---
title: nf-core/maintainers meeting notes 2024
tags: meetings,maintainers,2024
---

## 2024-11-29

### Attendance

- [x] Adam
- [ ] Anders J.
- [ ] Anders S. P.
- [ ] Carson
- [ ] Christopher
- [ ] Edmund
- [x] Florian
- [ ] Gisela
- [ ] Igor
- [ ] Harshil
- [x] James
- [ ] Jasmin
- [x] Jon
- [x] Joon
- [ ] Jonas
- [x] Jose
- [x] Júlia
- [ ] Lili
- [ ] Luisa
- [ ] Mahesh
- [ ] Matthias D. S.
- [ ] Matthias H.
- [x] Maxime
- [ ] Moritz
- [x] Nico
- [x] Nicolas
- [ ] Ram
- [ ] Rike
- [ ] Rob
- [x] Simon
- [ ] Sofia
- [x] Sateesh

### Agenda (Don't edit, only append!)

1. **@Jon** To what degree should we adhere to Semantic versioning in pipeline releases? Specifically, what qualifies as a major release?
- Option 1: we follow pretty closely. Any change to the interface (existing params/ inputs) should be a major release, so we should be doing a lot more of those. PR open at https://github.com/nf-core/website/pull/2842. (@Jon + ?) <<< pick this one!
- Option 2: PhilVer - VERY loosely SemVer, but not too many major release because people would get major release fatigue (@Phil + ?)
- Option 3: We use an LLM automation to tell us what the next release version should be so we don't argue about this (Florian + ?)
2. **@Rike** Should we push for CRAM support in the genomics pipelines? What are the blockers (@Rike - in an attempt to safe polar bears)
3. **@Júlia** Remove `workflowCitation()` function completely? or add it back to the pipelines template? https://github.com/nf-core/modules/pull/7094 (Júlia)
4. **@Florian** [Was contacted by a Nextflow dev at CDC about](https://nfcore.slack.com/archives/C043UU89KKQ/p1732763298086419) a checklist for Nextflow development they wrote: https://github.com/Ethan-Hetrick/ethans-toolbox/blob/main/nextflow/nextflow-best-practice-checklist.md#execution (Florian)
5. (@speri) CI Changes:
1. `nf-test.yml`: Runs Nextflow tests for workflows and modules across profiles using matrixed shards.
2. `gpu-tests.yml`: Executes GPU-specific Nextflow tests for changes in GPU-related workflows and modules.
3. `nf-test-action.yml`: Provides a reusable action to standardize and run Nextflow tests with configurable setups.
* [Implementation PR](https://github.com/nf-core/modules/pull/7007/files)
4. `detect-nf-test-changes`: Include tags / Exclude tags
* [Implementation PR](https://github.com/adamrtalbot/detect-nf-test-changes/pull/16/files)
6. **@jfy133** Please announce when experimenting on public repos!
7. **@Simon** Modules PRs are piling up again
8. **@maxulysse**: /!\ Updating AWS iGenomes cf #references
9. **@maxulysse**: Unify keys in igenomes / unify params name / unify fields in samplesheets / unify meta map fields (related to 1 and 7)
10. **@maxulysse**: Automatic release notes (Addressed by 1)
11. **@maxulysse**: Auto-generate CITATIONS for CITATIONS.md (all tools) and MultiQC (on run time) cf #regulatory

### Minutes

0. Has everyone recovered hackathon/summit? Some bitterness, some car mechanics
1. Should we be stricter on SemVer?
- Florian: We should try to make tooling to help developers on this
- Jose: concerned about people misintepreting what a major release?
- James: maybe just include in docs to tell Devs to avoid this

TANGENT: automation of releases?

- Floran, Maxime, etc. what a lot of automation around that
- Júlia reluctant because a lot of the AI stuff is really ugly
- Floran says doesn't have to be LLM, but we would like more automation (maybe with tools) to also help standardisation - there are a lot of manual steps
- GitHub action app for releases; some tooling stuff already used in tools, for example

2. Maxime (for Rike) CRAM everywhere rather than BAM?
- Adam/James: many tools won't work with CRAM, and then decompress/recompess procedures loses benefit
- Maxime: we should fix those tools!
- Adam/James: laughs in C and ancient DNA tools
- Simon: can't influence input
- Ultimately Rike wants bigger push
- More awareness during reviewing?
- Stronger wording in specifications
3. Júlia: OK for me to remove `workflowCitation()`
- Maxime: suggests replacing function within an nf-core nextflow plugin
- No one uses/needs it, agree to remove
4. Florian: checklists we could adopt?
- Florian: CDC checklist from CDC is old
- Sateesh: acually partly from when he was there
- James: where to put them is hard question as docs structure isn't great
- General agreement: he can make a PR if he wants, and when we do an overall we can check his checklist and use stuff we might be missing
5. @speri CI Changes
- TL;DR
- nf-test sharding
- GPU tests -> just needs a specific tag to run on a GPU runner
- James: can we do similar GPU test but for requesting larger runners (E.g. tools that need more HDD space)?
- Adam/Sateesh: somewhat yes, would require duplicating current action and set up a new/different queue for larger runs
- Simon: does it help preventing PRs that run 200 tests accidently
- Sateesh: Yes!
- Jon/Simon (and many others) hate the new uninformative names
- Sateesh to investigate, Maxime has a hacky implementation
- Simon: random linting issues?
- Sateesh: Edmun has a PR to fix this soon
6. James: ~~James wants Maxime to stop experimenting on Sarek and rnaseq~~ Sateesh and Edmund exprimenting on modules ;)
7. Simon: Sad that modules PRs are pilling
8. Maxime: updating Igenomes, thanks Sateesh and Adam for the reviews

