos/-/merge_requests/1595 (replace greylist with bloom filter)
- https://gitlab.com/tezos/tezos/-/merge_requests/2640 (prefix events with section names in logs)
- https://gitlab.com/tezos/tezos/-/merge_requests/2637 (Scheduling refactorisation in shell)
- https://gitlab.com/tezos/tezos/-/merge_requests/2598 (Update scripts to recent proto changes)
- https://gitlab.com/tezos/tezos/-/merge_requests/2656 (Remove proxy code for Carthage)

- Needs work from author, contributor, someone
- https://gitlab.com/tezos/tezos/-/merge_requests/2634 (logging in p2p, almost ready)
- https://gitlab.com/tezos/tezos/-/merge_requests/2539 (BLS packaging improvement)
- https://gitlab.com/tezos/tezos/-/merge_requests/2651 (consensus heuristic, preparation for further shell work)
- https://gitlab.com/tezos/tezos/-/merge_requests/2395 (Docker from CI for anyone, has threads opened, seemed nearly ready though)
- https://gitlab.com/tezos/tezos/-/merge_requests/2508 (reconfigure logging via RPC, nearly ready, nitpicky comments left over, ping Seb)
- https://gitlab.com/tezos/tezos/-/merge_requests/2585 (Factorisation in python tests)
- https://gitlab.com/tezos/tezos/-/merge_requests/2613 (new storage command for query/checks, nearly ready)
- https://gitlab.com/tezos/tezos/-/merge_requests/2506 (script to set up env)

## 2021-02-22

### Talking points

- Recall the semantics of assignee/reviewer
- Discussions about testing frameworks
- Prefer writing tests in OCaml frameworks instead of PyTest (for unit tests typically)
- White box testing vs black box testing

## 2021-02-15

### Talking points

- MRs from Metastate: should they be processed quickly?
- Nic with merging rights? => Merbocop

### MRs

MRs:

- proto MRs ?
- https://gitlab.com/tezos/tezos/-/merge_requests/2458 (baking accounts)
- https://gitlab.com/tezos/tezos/-/merge_requests/2602 (useless legacy flag, mergeable)
- https://gitlab.com/tezos/tezos/-/merge_requests/2352 (discussion in #modular-michelson)
- https://gitlab.com/tezos/tezos/-/merge_requests/2439 (benchmarking proto, efforts to be redirected)
- https://gitlab.com/tezos/tezos/-/merge_requests/2567 (Doc: add a page explaining the protocol naming convention.)
- https://gitlab.com/tezos/tezos/-/merge_requests/2593 (Add some comments related to the constants)

- Needs input from the merge team (or at least worth discussing)
- https://gitlab.com/tezos/tezos/-/merge_requests/2356 (update migration test, Metastate)
- https://gitlab.com/tezos/tezos/-/merge_requests/2381 (enable dune release mode)
- https://gitlab.com/tezos/tezos/-/merge_requests/2332 (PVSS client support, privacy team?)
- https://gitlab.com/tezos/tezos/-/merge_requests/2250 (CI: build on different ubuntu versions)
- https://gitlab.com/tezos/tezos/-/merge_requests/2455 (broken links in HTML doc)

- Let's discuss old MRs :
- https://gitlab.com/tezos/tezos/-/merge_requests/1988 (RPC to query ... @ AL Manning)
- some comments from vbot, both author and reviewer needs to respond
- Pietro stuff:
- https://gitlab.com/tezos/tezos/-/merge_requests/1918 (Consolidate ban)
- This is good to go right?
- https://gitlab.com/tezos/tezos/-/merge_requests/1915 (p2p: limit connections to 800 and add smother curve for limits)
- https://gitlab.com/tezos/tezos/-/merge_requests/1994
- I talked with Paul last week, he's on it but also busy with PhD
- https://gitlab.com/tezos/tezos/-/merge_requests/2159 (Build: make tezos-lmdb compile under FreeBSD (ocaml/opam-repository CI))
- Mehdi seems OK with it
- ... https://gitlab.com/tezos/tezos/-/merge_requests?draft=no&scope=all&sort=created_asc&state=opened&utf8=%E2%9C%93

- Ready for merging (has approvals, or at least one approval and seems mostly good to go)
- https://gitlab.com/tezos/tezos/-/merge_requests/2594 (Docs: Improve help on sandbox mode)
- https://gitlab.com/tezos/tezos/-/merge_requests/2562 (proxy_mode: use Irmin)
- Needs review but otherwise looks good (with suggestion for reviewer)
- https://gitlab.com/tezos/tezos/-/merge_requests/2324 (P2p: documenting P2p IO scheduler internal behaviour, Nic)
- https://gitlab.com/tezos/tezos/-/merge_requests/2230 (Stdlib/dropbox: specialise function to avoid dummy timeout)
- https://gitlab.com/tezos/tezos/-/merge_requests/2580 (lib_crypto : add hacl checks)
- https://gitlab.com/tezos/tezos/-/merge_requests/2569 (Codec: register N and Z)
- https://gitlab.com/tezos/tezos/-/merge_requests/2511 (shellcheck scripts)
- https://gitlab.com/tezos/tezos/-/merge_requests/2549 (fail early during CI tests)
- https://gitlab.com/tezos/tezos/-/merge_requests/2524 (remove legacy logging from signer-backends)
- https://gitlab.com/tezos/tezos/-/merge_requests/2509 (check that there are no references to alpha in proto doc)
- https://gitlab.com/tezos/tezos/-/merge_requests/2297 (FA1.2 client support)
- https://gitlab.com/tezos/tezos/-/merge_requests/2335 (new michelson interpreter, needs review)
- https://gitlab.com/tezos/tezos/-/merge_requests/2324 (documentation effort in P2p implem)
- https://gitlab.com/tezos/tezos/-/merge_requests/2526 (tezt can now use temp-dir across tests)
- I've reviewed but yet to test, also some new remarks
- https://gitlab.com/tezos/tezos/-/merge_requests/2411 (allow baker to fetch ops from remote server)
- https://gitlab.com/tezos/tezos/-/merge_requests/2447 (contract metadata support)
- https://gitlab.com/tezos/tezos/-/merge_requests/2369 (Add two new RPC to query the greylist tables)
- https://gitlab.com/tezos/tezos/-/merge_requests/2227 (OpenAPI : Test validity)
- https://gitlab.com/tezos/tezos/-/merge_requests/2260 (make: remove default opam repo)
- https://gitlab.com/tezos/tezos/-/merge_requests/2575 (Migrate Proxy Python tests to Tezt)

- Needs work from author, contributor, someone
- https://gitlab.com/tezos/tezos/-/merge_requests/2561 (Fix coverage job in CI)
- https://gitlab.com/tezos/tezos/-/merge_requests/2506 (add scripts/env.sh)
- https://gitlab.com/tezos/tezos/-/merge_requests/2601 (add README.md in scripts/)
- https://gitlab.com/tezos/tezos/-/merge_requests/2556 (add VSCode development support)
- https://gitlab.com/tezos/tezos/-/merge_requests/2426 (bootstrap rehaul)
- https://gitlab.com/tezos/tezos/-/merge_requests/2540 (update dep to lwt-canceler.0.3 to use new features)

- https://gitlab.com/tezos/tezos/-/merge_requests/1857 (multisig client support)

- On the shelf
- https://gitlab.com/tezos/tezos/-/merge_requests/2229 (OCaml 4.12 from Kate)

## 2021-02-08

(copied from last week, cleaned out merges)

### Talking points

- @arvidj: Scripts: https://gitlab.com/arvidnl/tezos-merge-team-scripts-v2
- @arvidj: MR description template ?
- a voir si c'est possible de mettre le template dans le repo ?
- RC: les liens pour ou sont les changelogs different ?
- La conclusion: on mets le template dans le repo.
- est-ce qu'on fait un one-liner pour dire aux gens de choisir un template?
- @pirbo: relax rules about merging doc
- techniquement, comment faire en GL ?
- RC: on est trop exigant, mais c'est mieux que merger n'importe quoi.
- solution: sortir un peu du doc (tutoriels? how to use?) dans un wiki ?
- conclusion selon arvid
- 1. mt memebers regarde l'etiquette doc de temps en temps et compte arpprove nic omme approve mt
- 2. merbocop qui change les regles à la volee
- next merge-dispatchers
- YRG
- pirbo: comment s'organiser pour florence ?
- snapshot ?
- branch ?
- juste le laisser tranquil dans proto_alpha tant qu'on trouve un hash ?