### Action Points

- [ ] Everyone to dump their ideas for breaking changes onto website [#2842](https://github.com/nf-core/website/pull/2842), and in 2 weeks @Jon to merge in, and then @Julia to design (release) linting rules around this
- [ ] @Rike come up with more examples of how to push
- [ ] @Júlia to remove the function!
- [ ] @Florian: respond to Ethan at CDC to make a PR if he wants
- [ ] @speri to improve naming of sharded CI tests
- [ ] Everyone: let's close PRs in modules
- [ ] Everyone: Announce it when you do stuff on common Repos :pray: -> fine to experiment, but be loud about it we're not in the Far West anymore, not more merge cowboy

## 2024-10-25

### Attendance

- [ ] Adam
- [ ] Anders J.
- [ ] Anders S. P.
- [ ] Carson
- [ ] Christopher
- [ ] Edmund
- [x] Florian
- [ ] Gisela
- [x] Igor
- [ ] Harshil
- [x] James
- [ ] Jasmin
- [x] Jon
- [ ] Joon
- [ ] Jonas
- [ ] Jose
- [ ] Júlia
- [ ] Lili
- [ ] Luisa
- [x] Mahesh
- [x] Matthias D. S.
- [x] Matthias H.
- [x] Maxime
- [ ] Moritz
- [ ] Nico
- [ ] Nicolas
- [ ] Ram
- [ ] Rike
- [ ] Rob
- [ ] Simon
- [ ] Sofia

### Agenda (Don't edit, only append!)

- [x] Action points from last time
- [x] Maxime to invite Lorena Pantano -> FOR NEXT TIME
- [x] Rike to really this time restart nf-core helpdesk
- [x] Sign up for re-booted helpdesk: https://docs.google.com/spreadsheets/d/1c1Pz97W5azlPph9iiORAxRCLc6q4KBQRfK4M0h127KM/edit?usp=sharing
- [x] nf-core hackathon!
- Who's coming?
- Who's doing what?
- Most people as leads already
- Jon -> beginner training whole time :scream: (but hackathon was planning DiffAbun improvements and references)
- Florian -> Protein design
- Mahesh -> genomeassembly/genomeqc
- Matthias -> drawing corner
- Igor -> expecting to work on CRG stuff
- Sateesh -> nf-test plugins
- Maxime -> Everything
- Matthias de S -> Modules

- [x] [name=Mahesh Binzer-Panchal]Mahesh PoC of nf-test config clean up: https://github.com/nf-core/modules/pull/6788
- [x] [name=Mahesh Binzer-Panchal]What to do about module maintainers holding up contributions? https://github.com/nf-core/modules/pull/5706
- [x] (sateesh) Discuss the depreciation of `conf/modules.config` in favor of individual configs. Example implementation - [methylseq dev](https://github.com/nf-core/methylseq/tree/dev)
- [x] (Mahesh): deprecation of conf/modules.conf: Makes it difficult to find settings when separate. User experience from a colleague, they like being able to see what flags are set given certain parameters. They don't want it split up.
- ~~[@jfy133] despite assumed knowlege, `multiMap` indeed does _not_ ensure syncronised subchannels - as listed in updated docs under the new [note block](https://www.nextflow.io/docs/latest/reference/operator.html#multimap). We should change modules specs to have a single tuple to enure all files come intogether syncronised~~
- ~~We shouldn't mass change all modules, just update over time~~
- [MultiMapOperator code:](https://github.com/nextflow-io/nextflow/blob/master/modules/nextflow/src/main/groovy/nextflow/extension/MultiMapOp.groovy)
- [x] [name=Matthias De Smet] Module guidelines points 3.3 and 3.6 seem to contradict each other. We should remove the ambiguous language and define whether or not each output file type MUST have their own output channel
- [x] [name=Mahesh Binzer-Panchal] Monthly update of changes in guidelines or PR reviewing. It's quite easy to fall behind and it would be nice to have monthly coverage of what's changed and go over how to review a PR. This could be a part of the office hours.
- Can we have a GH action or something that on review request puts a checklist linking in a comment for the reviewer to check off. There would be a link also to the current guidelines for clarification of any points to be checked off.