### MRs

MRs:

- proto MRs ?

- Needs input from the merge team (or at least worth discussing)
- https://gitlab.com/tezos/tezos/-/merge_requests/2332 (PVSS client support)
- https://gitlab.com/tezos/tezos/-/merge_requests/2250 (CI: build on different ubuntu versions)
- https://gitlab.com/tezos/tezos/-/merge_requests/2545 (group opam: redo of !2380 without the bug that made us revert it)

- Let's discuss old MRs :
- https://gitlab.com/tezos/tezos/-/merge_requests/1988 (RPC to query ... @ AL Manning)
- some comments from vbot, both author and reviewer needs to respond
- Pietro stuff:
- https://gitlab.com/tezos/tezos/-/merge_requests/1918 (Consolidate ban)
- This is good to go right?
- https://gitlab.com/tezos/tezos/-/merge_requests/1915 (p2p: limit connections to 800 and add smother curve for limits)
- Superseded by https://gitlab.com/tezos/tezos/-/merge_requests/2557 (Cap connections to 100) ?
- https://gitlab.com/tezos/tezos/-/merge_requests/1994
- I talked with Paul last week, he's on it but also busy with PhD
- https://gitlab.com/tezos/tezos/-/merge_requests/2159 (Build: make tezos-lmdb compile under FreeBSD (ocaml/opam-repository CI))
- Mehdi seems OK with it
- ... https://gitlab.com/tezos/tezos/-/merge_requests?draft=no&scope=all&sort=created_asc&state=opened&utf8=%E2%9C%93

- Ready for merging (has approvals, or at least one approval and seems mostly good to go)

- Needs review but otherwise looks good (with suggestion for reviewer)
- https://gitlab.com/tezos/tezos/-/merge_requests/2549 (fail early during CI tests)
- https://gitlab.com/tezos/tezos/-/merge_requests/2479 (rewrite some py tests into tezt)
- https://gitlab.com/tezos/tezos/-/merge_requests/2524 (remove legacy logging from signer-backends)
- https://gitlab.com/tezos/tezos/-/merge_requests/2509 (check that there are no references to alpha in proto doc)
- https://gitlab.com/tezos/tezos/-/merge_requests/2297 (FA1.2 client support)
- https://gitlab.com/tezos/tezos/-/merge_requests/2335 (new michelson interpreter, needs review)
- https://gitlab.com/tezos/tezos/-/merge_requests/2324 (documentation effort in P2p implem)

- Needs work from author, contributor, someone
- https://gitlab.com/tezos/tezos/-/merge_requests/2526 (tezt can now use temp-dir across tests)
- I've reviewed but yet to test, also some new remarks

- https://gitlab.com/tezos/tezos/-/merge_requests/2530 (run some forgotten tests)
- I've pinged Craig to fix outstanding threads
- https://gitlab.com/tezos/tezos/-/merge_requests/2556 (add VSCode development support)
- https://gitlab.com/tezos/tezos/-/merge_requests/2426 (bootstrap rehaul)
- https://gitlab.com/tezos/tezos/-/merge_requests/2540 (update dep to lwt-canceler.0.3 to use new features)

- https://gitlab.com/tezos/tezos/-/merge_requests/2511 (shellcheck scripts)
- https://gitlab.com/tezos/tezos/-/merge_requests/2411 (allow baker to fetch ops from remote server)
- mostly good, needs one more approval!
- https://gitlab.com/tezos/tezos/-/merge_requests/1857 (multisig client support)
- https://gitlab.com/tezos/tezos/-/merge_requests/2447 (contract metadata support)

## 2021-02-01

- topics
- Updating OCaml in env v2?
- blocked by !2021 (ocamlformat), coq-of-ocaml (ready)
- 4.11.1 is in Debian/Ubuntu, would simplify packaging
- opam-dune-lint also depends on 4.11
- need to look at the diff of the stdlib; compiler output probably safe wrt 4.09
- check that the order of folds in new versions preserve fold ordering, especially for data-encoding (4.12) => maybe subsumed by a testsuite in data-encoding to check for encoding stability for a large set of commonly used types.
- https://gitlab.com/tezos/tezos/-/merge_requests/1914 (p2p: add new rpc /config/p2p and /config/p2p/limits)
- reviewed by Raphaël Proust & Romain, yet no approval. Do we want it?
- why lib_base, why not p2p_services?
- wait for Romain?
- https://gitlab.com/tezos/tezos/-/merge_requests/2426 (bootstrap rehaul)
- François suggests that we need a better process for large changes. But splitting such changes introduces dead code.
- He suggests:
1. first submit TZIP-style doc describing changes to merge team for pre-approval (eg bootstrap rehaul; new mempool)
2. then split task in sub-MRs, some potentially introducing dead code (ok beprcause of 1.)
- Being tested, looks good, we're going to take this leap
- Sequentialize these big rehauls to prevent having to induce huge rebases; how to identity these tasks & organize?
- To reduce risk, can we make intermediate results with eg both mempools/bootstrap pipelines/... coexisting, the old one as a fallback
- other alternative: have two versions coexisting
- Big release of data-encoding: how do we vet the upgrade? in this case, in the future, in general?
- general question: how do we manage updates to critical dependencies of tezos (eg merge team involvement)
- tests pass when pinning, lots of benchmarks done
- regression tests? some work by Tomasz, not exhaustive
- registered encodings do not cover all potentially critical ones
- registered encodings correspond to data exposed to the user
- if we register more systematically, need to structure their namespace
- take all .mli's and take ground encoding values, automatically generate regression tests

MRs:

- proto MRs: https://hackmd.io/_tNRy4ioRRqchD8EbyU5qA

- Needs input from the merge team (or at least worth discussing)
- https://gitlab.com/tezos/tezos/-/merge_requests/2455 (fix some broken links in the doc, stalled because of a semi-related conversation)
- Conclusion: resolve python packaging later
- https://gitlab.com/tezos/tezos/-/merge_requests/2539 (use rust libs instead of vendored BLS, WE NEED TO BE CAREFUL ABOUT PERF REGRESSION!!)
- Conclusion: Wait
- https://gitlab.com/tezos/tezos/-/merge_requests/2181 (replace some py/flex tests with tezt, Romain started to review, we need to reattribute)
- Conclusion: Arvid will have a look

- Ready for merging (has approvals, or at least one approval and seems mostly good to go)
- https://gitlab.com/tezos/tezos/-/merge_requests/2102 (refactor and comment shell tests, needs one approval from default, Arvid)
- https://gitlab.com/tezos/tezos/-/merge_requests/2551 (add missing primitive in documentation)
- https://gitlab.com/tezos/tezos/-/merge_requests/2514 (add "mempool" to glossary)
- https://gitlab.com/tezos/tezos/-/merge_requests/2515 (improve glossary)
- https://gitlab.com/tezos/tezos/-/merge_requests/2530 (run some forgotten tests)

- Needs review but otherwise looks good (with suggestion for reviewer)
- https://gitlab.com/tezos/tezos/-/merge_requests/2536 (add README to tezt, RaphaelP)
- https://gitlab.com/tezos/tezos/-/merge_requests/2486 (improve documentaiton of `tezos-client man`)
- https://gitlab.com/tezos/tezos/-/merge_requests/2526 (tezt can now use temp-dir across tests)
- https://gitlab.com/tezos/tezos/-/merge_requests/2549 (fail early during CI tests)
- https://gitlab.com/tezos/tezos/-/merge_requests/2500 (mockup config is not ignored anymore, fix #938)
- https://gitlab.com/tezos/tezos/-/merge_requests/2479 (rewrite some py tests into tezt)
- https://gitlab.com/tezos/tezos/-/merge_requests/2535 (document logging levels/sinks to users, RaphaelP)
- https://gitlab.com/tezos/tezos/-/merge_requests/2332 (PVSS client support)
- https://gitlab.com/tezos/tezos/-/merge_requests/2524 (remove legacy logging from signer-backends)
- https://gitlab.com/tezos/tezos/-/merge_requests/2496 (don't stdout during micheline tests)
- https://gitlab.com/tezos/tezos/-/merge_requests/2509 (check that there are no references to alpha in proto doc)
- https://gitlab.com/tezos/tezos/-/merge_requests/2297 (FA1.2 client support)
- https://gitlab.com/tezos/tezos/-/merge_requests/2250 (CI: build on different ubuntu versions)

- Needs work from author, contributor, someone
- https://gitlab.com/tezos/tezos/-/merge_requests/2556 (add VSCode development support)
- https://gitlab.com/tezos/tezos/-/merge_requests/2426 (bootstrap rehaul)
- https://gitlab.com/tezos/tezos/-/merge_requests/2540 (update dep to lwt-canceler.0.3 to use new features)
- https://gitlab.com/tezos/tezos/-/merge_requests/2555 (gas consumption instrumentation, needs some rebasing and some reviews)
- https://gitlab.com/tezos/tezos/-/merge_requests/2545 (redo of !2380 without the bug that made us revert it)
- https://gitlab.com/tezos/tezos/-/merge_requests/2511 (shellcheck scripts)
- https://gitlab.com/tezos/tezos/-/merge_requests/2324 (documentation effort in P2p implem)
- https://gitlab.com/tezos/tezos/-/merge_requests/2411 (allow baker to fetch ops from remote server)
- mostly good, needs one more approval!
- https://gitlab.com/tezos/tezos/-/merge_requests/1857 (multisig client support)
- https://gitlab.com/tezos/tezos/-/merge_requests/2447 (contract metadata support)

## 2021-01-25

- Talking points
- Protocol depends on folding order [!2527](https://gitlab.com/tezos/tezos/-/merge_requests/2527)
- let's take this to fix the emergency
- Arvid2k https://arvidnl.gitlab.io/tezos-merge-team-scripts/
- Host under the tezos org? Let's not add burden to @pirbo
- We should probably mention the tool in the dev doc.
- Clément Hurlin (@smelc) from Tweag expressed interest in joining the merge team
- He is welcome!

#### MRs:

- Needs input
- Check that protocol doc does not refer alpha https://gitlab.com/tezos/tezos/-/merge_requests/2509
- Client support for contract metadata https://gitlab.com/tezos/tezos/-/merge_requests/2447
- Add versions to all binaries https://gitlab.com/tezos/tezos/-/merge_requests/2487
- Bytecode compilation https://gitlab.com/tezos/tezos/-/merge_requests/2506 and https://gitlab.com/tezos/tezos/-/merge_requests/2507
- Pietro's work on loging https://gitlab.com/tezos/tezos/-/merge_requests?scope=all&utf8=%E2%9C%93&state=opened&author_username=abate&search=logging

- Approved
- [Proto doc linking](https://gitlab.com/tezos/tezos/-/merge_requests/2345)
- [Mockup Carthage and test cleanup](https://gitlab.com/tezos/tezos/-/merge_requests/2502)
- [Consolidate ban/unban trust/untrust rpc calls in one / clarify the semantic of the greylist and whitelist/blacklist / change the semantic on unban addr to also unban the all associated peers.](https://gitlab.com/tezos/tezos/-/merge_requests/1918)

## 2021-01-18

- Talking points
- CI Problems: opam-check.sh
- → **2** problems timeout + actual test failure
- ⇒ we set it to allow to fail (Raph P), and @pirbo will fix it properly
- Discussion on moving Tezos Developer Dashboard (aka arvid2k)
- ⇒ Next Week
- Standardize Ocaml comment/doc line length: https://gitlab.com/tezos/tezos/-/issues/1066
- New version of OCamlFormat first ([MR](https://gitlab.com/tezos/tezos/-/merge_requests/2021) open Raph P)
- Consistent use `Assignee` Vs `Reviewer`:
- Merge-dispatcher: assign 2 merge-teamsterates per MR.
- Review-field: used to delegate to other devs.
- Let merbocop issue a warning whenever the diff touches/creates `.*opam`|`dune`
- ⇒ https://gitlab.com/smondet/merbocop/-/issues/20
- Feb 01 Merge Dispatcher.

#### MRs:

- Needs input from the merge team (or at least worth discussing)
- https://gitlab.com/tezos/tezos/-/merge_requests/2494 (URGENT! fix opam packaging for mockup tests, says it FIXES MASTER
- ilias: the code for the mockup test MR needs to be refactored, this MR will be closed
- https://gitlab.com/tezos/tezos/-/merge_requests/2345 (Doc linking of multiple protocol versions, Raphaël C: what's the status?)
- https://gitlab.com/tezos/tezos/-/merge_requests/2381 (Use `release` mode in dune by default for performance)
- https://gitlab.com/tezos/tezos/-/merge_requests/2442 (Enable `-O3` and `--flambda` by default for performance)

- Ready for merging (has approvals)
- https://gitlab.com/tezos/tezos/-/merge_requests/2199 (Add regression tests for protocol RPCs)
- https://gitlab.com/tezos/tezos/-/merge_requests/2425 (Add test for nonce revelation)
- https://gitlab.com/tezos/tezos/-/merge_requests/2380 (Group opam jobs by batches of 8 to improve CI)
- https://gitlab.com/tezos/tezos/-/merge_requests/2461 (rename some functions in the context for compatibility with upcoming irmin/storage)