### Minutes

- Maxime didn't invite Lorena
- Hackathon
- Who's doing what?
- Most people as leads already
- Jon -> beginner training whole time :scream: (but hackathon was planning DiffAbun improvements and references)
- Florian -> Protein design
- Mahesh -> genomeassembly/genomeqc
- Matthias -> drawing corner
- Igor -> expecting to work on CRG stuff
- Sateesh -> nf-test plugins
- Maxime -> Everything
- Matthias De S -> Modules
- Mahesh' PoC clean up
- Everyone loves it!
- Only Maxime's concern: people might start doing this in a pipeline
- Mahesh to update specifications and docs with warning!
- Waiting for maintainers to review/approve
- Guideline: if no (reasonable) action after 3 months, then go ahead and merge if reviewed thoroughly and tests passing
- Maybe ask for another reviewer (so it's not steamrolling)
- Splitting of `modules.conf` into one per module
- Maxime/Sateesh reasoning: to help reduce triggering automated tests on file changes (i.e., don't trigger all module tests jsut because `modules.conf`)
- Paolo/Mahesh/Jon/James/Matthias:
- too hard to find for users to know exactly what arguments
- not natively suppported by nextflow (a main.nf automatically imports the neighbouring `nextflow.config`)
- cascading configs in rnaseq was also bad (Maxime)
- (Mahesh) user will be confused - they find the modules modules.conf and think that's what is being used, but won't be awaremof being overridden by subworkflows.conf
- Testing shouldn't be the priority over user/usability (Matthias) -> `nextflow config` should properly print out stuff, then it's OK (a user can use that to view)
- To improve testing efficiency, we should add our own layer to detect changes
- Igor: there are some additional subtitles to configs that won't work by passing stuff through channels
- Vote: 6 nay for splitting vs 2 splitting
- Modules specifications inconsistency
- First solution: relax both 3.3/3.6
- Matthias de S. wants to make strict again (James remembers discussion to relax this -> Mahesh also says it simplifies code a lot if we share channels for files that serve same purpose, less conditional channel combining)
- Jon doesn't like unpredictable types
- Mahesh/James don't like it because code becomes extremely complicated with loads of if/else statements to combine each channel back to the main file
- James: users _shouldn't_ modify args to add unsupported stuff that could influence this output file
- Being informed about specification updaes
- Versioning of docs again?
- Better idea: more linting! Matthias says make issues for linting when you want it!

### Action points

- [ ] Maxime: blog update!
- [ ] Maxime: invite Lorena
- [ ] Mahesh: update nf-core modules specifications and other relevant module/nf-test documentations
- [ ] Matthias: follow up with Simon P. if he remembers where allowing the combining of e.g. .bai,.csi came from/history

## 2024-09-27

- [ ] Adam
- [ ] Anders J.
- [ ] Anders S. P.
- [ ] Carson
- [ ] Christopher
- [ ] Edmund
- [x] Florian
- [ ] Gisela
- [ ] Igor
- [ ] Harshil
- [ ] James
- [x] Jasmin
- [ ] Jon
- [ ] Joon
- [x] Jonas
- [x] Jose
- [ ] Júlia
- [ ] Lili
- [ ] Luisa
- [x] Mahesh
- [ ] Matthias D. S.
- [ ] Matthias H.
- [x] Maxime
- [ ] Moritz
- [ ] Nico
- [ ] Nicolas
- [ ] Ram
- [ ] Rike
- [ ] Rob
- [ ] Simon
- [ ] Sofia

### Agenda

- Proposal for new maintainer: Lorena Pantano
- Early nf-core adopter, one of the original maintainers for #smrnaseq and some other pipelines.
- Reduced activity when she left academia to go to industry, but now she's back in academia again at Harvard.
- She's now in the group that the bcbio workflow manager came from, and they're in the process of switching to nf-core and porting any missing functionality from bcbio into nf-core.
- YES - Maxime will contact and ASK her
- Coming nf-core/tools release
- MASSIVE modules MASSIVE updates
- GOOD progress, but still things to do
- [new modules page](https://github.com/nf-core/website/pull/2744)
- New modules page looks AWESOME
- Topics channel Edmund?
- Edmund created a POC here https://github.com/nf-core/nascent/pull/150/files and will test this out. TBD: can topics and version.yml be kept in parallel to allow testing in some pipelines without destroying it all. He will test.
- TEMPLATE update
- Going further done the line of simplifying it
- @maxulysse to do a speedrun to convert template to non-genomics pipeline at hackathon. Other than that people generally happy with the update: better support for non-nf-core pipelines, easier template merges for non-genomics pipelines
- Conda lock files, where are they? (Maxime will look):
- They will improve reproducibility
- helpdesk is the new coffee-n-code
- Rike to set it up. Really this time. Sorry!
- Back to the discussion whether modules should have mandatory parameters as input (James vs. Simon)
- pipeline level test in nf-core/rnaseq and nf-core/methylseq: https://github.com/nf-core/rnaseq/tree/master/tests / https://github.com/nf-core/rnaseq/pull/1379 and https://github.com/nf-core/methylseq/pull/414:
- Discussion what to snapshot
- Maxime to give a bytesize in how to do pipeline level snapshot
- Mahesh shows how nf-test can be run without additional config files by putting the params into nf-test https://nfcore.slack.com/archives/C049MBCEW06/p1721042240473589:
- Generally everyone happy about it. Mahesh will implement it for 1-2 modules.

## 2024-08-30

### Attendance

- [ ] Adam
- [ ] Anders J.
- [ ] Anders S. P.
- [x] Carson
- [ ] Christopher
- [x] Edmund
- [x] Florian
- [ ] Gisela
- [x] Igor
- [ ] Harshil
- [ ] James
- [ ] Jasmin
- [x] Jon
- [ ] Joon
- [ ] Jonas
- [ ] Jose
- [x] Júlia
- [ ] Lili
- [x] Luisa
- [ ] Mahesh
- [ ] Matthias D. S.
- [ ] Matthias H.
- [x] Maxime
- [ ] Moritz
- [ ] Nico
- [x] Nicolas
- [ ] Ram
- [x] Rike
- [ ] Rob
- [x] Simon
- [ ] Sofia

### Agenda (Don't edit, only append!)

- (Phil) Seqera Containers
- https://github.com/nf-core/modules/issues/6505:
- Reviewing upcoming blog post by Edmund & Phil on Seqera Containers roll out
- Summary: We are all happy. Lock files will be rolled out to ensure reproducibility. Container URLs will be autogenerated and hard coded into a config file.
- Phil/Edmund will post a PR with the finalised Blog for discussion
- (Edmund) Quick vote on [Add renovate comments to environment.yml](https://github.com/nf-core/modules/issues/6504):
- Keep bioconda declaration
- (James) Review last meetings action points
- [ ] Matthias to attempt again to list modules on subworkflow page
- [ ] Postponed to next time
- [x] Julia to include singularity/conda tests in template (deadline: next template release).
- Done, will come with next template update
- Do we want the 0.9.0 nf-test CI in the template? For 3.0.0 release
- [x] Nicolas to update module guidelines to say all assertions in one snapshot (definitely not one per file):
- Done
- [ ] Mahesh to talk about params ext.args next time
- [ ] Postponed to next time
- [ ] Matthias to think about new ways of versioning module guidelines
- [ ] Postponed to next time
- [ ] Maybe annotate modules with guideline version
- [ ] Maybe don't spend too much time on that cf Simon
- (Rike) Office-hours continuation
- We have volonteers to host session
- Rebranding to (let's vote)
- coffee-n-code
- nf-coffee-chat
- fireside-chat
- (Simon)
- [ ] Discuss allowing a [default but overwritable prefix for modules](https://github.com/nf-core/website/pull/2608) (especially to convince James and Nicolas)
- [ ] Nicolas is willing to compromise if we use task name as default. It would ensure it should work always
- [ ] The disambiguate function should still be present because suffixes can overlap
- [ ] the default behavior by many people is to add a suffix at the moment.
- [ ] The output definition should in addition help with clearing up filename
- [ ] Inviting any additional maintainers, particularly Usman Rashid
- [ ] YES!
- [ ] Maxime will ask
- [ ] [nf-test assertions page](https://nf-co.re/docs/contributing/nf-test/assertions) needs updating
- [ ] Someone needs to do it
- [ ] Maxime will lead and ping Nicolas/Simon/Usman
- [ ] Any plans for meeting up at the Barcelona summit?
- [ ] Maxime will figure out who will be around when

## 2024-07-28

### Attendance

- [ ] Adam
- [ ] Anders J.
- [ ] Anders S. P.
- [ ] Carson
- [ ] Christopher
- [x] Edmund
- [ ] Florian
- [ ] Gisela
- [x] Igor
- [ ] Harshil
- [x] James
- [ ] Jasmin
- [ ] Jon
- [x] Joon
- [ ] Jonas
- [x] Jose
- [x] Júlia
- [ ] Lili
- [x] Luisa
- [ ] Mahesh
- [ ] Matthias D. S.
- [x] Matthias H.
- [ ] Maxime
- [ ] Moritz
- [ ] Nico
- [x] Nicolas
- [ ] Ram
- [x] Rike
- [ ] Rob
- [ ] Simon
- [ ] Sofia

### Agenda (Don't edit, only append!)

- Last meetings action points
- [ ] Matthias to attempt again to list modules on subworkflow page
- [ ] Jon to update guidelines based on notes above (everyone to comment - with a poll)
- [x] Rike to set up a working group for pushing forward daisy-chaining
- [x] Test-data working group set up by Simon
- [x] Modules bulk update group set up by Julia
- [x] (Júlia) Test pipelines with singularity and conda
- [x] (Júlia) Bulk update of modules (https://github.com/nf-core/modules/issues/5828)
- (Júlia) Restructuring of meta.yml (https://github.com/nf-core/modules/pull/5867)
- (Nicolas) nf-test - a single snapshot per test vs a single snapshot per file (meaning many per test)
- (Mahesh) use params for ext.args: https://nfcore.slack.com/archives/C049MBCEW06/p1721042240473589
- [ ] Versioning module guidelines? Suggested by Arthur Gymer.