- Needs review but otherwise looks good (with suggestion for reviewer)
- https://gitlab.com/tezos/tezos/-/merge_requests/2432 (Use MR labels to control CI optional stages, Reviewers: Arvid, Mehdi)
- https://gitlab.com/tezos/tezos/-/merge_requests/2332 (Add `pvss` command (`pvss gen`, `pvss list`, etc.) to the client, suggestion: Raphaël P)
- https://gitlab.com/tezos/tezos/-/merge_requests/2440 (Modernise the storage/context API to match upcoming irmin release, suggestion: Raphaël P, Vincent, Mehdi)
- https://gitlab.com/tezos/tezos/-/merge_requests/2054 (Add `Lwtreslib.{Option,Result,Unsafe}`, suggestion: Mehdi)
- https://gitlab.com/tezos/tezos/-/merge_requests/2488 (Bump irmin version, suggestion: Pierre, Vincent)
- https://gitlab.com/tezos/tezos/-/merge_requests/2496 (Remove some printf statement in some tests, small and easy, suggestion: Seb)
- https://gitlab.com/tezos/tezos/-/merge_requests/2493 (Tweak crypto tests build rules for macOS compatibility, suggestion: Pierre)
- https://gitlab.com/tezos/tezos/-/merge_requests/2479 (Port proxy-mode tests from python to tezt, suggestion: Arvid)
- https://gitlab.com/tezos/tezos/-/merge_requests/2487 (Add `--version` flag to all binaries, suggestion: Raphaël P, Seb)
- https://gitlab.com/tezos/tezos/-/merge_requests/2411 (Allow fetching mempools from a remote server)
- review ongoing by Tomasz, needs feedback from vbot
- https://gitlab.com/tezos/tezos/-/merge_requests/2480 (Uniform method to activate protocol in python tests)
- https://gitlab.com/tezos/tezos/-/merge_requests/2080 (Fix typo (significant digits in rounding) in doc)
- https://gitlab.com/tezos/tezos/-/merge_requests/2297 (Support FA1.2 in client, suggestion: Arvid, Raphaël C)

- Non-Tezos MRs (resto, data-encoding, etc.)

- Needs work from author, contributor, someone
- https://gitlab.com/tezos/tezos/-/merge_requests/2497 (Use non-`_repr` modules outside of the protocol to enforce abstraction barrier, opened today)
- https://gitlab.com/tezos/tezos/-/merge_requests/2405 (Add `vendors/README` for in-repo dev documentation)
- https://gitlab.com/tezos/tezos/-/merge_requests/2472 (Add (manual) script for "How to get" instructions)
- https://gitlab.com/tezos/tezos/-/merge_requests/2492 (Code quality improvements to python test, prepares for a multi-protocol test maybe)
- https://gitlab.com/tezos/tezos/-/merge_requests/2486 (tezos-client doc improvement)
- https://gitlab.com/tezos/tezos/-/merge_requests/2395 (Allow docker for any branch)
- https://gitlab.com/tezos/tezos/-/merge_requests/2353 (Use `unsafe` when it's safe in crypto)

## 2021-01-11

- Talking points:
- the return of the opam solver timeout (following mirage-crypto downgrade); needs fixing because it slows down everything; hacky fix https://gitlab.com/tezos/tezos/-/merge_requests/2468 (Seb to make a solution to the problem based on opam switch importing, Raphaël P to review)
- List of MRs
- Arvid2000: https://arvidnl.gitlab.io/tezos-merge-team-scripts/

- Ready for merging (has approvals)
- https://gitlab.com/tezos/tezos/-/merge_requests/2415 (coq_of_ocaml tweaks, needs rebasing)

- Needs review but otherwise looks good (with suggestion for reviewer)
- https://gitlab.com/tezos/tezos/-/merge_requests/2422 (improve error management for sapling lib, this is blocking further MRs, suggestion: Marco)
- https://gitlab.com/tezos/tezos/-/merge_requests/2459 (documentation support for multiple proto version, reviewer suggestion: Raphaël C, Arvid)
- https://gitlab.com/tezos/tezos/-/merge_requests/2403 (split documentation for proto-specific and proto-independant commands, suggestion: Raphaël C)
- https://gitlab.com/tezos/tezos/-/merge_requests/2348 (big p2p MR, suggestion: Ilias)
- https://gitlab.com/tezos/tezos/-/merge_requests/2353 (avoid some copies by using `Bytes.unsafe_{to,of}_strings` in places where it is safe, suggestion: Mehdi, Yann)
- https://gitlab.com/tezos/tezos/-/merge_requests/2455 (fix broken links in documentation, update poetry lock, suggestion: )
- https://gitlab.com/tezos/tezos/-/merge_requests/2456 (refactor some code in tezt, suggestion: Arvid, Romain)
- https://gitlab.com/tezos/tezos/-/merge_requests/2343 (catch Unix error when spawning validator, suggestion: Vincent)
- https://gitlab.com/tezos/tezos/-/merge_requests/2346 (delay start of prevalidator until bootstrapped, suggestion: Raphaël P, Yann)
- https://gitlab.com/tezos/tezos/-/merge_requests/2446 (extend mempool plugins for doing protocol RPCs)

- Needs input from the merge team (or at least worth discussing)
- https://gitlab.com/tezos/tezos/-/merge_requests/2236 (only ask for password three times, very old MR, maybe stale, the MR seem to do more than just what it says on the tin)
- https://gitlab.com/tezos/tezos/-/merge_requests/2442 and https://gitlab.com/tezos/tezos/-/merge_requests/2381 (use flambda and `-O3`, and use `release` profile for dune, needs input from Pierre mostly)

- Non-Tezos MRs (resto, data-encoding, etc.)

- Needs work from author, contributor, someone
- https://gitlab.com/tezos/tezos/-/merge_requests/2345 (scripts to automate multo-proto-version documentation)
- https://gitlab.com/tezos/tezos/-/merge_requests/2461 (some renaming of Context functions)
- https://gitlab.com/tezos/tezos/-/merge_requests/2199 (additional regression tests using tezt)
- https://gitlab.com/tezos/tezos/-/merge_requests/2443 (documentation about the voting periods)
- https://gitlab.com/tezos/tezos/-/merge_requests/2380 (Groups some opam jobs together)

## 2021-01-04

- Ready for merging (has approvals)
- https://gitlab.com/tezos/tezos/-/merge_requests/2377 (mostly doc)
- Needs review but otherwise looks good (with suggestion for reviewer)
- https://gitlab.com/tezos/tezos/-/merge_requests/2376 (small, review by Vincent)
- https://gitlab.com/tezos/tezos/-/merge_requests/2421 (documentation of sapling protocol, review by Marco)
- https://gitlab.com/tezos/tezos/-/merge_requests/2427 (documentation, small, review by Raphaël P)
- https://gitlab.com/tezos/tezos/-/merge_requests/2425 (python test for nonce reveal, review by anyone or by default Arvid and Eugen)
- Needs input from the merge team (or at least worth discussing)
- https://gitlab.com/tezos/tezos/-/merge_requests/2360 (and https://gitlab.com/tezos/opam-repository/-/merge_requests/132) (externalisation of lwt-exit)
- https://gitlab.com/tezos/tezos/-/merge_requests/2239 (use chunked transfer encoding for large RPCs, still early WIP, will make significant changes, needs benchmarking)
- https://gitlab.com/tezos/tezos/-/merge_requests/2329 https://gitlab.com/tezos/tezos/-/merge_requests/2330 (protocol: gas arithmetic, non-trivial change that needs to be done right)
- https://gitlab.com/tezos/tezos/-/merge_requests/2335 https://gitlab.com/tezos/tezos/-/merge_requests/2374 (protocol: Michelson interpretation, non-trivial change that must be semantically correct)
- Non-Tezos MRs (resto, data-encoding, etc.)
- https://gitlab.com/nomadic-labs/resto/-/merge_requests/24 (local mode)
- https://gitlab.com/nomadic-labs/lwt-canceler/-/merge_requests/1
- Needs work from author, contributor, someone
- https://gitlab.com/tezos/tezos/-/merge_requests/2345 (documentation of protocols)
- https://gitlab.com/tezos/tezos/-/merge_requests/2314 (release with static assets, ping Pierre Boutiller)
- https://gitlab.com/tezos/tezos/-/merge_requests/2375 (less parallelism in CI)
- https://gitlab.com/tezos/tezos/-/merge_requests/2422 (improvements in error management in Sapling, ping Marco)
- https://gitlab.com/tezos/tezos/-/merge_requests/2315 (test for migration using mockup mode)
- https://gitlab.com/tezos/tezos/-/merge_requests/2297 (FA1.2 support in client)
- https://gitlab.com/tezos/tezos/-/merge_requests/2084 (multisig CLI support)
- https://gitlab.com/tezos/tezos/-/merge_requests/2266 (protocol testing doc)