### Minutes

- Last meeting action points:
-
- Test pipelines with singularity/conda:
- crisprseq: only testing with docker in CI, but conda was failing! Singularity also failed horribly
- ✅ singularity & conda also run on PR to master + workflow dispatch?
- Make singularity required
- Conda optional opt-out
-
- Workflow dispatch?
- Once a month? (only if new commits detected)
- Only implement once nf-test is in pipelines
- But what about pipelines that use custom docker containers(?) -> conda doesn't exist
- But: nf-test in template will be quite a while
- Bulk updates of modules:
- Plan: update all modules with:
- meta.yml -> channel structure
- subworkflows ->
- bio.tools identifiers
- edam ontologies for all input/output files
- removing defaults from conda channel
- adding lock files
- adding version topic channels
- adding lint for stub tests for all modules
- Version topic channels:
- Maybe should be gradual overtime
- Requries pipeline changes anyway
- You can actually mix and match topic/non-topic channels
- Hackathon task
- subworkflow meta.yml restrucutre:
- Only has name of the channel, being expanded to better match the channel construct (not just files)
- What is the name of the overall channel XYZ - NMaming stuff: now FIGHT!
- Structure -> Accepted by majority (only Mattias unhappy)
- nf-test snapshot format:
- One snapshot per file is bad!
- All assertions in a snapshot
- params for ext.args:
- For next time
- Versioning guidelines
- Everyone think it's a good idea
- Linked with too tools version (but docs independent) or date?
- Linting against tools release, but can have a `dev` version of modules guidelines?
- Problem: linting can be against tools dev, but this is a not a stable modules version guidelines specification... could change a lot prior next tools release
- Commits to website markdown file (too regular/fast)?
- Major issue: how to render on website, without copying the entire markdown page every time etc
- Also plans (although James found 🤢) to split current guideline pages into multiple pages

### Action points

- [ ] Matthias to attempt again to list modules on subworkflow page
- [ ] Julia to include singulairty/conda tests in template (deadline: next template release)
- [ ] Nicolas to update module guidelines to say all assertions in one snapshot (definitely not one per file)
- [ ] Mahesh to talk about params ext.args next time
- [ ] Matthias to think about new ways of versioning module guidelines

## 2024-06-28

### Attendance

- [ ] Adam
- [x] Anders J.
- [ ] Anders S. P.
- [x] Carson
- [ ] Christopher
- [x] Edmund
- [x] Florian
- [ ] Gisela
- [x] Igor
- [ ] Harshil
- [x] James
- [ ] Jasmin
- [x] Jon
- [ ] Joon
- [ ] Jonas
- [x] Jose
- [x] Júlia
- [ ] Lili
- [ ] Luisa
- [ ] Mahesh
- [x] Matthias D. S.
- [x] Matthias H.
- [ ] Maxime
- [ ] Moritz
- [x] Nico
- [x] Nicolas
- [x] Ram
- [x] Rike
- [ ] Rob
- [x] Simon
- [ ] Sofia

### Agenda

- Actions from last time:
- [x] Rike to ask Chris about office hours
- [x] James to invite new people/remove old people to team
- [ ] preprocess_rna subworkflow: each maintainer come up with a renaming proposal for that workflow, and vote
- [x] Júlia to make tools dev branch much more strict (require tests passing etc.)
- [x] Rike to bring up expanding new pipeline approval rights to also matinainers?
- [x] James maintainrs minutes
- [x] Rike to add maintainers approval and namespace squatting timeout to core agenda
- Jon's subworkflows name proposal:
- Maxime: `PREPARE_RNASEQ_READS_FOR_ALIGNMENT`
- Matthias: `FASTQ_TRIM_QC_SUBSAMPLE_FASTQC_UMITOOLS_TRIMGALORE_SUBSAMPLE_FQ_SALMON_SORTMERNA`
- Matthias: `MY_SUBWORKFLOW_FINAL_FINAL_3_FINAL`
- Jose: `FASTQ_QC_TRIM_FILTER_REMOVE_STRANDEDNESS`
- Jose: `FASTQ_PREPROCESS_FOR_RNASEQ`
- Jose: `FASTQ_PREPROCESS_RNASEQ`
- Jon: `PREPROCESS_RNASEQ`
- Jon: `PREPARE_RNASEQ_READS_FOR_ALIGNMENT`

- Do we give Maxime his own pipeline (an extra one since he already has many toys)?
- cf https://nfcore.slack.com/archives/C043UU89KKQ/p1718114169742709
- He wants one now, but he will ask for more
- (Maxime wrote all that)
- Discuss problems with test-datasets
- Very messy outside genomics
- Getting reviews are slow
- Annoying to have to get two reviews each time (test-datasets, then module)
- Suggestions to allow test-data files alongside test files in modules repo
- But file polultion etc...
- [See here](https://nfcore.slack.com/archives/C043UU89KKQ/p1718129591667489)
- Non-nf-core test-data repos allowed?
- Branch protection for the modules branch at least?
- Linking external github data
- How can we ensure the full size tests are run before a release?
- We agreed in the past to check in PRs that the full size test is run (especially for first releases)
- We recently had many released pipelines fail the full size test
- Clearly we are very bad at remembering without some form of automated checking/blocking the release.
- How can we do better?
- Do we need more frequent/longer meetings, or sub-group meetings (e.g. modules/subworkflows)

### Minutes

- Jon's subworkflows name proposal:
- Suggestions
- Maxime: `PREPARE_RNASEQ_READS_FOR_ALIGNMENT`
- Matthias: `FASTQ_TRIM_QC_SUBSAMPLE_FASTQC_UMITOOLS_TRIMGALORE_SUBSAMPLE_FQ_SALMON_SORTMERNA`
- Con: if you add a new module, renaming is bad
- Pro: you know that your tool of choice is in there
- Con: doesn't allow switches inside subworkflow to allow different tools
- Matthias: `MY_SUBWORKFLOW_FINAL_FINAL_3_FINAL`
- Jose: `FASTQ_QC_TRIM_FILTER_REMOVE_STRANDEDNESS` (input, generic process description)
- Jose: `FASTQ_PREPROCESS_FOR_RNASEQ`
- Jose: `FASTQ_PREPROCESS_RNASEQ`
- Jon: `PREPROCESS_RNASEQ`
- Jon: `PREPARE_RNASEQ_READS_FOR_ALIGNMENT`
- Thoughts
- No strict naming, guidelines instead and approval?
- What about renaming? Symlink to new one?
- Noted shifting of feeling - subworkflows are too variable to standardise
- Guidelines:
- If short with single tool options, we suggest explicitly name the tools (approval not required)
- Only main modules, not auxilary (e.g. index/sort)
- If more generic subworkflow, multiple options or very long tools chains don't
- Guideline: describe the the workflow in 1 sentence, and then condense based on main points
- Generally: approach is finadability via keywords, but understanding by descirptive name
- Maxime's proposal (variant part of sarek as separate pipeline - i.e., pipeline chaining)
- Lots of customers people want this at Seqera, as do in other places
- Should push forward
- Where does documentation then go (if you start merging pipelines together, can default merge them)
- Nextflow code should be relatively easy, but it's everything around it that is the issue
- Working group?
- Turn down for now, because e.g. docs maintaince burden is too high; as it's still an early PoC -> Still too many questions about what is a subworkflow/workflow
- test-data
- Fundmental issues with 'delete-me'
- Need a working group
- Need to remove a lot of stuff linked outside nf-core test-datasets (when not necessary)
- README is very messy, makes it hard to find stuff too (puts people off)
- e.g. just a simple protein sequence
- Some nf-tests have huge setup, could be reduced by putting the reuqired input on test-dataset -> if we have a good place put them!
- Agreement to have better docs: where is what? how can I produce small test files on the fly?
- Render the documentaiton on the website
- AWSMegatests:
- Pipelines fail
- One reason for pipelines failing is interactions with FUSION, which most users don't test during development. Are we clear enough about this and do people know they can trigger Megatests manually?
- Improve documentation that everyone can submit this job
- add it to https://nf-co.re/docs/checklists/pipeline_release
- Can we run the action on open dev to master?
- GET BACK TO MAKE THE BOTS DO ALL THE WORK: https://github.com/nf-core/tools/issues/2808
- Bot should work on 2 passing PR reviews
- Matthias will look into the action
- Meeting frequency:
- longer tough
- more meetings also tough to schedule
- vote on important topics
- Modules bulk update:
- Julia will organize working group

### Action Points

- [ ] Matthias to attempt again to list modules on subworkflow page
- [ ] Jon to update guidelines based on notes above (everyone to comment - with a poll)
- [ ] Rike to set up a working group for pushing forward daisy-chaining
- [ ] Test-data working group set up by Simon
- [ ] Modules bulk update group set up by Julia

## 2024-05-31

### Attendance

- [x] Adam
- [x] James
- [x] Jon
- [x] Jose
- [x] Júlia
- [x] Maxime
- [x] Nicolas
- [x] Rike
- [ ] Anders J.
- [ ] Anders S. P.
- [ ] Christopher
- [ ] Daniel
- [ ] Edmund
- [ ] Gisela
- [ ] Harshil
- [ ] Jasmin
- [ ] Jonas
- [ ] Louisa
- [ ] Matthias D. S.
- [ ] Matthias H.
- [ ] Moritz
- [ ] Ram
- [ ] Rob
- [ ] Simon
- [ ] Sofia

### Agenda

- Show gratitude to the module-meister
- Check last times action points
- [x] Tell people to post request-review
- [x] Tell people short and quick merges
- [x] Spring cleaning! -> not this year (modules, configs ✅)
- [x] Office hours -> Rike to shout at Chris if he can finish in two week, otherwise we will go ahead and work it out
- [x] Maintainers hackathon day?
- [x] Yes to meeting discussion summary-> maxime to get a new zoom like channel fro mseqera to try treansicprt funcionality
- [x] Everyone who feel strongly abut subworkflows, should go throug nf-core/subworklows, and pick ones they don't like and ones they think have very good naming scehes, and have this read for next time
- Good byes and invitations
- Reviewing subworkflow naming homework
- Be considerate of other when updating TEMPLATE.
- cf https://nfcore.slack.com/archives/C043UU89KKQ/p1715178861026539
- news from the tools team: https://github.com/nf-core/tools/pull/2978
- Fix for modules `meta.yml` (https://github.com/nf-core/modules/issues/4983 and https://github.com/nf-core/tools/pull/2789)