- Announcement from Metastate
- Some more background: https://medium.com/metastatedev/update-on-metastates-plans-as-the-second-tezos-core-development-team-17741aa9ec47
- Feel free to message me (Adrian) if you have any questions
- Also, hoping that we can have lunch in Paris in summer if covid allows
- Welcome Ilias and Yann
- Regression with 8.0
- a lot of endorsement misses on the network as a whole
- Discussion around the migration time
- Storage flattening
- Sapling breaks the invariant that all directories are small

## 2020-12-14

- Merged:
- Lwtreslib
- Fake baking
- Not merged:
- empty traces
- ocamlformat
- proxy
- snoop
- Growing merge team
- Mehdi wrote:
> What would you think about growing the merge team again?
- Yann happy to join if not eaten by that
- Ilias too for client/snoop
- Approval/merge rules
- Pirbo wrote:
> For reasons, "everybody" is "developper" in tezos/tezos and therefore has the right to merge so we encode the 1 approuver 1 merger workflow by the 2 approuver one but the second one is still moraly the merger :wink: By the way, I would like to rediscuss this organisation this afternoon during the meeting :slightly_smiling_face:
- Mehdi wrote:
> To improve @Arvid Jakobsson's wonderful job, I'd like a page where you can easily identify the actions you have to take.
> I propose this algorithm: https://codimd.nomadic-labs.com/QMmLws_jRkWh1QU-65w6OA?edit comments are welcome!
> Should we move tezos-merge-team-scripts to gitlab.com/tezos?
- Arvid: making tools are always fun, but one should always be wary of trying to solve social/organisational issues with technological solutions (sorry if I repeat this often :P)
- Edo fix
- What's left?
- Normalize RPC for 009+: https://gitlab.com/tezos/tezos/-/merge_requests/2354
- Rafoo on it for Edo, use Vincent/Benjamin work on mempool plugins
- TZIP will accept protocol RFCs
- Protocol development still blocked?
- Python test duplication merged
- Opened MR: duplicating doc
- Proto MRs not blocked, but
- new features require docum
- MRs to discuss
- https://gitlab.com/tezos/tezos/-/merge_requests/1988 (Adrian, vbot)
- Spoke to vbot on slack :-)

## 2020-12-07

- Protocol RFCs
- There seems to be a consensus among protocol devs and protocol users about the need to move at least Michelson feature proposals to a more formal, RFC-based, repo. Sophia Gold has launched an initiative https://gitlab.com/future-design/rfcs. Seb has suggested to host it under the tezos Gitlab organisation.
- We agree to host the repo. As a organisation owner, Pirbo can help launch this. Raphael C. agrees to be one of the maintainers.

(message from Romain, MD of last week: here is a summary of MRs from last week)

Merged :
- !2106 (python test lint)
- !1927 (RPC DELETE /network/greylist)
- !2318 (proto doc fixes)
- !2288 (parameterize Tezt tests by the protocol)
- !2338 (split make fmt)

To merge:
- !2245 (lwtreslib)
- should be merged quickly: it's ready and is annoying to rebase
- I did a full review and approved it
- needs one more approval - Mehdi is on it
- includes !1967 which can thus be closed once !2245 is merged
- !2162 + !2300 (snoop improvements)
- I think we said those could be merged blindly
- !1943 (proxy mode)
- approved by Ilias
- has some Tezt tests than will need to be parameterized by the protocol
now that !2288 is merged
- !2096 (mockup fake baking)
- already has 4 approvals, including mine
- may need to fix conflict because !2288 was merged
- !2021 (ocamlformat)
- should be merged last of the big MRs, because it is not hard to rebase
(one just needs to re-run ocamlformat)
- don't forget to remove the DO NOT MERGE commit
- depends on https://gitlab.com/tezos/opam-repository/-/merge_requests/108
which has many open threads (I pinged Raphaël)

## 2020-11-30 “Like a Release Monday”

- **Release** status update.
- Should be done within the hour after the MTM.
- **Proto-proposal → master** status update.
- All almost ready.
- One problem with python tests: WIP.
- Should be able to merge today.
- Big MRs coming in.
- CI improvements.
- Next `MD`.
- merge order:
- proto-proposal into master => first
- python test lint
- lwtreslib
- interop with JS => @pirbo
- snoop
- mockup
- proxy
- ocamlformat

## 2020-11-23 « À L'Arrache »

- Wednesday: 8.0rc2
- Compatibility with OCaml 4.10.x
- New warning: unused functor argument.
- (Only) Protocols are raising that warning.
- Just replacing `Arg` with `_` is OK.
- Protocol development in tezos/tezos:master
- Documentation and tests.
- https://hackmd.io/NdvnEAOPRVuSdSqggY0PoA - a section summarizes the discussion that happened today

## 2020-11-16

- MRs for proto-008
- don't merge changes to the protocol in master w/o changing proto-proposal ==> let's merge proto-proposal in master
- development of the protocol to be done in `master` (still in `proto_alpha`)?
- need to convince ourselves that it does not harm decentralization / cannot be used against us
- would alleviate the need for complicated rebases
- people who want to contribute a new protocol would have an easier time contributing to `proto_alpha` actually
- people who want to contribute a completely different protocol would have to fork / create a new `proto_beta` directory though
- discussion to be continued in the protocall, ideally with Perry and Arthur
- development version of protocol environment would need to be clearly identified as such
- name the modules `V2dev`, `V3dev` etc.
- or use an odd-even rule
- less self-documenting though
- would have to take this decision right now, because Edo needs v1 very soon
- or have a built-in list of stable versions and refuse to run them on Mainnet
- Tests on `master` that are not reliable?
- Next MD (Merge Dispatcher, not MarkDown) => Pierre
- @arvid: https://github.com/smarkets/marge-bot ?
- advantages: automatic and we can space out merges to reduce pressure on CI
- to be noted: Gitlab finally working on it https://gitlab.com/groups/gitlab-org/-/epics/4911#note_446967572, so don't put too much effort on it

## 2020-11-09

- New store
- to be merged at the same time as the context irmin improvements (that require conversion too)
- possible double-migration?
- still need to bench and monitor
- RAM issue killing the bootstrap
- takes more RAM than the previous store (but is faster)
- needs more tests
- wait a bit before review
- needs a couple of weeks
- OCaml 4.11?
- implies big ocamlformat patch
- would be useful to debug RAM issues
- could just remove ocamlformat temporarily?
- wait for big MRs to be merged first
- [Tests of lwtreslib](https://gitlab.com/tezos/tezos/-/merge_requests/2131)
- discuss merging the yes-node and proto-proposal
- currently a patch
- especially useful to test migration on Mainnet
- merging this may be risky
- what if a user uses it by mistake?
- what if the code is buggy and runs even when we don't ask it to?
- what if a refactoring makes this active by mistake?
- one people running a yes node is not an issue but many people is
- prompt to confirm?
- not automatable (that's the point, but the point of merging is also to automate)
- question is: how to mitigate the risks
- implement as another dune target?
- node that self-tests whether it can refuse a signature
- copyright script now capable of updating the files
- https://gitlab.com/tezos/tezos/-/merge_requests/2120

## 2020-11-02
- merbocop updates to stop complaining about changes in proto tests (thanks RaphaelP); before relaunching the bot, please add issues if any <https://gitlab.com/smondet/merbocop/-/issues>
- [!1988](https://gitlab.com/tezos/tezos/-/merge_requests/1988)
- to address new comments

## 2020-10-26

Discuss:

- Arvid: Ping everyone on #devteam regarding outstanding threads?
- Yes for now, bot for later
- Update on node(s) with operation hash indexing? ([!1988](https://gitlab.com/tezos/tezos/-/merge_requests/1988))
- Adrian logging back in … → still sinc'ing archive-nodes. :-(

28G     /home/ubuntu/tezos-mainnet/context/index
60G     /home/ubuntu/tezos-mainnet/context
24G     /home/ubuntu/tezos-mainnet/store
84G     /home/ubuntu/tezos-mainnet
https://tzstats.com/BLmB6kEEQxcvkDVjgWDKu55cAzySG4wVqeeUNGvSjRwaedKLo6W

- [`!2182: Add tezos-packaging to the how-to-get instructions`](https://gitlab.com/tezos/tezos/-/merge_requests/2182)
- Pierre B: Pierre B @ NL is working on something similar.
- These are static binaries, but that is ok
- they (serokell) are payed by TQ to do this
- should be upstreamed to tezos/tezos instead of linking to serokell
- lack of communication in the ecosystem leads to redundancies
- looks weird to have an external ppa
- if it is not in tezos/tezos then it should be made more clear that the static binaries are by "another company". who is reponsible etc.
- in general good, just some pratical concerns.
- how to test if it is good work?
- Pierre deals with it.
- Summary of what to say:
- Good job!
- We want to have this in tezos/tezos-packaging
- Otherwise, the documentation must be made clearer that these packages are external.
- [`!2184: Draft: Add documentation guidelines`](https://gitlab.com/tezos/tezos/-/merge_requests/2184)
- Add something to MR template ?
- RC, Romain is for. Noone against.
- Can we check automatically? By ocamldoc? Can be a bit complicated. Maybe in a manual job. Give difference between number of undocumented functions in target branch / source branch. Maybe using the website (developer docs) to store the number?
- Add a badge.
- [ `!2024: split lib_base in smaller libraries` ](https://gitlab.com/tezos/tezos/-/merge_requests/2024)
- could be interesting to split stuff, but some motivation is lacking.
- the dependency graph needs a review
- having more split between pure / unixy things would be useful for e.g. js-of-ocaml.
- vbot: agree with the mr / and what has to be done. but why now?
- pb: belongs to the js effort. so it makes sense and it is a priority
- vbot: we need it but the MR lacks context & motivation.
- pb: maybe merging this stuff would be inconsitent with teh more recent effort to compile to js. ask raphael/pietro his opinion.
- Good start, but with some issues currently. Pietro should work with Raphaël to package this correctly. Check how it relates to more recent JS work. The MR lacks motivation and context.

## 2020-10-19

- It looks like under some conditions, some non-merge-team members can merge.
- Nomadic-Labs group had “developer” status → back to “reporter”.
- Update on node(s) with operation hash indexing? ([!1988](https://gitlab.com/tezos/tezos/-/merge_requests/1988))
- OK, it's optional, we _want_ this.
- Adrian: 1st approver, then VBot merger.
- [x] Need volunteers for _merge-dispatcher_ for incoming weeks.
- statsmemprof tool from JaneSt
- Events most usage
- More experiments
- Review Party Fuel:
- [`!1595`](https://gitlab.com/tezos/tezos/-/merge_requests/1595): **Replace IP greylist structure by a generational Bloom filter**
- Opened 7 months ago by Benjamin Canou
- Arvid has done a lot of reviewing but wants more eyes.
- Ben C to address comments.
- No integration tests (was already like this).
- Need to find a Bloom-filter expert for extra-review.
- [`!2056`](https://gitlab.com/tezos/tezos/-/merge_requests/2056): **Client: add import secret key from mnemonic**
- Opened 1 month ago by Alexandre Doussot
- 7 participants, *no consensus* :)
- couple of pushes since last comments
- Need documentation about this algorithm Vs the ledger one.
- Ben C to do next review.
- [`!2044`](https://gitlab.com/tezos/tezos/-/merge_requests/2044): **node,client: add env vars for -base-dir and -data-dir**
- Opened 1 month ago by pietro👷
- Assignee: Romain
- Easy in itself but pretty fundamental change to node setup worth discussing (?)
- OK → Romain to push Pietro to fix help message.
- [`!2034`](https://gitlab.com/tezos/tezos/-/merge_requests/2034) (WIP) claims to be an alternative to [`!1620`](https://gitlab.com/tezos/tezos/-/merge_requests/1620) (7 months old) should we **close?**
- 2034 indeed more flexible
- Romain to close 1620 with message about 2034
- **@all**: ask people who manage public nodes to give opinions on !2034
- [`!1838`](https://gitlab.com/tezos/tezos/-/merge_requests/1838): **lib_shell: Factorize the timestamp drifting check**
- Opened 4 months ago by François Thiré
- stuck a month ago.
- → Ping Raphaël P.
- [`!1914`](https://gitlab.com/tezos/tezos/-/merge_requests/1914): **p2p: add new rpc /config/p2p and /config/p2p/limits**
- Opened 4 months ago by pietro
- 4 Assignees, no approval, comments seem answered
- → Romain
- [`!1927`](https://gitlab.com/tezos/tezos/-/merge_requests/1927): **p2p: new RPC DELETE /network/greylist**
- Opened 3 months ago by pietro
- → P2P Meta-Coordination
- [`!1994`](https://gitlab.com/tezos/tezos/-/merge_requests/1994): **Improve node configuration validation support**
- Opened 2 months ago by Paul Laforgue
- Kinda a stuck a couple of weeks ago.
- → Romain

## 2020-10-12

- Memory requirements of the node:
- Now 8 GB for bootstrapping, then 4 GB is fine for running.
- Should be documented.
- Result of the operation hashes RPC experiment: “They keep crashing :arrow_upper_left: ”
- To be published on the slack channel in the coming days
- Status of the test documentation MRs.
- Is Jun's "Context reconstruction resume" already in the new store? Should we close MR [!2103](https://gitlab.com/tezos/tezos/-/merge_requests/2103) ?
- Yes useful, but needs harmonization of UX with new store.
- Assignee: **@VBot**.
- Release: is anything still blocking the release?
- Romain: we're ready, making branch and filling `CHANGES.md` today/tomorrow
- Francois patch for next subversion
- Precise the role of the merge dispatcher: to put up there :arrow_up:
- How to motivate people to do reviews?
- Arvid has a script to let people auto-evaluate the ratio between reviews they request ans review they make.
- RC: How much time should it take?
- Pierre B: should be a full-time job for a week
- Arvid: maybe a lot for the first day and then about one hour every morning

## 2020-10-05

- Moved back to 15:00 CEST for next week.
- Operation hashes table/RPC: currently running experiments → results next week.
- New store:
- Taking into account first of reviews.
- Needs reviewers for 2nd round.
- Next release:
- Need: fix for warning for incorrect protocol hash.
- Need: bootstrapping problems (adding seeded hash-tables to avoid resource-consumption attacks) → we can revert the problematic commit but would like better fix.
- WIP: Bootstrap heuristic imporvements.
- Need: `tezos-codec` bug; duplicate encoding name for an error in Delphi.
- Pietro's log level changes.
- Opam solver non-determinism problems in CI.
- Romain's Copyright script (cf. `headache` tool).