### Minutes

- Proposals / goodbyes
- Proposals: ✅ Igor - ACCEPTED, ✅ Carson - ACCEPTED, ✅ Florian Wuennemann - ACCEPTED, ✅ Nico Trummer - ACCEPTED, ✅ Thomas Danhorn - TURNEDDOWN, ✅ Mahesh - ACCEPTED, ✅ Luisa Santus, ✅ Joon Klaps - ACCEPTED, ✅ Lili - ACCEPTED
- Goodbyes: Louisa, Alex P., (Matthieu?), Daniel Lundin
- More ideas for women please shout?
- Apologies people: please consider
- Subworkflow naming:
- Adam's idea:
- remove tools (final suffix) OR input file
- input + verb/noun
- bed_scatter_bedtools -> bed_scatter
- but what if subworkflow with same function but different tools
- e.g. bam_deduplicate_samtools vs bam_markduplicates_picard
- Adam's issue: long isn't bad, just more info
- Jon: main issue is can have so many tools (so suffix gtets very long)
- Jon: just be more specific: `preprocess` is bad, but rather e.g. trim, removerna, -> semantics in names is bad
- Tools team update: nf-core version compatibility will now be tracked in pipelines
- Ken Brewer developing a minimal _nextflow_ template (by Monday?!) with nf-core compatibility features
- Júlia and tools team is already starting with this from nf-core side (new TUI has turn on/off functionality for example)
- But will take time until released, needs a lot of testing: please make a lot of pipelines!
- Tools team: Modules meta.yml is not well formatted - Jon made proposed (made on issue 4983)
- Jose makes points on parsing
- Maxime: wants additional default fields into modules (bibtex as channel), he will speak to Rike/Júlia
- New pipeline proposals:
- Core team is slow, give Maintainers approval permissions (one core, and at least one maintainer)
- Rike to add to core
- Release reviews?
- Keeping up to date so seems OK
- namespace squatting (e.g. scflow) -> release deadline before otehrs can take over
- Core team has already discussed, debating how long to require

### Actions

- [ ] Rike to ask Chris about office hours
- [ ] James to invite new people/remove old people to team
- [ ] preprocess_rna subworkflow: each maintainer come up with a renaming proposal for that workflow, and vote
- [x] Júlia to make tools dev branch much more strict (require tests passing etc.)
- [ ] Rike to bring up expanding new pipeline approval rights to also matinainers?
- [ ] James maintainrs minutes
- [ ] Rike to add maintainers approval and namespace squatting timeout to core agenda

## 2024-04-26

### Attendance

- [x] Adam
- [x] Edmund
- [x] James
- [x] Jasmin
- [x] Jon
- [x] Matthias H.
- [x] Maxime
- [x] Nicolas
- [x] Simon
- [x] Sofia
- [ ] Anders J.
- [ ] Anders S. P.
- [ ] Christopher
- [ ] Daniel
- [ ] Gisela
- [ ] Harshil
- [ ] Jonas
- [ ] Jose
- [ ] Júlia
- [ ] Louisa
- [ ] Matthias D. S.
- [ ] Moritz
- [ ] Ram
- [ ] Rike
- [ ] Rob

### Agenda

- Spring cleaning?
- Matthias H to lead
- New Q from @jfy133 from last time: When? 😬
- Maintainers team review?
- Maxime to lead
- New Q from @jfy133 from last time: When? 😬
- @jfy133: I've been getting questions/comments about sometimes long review times in pipeline reviews and smaller pipelines; guidlines/suggestions/strategies to improve?
- @jfy133: further subteams (responsibility for: configs, modules, new-pipelines etc?)
- Core proposal: Office hours with rotating volunteers
- Subworkflow naming guidelines

### Minutes

- PRs are mounting 😱
- Particularly on modules
- Check repo: 3
- Github notficaitons: 3
- Request-review slack: 6
- Emails: 2
- How to improve this?
- #request-review keeping up to date
- Maybe do need more ruthless with auto-closing stuff
- Need spring cleaning
- OFFICE HOURS
- Everyone on board, x1 month is probably good for everyone
- Reviewing is slow
- Particularly large templates complicated
- Need to make smaller
- Maybe office hours also for tradiginv revierws
- post-tools special office hours?
- Better guidance on short and requick
- Maybe do an annoucnemtn
- Summary end of each maintaniers meeting of discussion and publish on PI
- All agree!
- subworkflow naming
- Too much complexity if strictly follow subworkflow structure
- Too many tools in some cases
- Risk of people being too broad -> 'rnaalign'
- Which ? All?
- Discoverability is too low
- But what if you add another tool? Have to change the name
- Already problem with deprecation
- Risk of laziness
- Name has to at least been very specific
- nf-core subworkflows SHOULD be as short as possible, not huge mega ones
- You can embed subworkflows in local umbrella subworkflows
- Don't always agree -> want to be able to copy paste between different, if it's too much embedding too complicated
- Maybe each subwk must have a DAG/workflow grpah, if too complicated shouldn't be included
- Popular!
- Keywords isn't descriptive enough
- Graphical represntation will help
-