## 2020-09-28

- Pushed from last week:
- (Seb asked:) Client code duplication (impacts rebasability/mergeability of proto-propsal + maintenance):
- example of symlink: https://gitlab.com/smondet/tezos/-/merge_requests/7
- Will try with dune's `copy` stanza instead of symlink.
- We'll come back to it if someone comes up with better idea.
- (Danny asked:) Can we run the test coverage when we push automatically on master? (And can we also add a small link to the results in the doc/README.md of tezos/tezos?)
- Yes we can
- Run coverage every 24 hours with GitLab schedulers
- Cannot make CI slower
- Propose something that is parallisable and does not slow down the CI
- Copyrights
- maybe make a script that looks at commits to find out if some are missing?
- Bootstrap heuristic
- https://gitlab.com/tezos/tezos/-/merge_requests/2083
- Prioritise merging this!!!
- Review this, this needs to be merged before v8
- not trivial and we should be happy with it
- Context reconstruction
- https://gitlab.com/tezos/tezos/-/merge_requests/2103/
- merge this now or close it and wait for the new store
- https://gitlab.com/tezos/tezos/-/merge_requests/1512
- backport from the new store into the Jun's patch
- keep the same interface at the command line
- Discussion on store
- Old store, new store, Plebeia
- Talk to Jun about it next week
- Operation hash to operation
- https://gitlab.com/tezos/tezos/-/merge_requests/1988
- Check the impact on the size of the store
- MRs that can be looked at:
- Mmenomic import: https://gitlab.com/tezos/tezos/-/merge_requests/2056
- Vbot
- Exit codes: https://gitlab.com/tezos/tezos/-/merge_requests/2092
- Non-terminal sequence: https://gitlab.com/tezos/tezos/-/merge_requests/1678
- Raphael C.
- RPC calls for peers: https://gitlab.com/tezos/tezos/-/merge_requests/1918
- Update copyright years: https://gitlab.com/tezos/tezos/-/issues/923#note_419499822

## 2020-09-22

- Metaquestion
- How to better align shell and protocol development?
- currently too many merge conflicts, in particular in regression tests
- also the problem of backporting protocols to latest releases: should the protocol be based on latest-release (maybe + new environment)?
- Conclusion: start by tweaking the current approach as follows.
- Be a little more discerning while writing new tests as to whether they should go in master or proto-proposal.
- Think a bit more about when we do rebases, this could be done by rebasing on top of a more stable prefix to master.
- Work on the CI file and test framework so that we can run the script more quickly.
- (seb asks:) Delphi in `master` timeline.
- We should merge @Pirbo's MR **now**.

## 2020-09-14

- Arvid: Discussion about testing and testing guidelines
- pirbo: difference shell / protocol
- discussion should be more nuanced
- more testing in the protocol? -> good
- in the shell, not so clear. because harder to test. and this meeting is more about the shell.
- 100% coverage makes no sense in OCaml
- testing also has a price.
- current testing guidelines (and the merge request description template) already goes a long way
- no relevance in automatic enforcement of coverage
- move discussion to proto-call
- Discussion about CI
- two orthogonal limiting factors: (1) how fast jobs run in the pipeline, and (2) the number of jobs launched in parallel
- (1) work has been done to reduce the size of downloads in pipelines by factor 5
- (2) however, number of jobs is also a limiting factor. testing especially launches many jobs
- limit: <= 500 jobs in the pipeline.
- typical tezos pipeline has 100 jobs.
- 25 unit: tests in .gitlab-ci.yml
- 42 integration: tests in .gitlab-ci.yml
- \_opam pipelines double the no of jobs. 120 opam packages -> each one is a job.
- we can merge tests, but the script for updating the .gitlab-ci.yml must still work. no manual work.
- can't throw money at the problem (would have to run own gitlab instance)
- questions:
- how many pipelines are \_opam?
- how many pipelines are typically run in parallel?
- Next week's merge coordinator: Adrian

## 2020-08-31 -> moved to 2020-09-07

- Last week Merge-Team-Dispatcher: Sylvain (@sribaroud)
- Discussion on our reflections on how to improve the merge process.
- improve Arvid's python script to easily update the google doc
- Activate merge-train ? https://docs.gitlab.com/ee/ci/merge_request_pipelines/pipelines_for_merged_results/merge_trains/
- not working for cross repository MRS
- maybe a new test to check if still true
- only works for non fast-forward MR
- https://gitlab.com/gitlab-org/gitlab/-/issues/35628
- thumbs up vs approuval vs LGTM
- incrementing approuval counter
- thumbs up does not modify history of MR, where we would like to have
- improved gitlab MR search, now we can check that reaction (approuved,...)
- https://gitlab.com/tezos/tezos/-/merge_requests?scope=all&utf8=✓&state=opened&approved_by_usernames[]=None
- https://gitlab.com/tezos/tezos/-/merge_requests?scope=all&utf8=✓&state=opened&approved_by_usernames[]=Any
- should we block MR that contains fix-up ?
- no, should be rebased locally
- maybe block such MR
- @arvid: Discussion on how to improve interaction between merge team / review party
- a list of MR that is synced between merge team and review party with label `to be reviewed` (or `ready for review` (?))
- MRs:
- https://gitlab.com/tezos/tezos/-/merge_requests/2000 (shell: demote Terminated and Started events to Info level)
- Maybe breaking change for tools using log level ?
- independant levels for stdout and rest ? (might be too complicated)
- needs longer discussion
- https://gitlab.com/tezos/tezos/-/merge_requests/1915 (p2p: limit connections to 800 and add smother curve for limits)
- long standing issue (#?) to check that the config is not too crazy.
- Add a config checker that check for any iconsistency.
- there is already an MR for that
- The trade-off is work for another MR
- ask the P2P team about ther opinion

- https://gitlab.com/tezos/tezos/-/merge_requests/2032 (Add a reset protocol)
- usecase :
- help to reset a test net (Proto_alpha -> Reset protocol -> Proto_alpha')
- maybe protocol0 can replace that reset protocol ?
- if we can vote for genesis this MR is not useful

## 2020-08-24

Merge-Team-Dispatcher: `@vbot`

Discussion: how do we get on top of 5-6 pages or MRs?

@arvidj MR status tracker:
https://docs.google.com/spreadsheets/d/1Vh05penfFX8dai9HTO-Su3OsCfbzmSd3J30VTUSugLY/edit

MRs:
- https://gitlab.com/tezos/tezos/-/merge_requests/2021
- ocamlformat.0.15 ?
- https://gitlab.com/tezos/tezos/-/merge_requests/1450
- Automatic identity file generation?
- https://gitlab.com/tezos/tezos/-/merge_requests/1918
- P2P whitelist/blacklist semantics rework
- https://gitlab.com/tezos/tezos/-/merge_requests/1994
- Node config sanity check
- https://gitlab.com/tezos/tezos/-/merge_requests/1907
- OpenAPI/Swagger RPC doc generation
- Public test server?
- https://gitlab.com/tezos/tezos/-/merge_requests/2018
- Distill generic part from error monad

## 2020-08-17

MRs:
- https://gitlab.com/tezos/tezos/-/merge_requests/2024
- refactoring lib_base into smaller libraries
- needs discussion
- overcomplicated
- p2p should not be in lib_base
- https://gitlab.com/tezos/tezos/-/merge_requests/1980
- adds crowbar tests for lib_base, introduces randomness in CI, fixes bug
- could fix the seed later when crowbar is updated
- idea: do it manually (deterministic / non-deterministic)
- idea: fix seed on regression tests & separate tests with new seeds in longer tests
- idea: 2 test: fixed / non-fixed (allow to fail)
- => merge now and re-evaluate later.
- https://gitlab.com/tezos/tezos/-/merge_requests/2035
- What should we do about the links to protocol's MRs that have been removed? I've added a pattern for them to ignore list for now.
- they're not removed, just private
- https://gitlab.com/nomadic-labs/tezos/-/merge_requests/73 as an example try to open in incognito

## 2020-08-10

Knowing who to assign MRs to:

- :palm_tree: slack-emojis are nice, but missing _“back to work”_ information.
- Maybe at the top of this file, things like: “Seb: OOO Aug 28”?

MRs:

- https://gitlab.com/tezos/tezos/-/merge_requests/2019/commits
- do we still ask for squashing those kinds of histories?
- YES
- https://gitlab.com/tezos/tezos/-/merge_requests/2005
- @klakplok approved, then bunch of comments & pushes, it it still approved?
- still assigned to Benjamin
- [RPC to query by operation hash](https://gitlab.com/tezos/tezos/-/merge_requests/1988)
- Indexer functionality: do we want that **at all**?
- Shipping `tezos-indexer` with Tezos? ⇒ improved tezos-client
- MR assigned to Adrian.
- [!1980](https://gitlab.com/tezos/tezos/-/merge_requests/1980/diffs): We need to decide if want crowbar in the CI
- if yes, opam-repo MR needs to go through first
- → `test` meeting discussion
- https://gitlab.com/tezos/tezos/-/merge_requests/1914 (`/config`)
- ⇒ register “untyped” `/config` end-point: `Data_encoding.json`
- Romain to ask Pietro about this solution

Asked by François:
- https://gitlab.com/tezos/tezos/-/merge_requests/2006
- https://gitlab.com/tezos/tezos/-/merge_requests/2004
- https://gitlab.com/tezos/tezos/-/merge_requests/2002 (François is interested in feedback)

Merge-Team Dispatcher: Tomáš

## 2020-08-03

Dalphanet:
- spawning the network was smooth, but not making the branch because of issues with opam packages

MRs:
- https://gitlab.com/tezos/tezos/-/merge_requests/2000
- we may need different sinks to have different log levels
- at least need documentation on tezos.gitlab.io
- https://gitlab.com/tezos/tezos/-/merge_requests/1914
- actually adds `/config/p2p` and `/config/p2p/limits`
- is it a problem in case we want to add `GET /config` later? (to do that we would need to move many types in `lib_base`)
- if we do add `GET /config`, will we want to remove `/config/p2p/*`? Or is it interesting to have it on its own as well?
- `/config` would really be better
- can the RPC be defined in `Node_config_file`?
- what would we need to move to have `/config` / how hard would it actually be?
- https://gitlab.com/tezos/tezos/-/merge_requests/1970
- can probably be merged but need a DDB expert to review => Philippe?
- https://gitlab.com/tezos/tezos/-/merge_requests/1725
- LGTM from Tomas and Marco => merge?

Other subjects:
- P2P team now uses a `p2p_master` branch with several MRs merged into it if I understand correctly
- are we OK with this, in particular are we OK to merge the whole branch at once at the end?
- ok if their review process is good, with a good MR with a changelog, easy to review
- not ok if the goal is just to bypass the merge team review process
- => wait and see?
- if `p2p_master` contains like 20 MRs it's too much
- why do they want to do this?
- Next merge master => Sébastien

## 2020-07-27
### Information

- toynet serves its purpose, it raises problems :-)
- Bootstrap heuristic again
- Client config init x2 (fields in the JSON file and clash with mockup)
- stuff on the protocol side to be discussed in the protocall

### General question to discuss

- RPC deprecation mechanism https://gitlab.com/tezos/tezos/-/merge_requests/1982
- Generate an OpenAPI from the doc of RESTO https://gitlab.com/tezos/tezos/-/merge_requests/1907
- Merge master volunteers?
- Romain this week
- Seb next week but not this week

## 2020-07-20
- v7.3
- Quick information
- https://gitlab.com/tezos/tezos/-/merge_requests/1960 is waiting for a review by an HaCL dev
- A concensus has been reached about https://gitlab.com/tezos/tezos/-/merge_requests/1711 (to be implemented after vacations)
- Needs a merger
- Lwt exit : https://gitlab.com/tezos/tezos/-/merge_requests/1880
- {client,sandbox}.py APIs : https://gitlab.com/tezos/tezos/-/merge_requests/1887
- Needs discussion
- Shell/DDb: remove deadcode : https://gitlab.com/tezos/tezos/-/merge_requests/1969
- add new rpc /network/config : https://gitlab.com/tezos/tezos/-/merge_requests/1914 (new RPC for intenals, is it right? the right place? the right names?)
- Trends
- lwtreslib : https://gitlab.com/tezos/tezos/-/merge_requests/1872
- Baker reorg (deterministic nonces 1245 (require to decouple nonces and baking) and error management 1972)
- DistributedDB & P2P (1970)
- Still time
- P2P/io_scheduler refactoring : https://gitlab.com/tezos/tezos/-/merge_requests/1921
- Store : https://gitlab.com/tezos/tezos/-/merge_requests/1954 (& https://gitlab.com/tezos/tezos/-/merge_requests/1512 )
- Elect a new merge master
- Pierre won :-)

## 2020-07-13

- New Merge Master for the week is Pierre
- Is it our responsibility to prod MR authors to address the comments? Should we push authors forward?
- We should push authors. It's our responsibility to push things forward. It isn't our responsiblity to assign priorities for people though.
- Long discussion about Python tests and potential security vulnerabilities
- https://gitlab.com/tezos/tezos/-/merge_requests/1840 (client binary accept an --endpoint argument) - merged
- https://gitlab.com/tezos/tezos/-/merge_requests/1936 - merged
- https://gitlab.com/tezos/tezos/-/merge_requests/1881 - mergeable (see last comments for some new commits to review quickly though)
- https://gitlab.com/tezos/tezos/-/merge_requests/1711
- is poetry available in debian stable?
- is pip to be deprecated and replaced by poetry?
- what is the general rationale for this change? this adds contraints, is it really necessary?
- security (notably in CI): hashes of python packages? how can we trust the packages we use? how can we trust the version of poetry ? should python be run in protected branches?
- Conclusion: talk to Danny and say we're not feeling sure about this.
- https://gitlab.com/tezos/tezos/-/merge_requests/1430
- first fixup
- then merge
- question about storing socket in tmp, if mounted in certain configuration, might affect performance.
- put it in the data dir. @vbot
- https://gitlab.com/tezos/tezos/-/merge_requests/1960
- Pierre: clean MR, but gives creepy flashbacks to previous bugs. Requires careful reading. But requires reading of HACL*, is that necessary?
- Adrian: general question, should we get subsequent audits when changing stuff? Mostly out of our hands.
- https://gitlab.com/tezos/tezos/-/merge_requests/1519
- Assigned to vbot for second approval
- https://gitlab.com/tezos/tezos/-/merge_requests/1962
- merged
- https://gitlab.com/tezos/tezos/-/merge_requests/1845
- merged