### Action points

- Tell people to post request-review
- Tell people short and quick merges
- Spring cleaning!
- Office hours
- Maintainer shackathon day?
- Yes to meeting discussion summary-> maxime to get a new zoom like channel fro mseqera to try treansicprt funcionality
- Everyone who feel strongly abut subworkflows, should go throug hnf-core/subworklows, and pick ones they don't like and ones they think have very good naming scehes, and have this read for next time

## 2024-03-29

- Cancelled due to Easter holidays for much of Europe

## 2024-02-23

### Attendence

- @maxulysse
- @mashehu
- @nvnieuwk
- @JoseEspinosa
- @sppearce

### Agenda

- @ChrisHakk: nf-core [office hours proposal](https://github.com/nf-core/website/pull/2292)
- Blog preview:point_down: [here](https://deploy-preview-2292--nf-core.netlify.app/blog/2024/office_hours)
- Want to build capacity by creating opportunities to upskill
- Aims to complement Slack
- It's a pilot, we want feedback!
- Time needs to be decided
- Action: @ChrisHakk to organise a poll in maintainers channel to find "best" time.
- Aim to kick off after the hackathon so be can benifit from paid Gather space.
- Conda lock files? - cf https://nfcore.slack.com/archives/C049MBCEW06/p1706195879697079
- Wave mulled containers https://github.com/nf-core/modules/pull/4080#issuecomment-1948953129
- Agreeing on test names + matches' names - cf https://nfcore.slack.com/archives/C049MBCEW06/p1706726078597519
- Dropping `ch_` prefix for channels.
- @maxulysse feels responsible for this since he started pushing for that in DSL1 (due to the need to duplicate channels)
- He do feels it's no longer useful in DSL2, and thinks we should rather prefix the rest
- Advantages, the channel name can directly be used in the emit statement.
- @jfy133 disagrees -> I think it makes code easier/clearer to follow (standard variable vs channel etc)
- @jfy133: I've been getting questions/comments about sometimes long review times in pipeline reviews and smaller pipelines; guidlines/suggestions/strategies to improve?
- @mashehu: Please pin all nextflow plugins: https://github.com/nf-core/tools/issues/2788
- Combining tools together in a single modules
- check out https://github.com/nf-core/modules/blob/master/modules/nf-core/samtools/collatefastq/main.nf
- Triage PRs in modules, add labels
- Matthias H will lead the SPRING CLEANING TORNADO TASK FORCE
- PRs, issues, branches, pipelines?

Actions points from last meeting:
- Volunteers wanted: Need group leaders for the hackathon in march
- Maxime will ask people
- Deprecation guidelines: https://github.com/nf-core/website/pull/2217
- Update docs with latest ideas for file name capture or snapshots cf qualimap + @jfy133 PR
- Nicolas to have a look
- Agreeing on a naming convention for the tests
- Nicolas will lead the focus group
- (@jfy133) Spring cleaning?
- Matthias H to lead
- (@jfy133) Maintainers team review?
- Maxime to lead

## 2024-01-26

### Attendence

@jfy133, @maxulysse, @kOcBx-beQZW6WOsI5UVQcA, Ram, @emiller, @YO3oG0hdSw2Gv5IWHc0nfA, Jasmin, Jon, Jonas, @JoseEspinosa, @jmir, @mashehu, Nicolas, Simon, Sofia,

### Agenda

- Was planned on 2023-12-15. but @maxulysse forgot about it, so will happen in 2014-01
- Improvement on the modules regarding multiple meta maps (Wait until Mahesh is a maintainer, or until we picked up his brain)
- Need to make an example and see if it's really useful or not in the end.
- Julia Mir and Mahesh might have good examples on how to deal with this without the extra meta maps messes
- -> https://github.com/nf-core/crisprseq/blob/master/modules/nf-core/minimap2/align/main.nf
- -> https://github.com/nf-core/modules/blob/27aa993895c29081f413fa91d6e981e134d41421/modules/nf-core/bwamem2/mem/main.nf#L11-L12
- James will look into that
- Let's ask if people is using the meta2 and such
- `meta2` -> `meta_file`?: https://github.com/nf-core/modules/pull/4594
- Matthias H
- Guidelines for pipeline developers could be of interest
- Implementation patterns
- Mahesh will start the focus group
- Thoughts on cmd for versions output -> https://github.com/nf-core/rnaseq/pull/1115
- Probably not the best solution (especially with optional tools and weird tools output)
- Default modules on template -> https://github.com/nf-core/tools/issues/2285
- 🎉 Julia did it already 🎉
- Piping samtools in a single module -> https://nfcore.slack.com/archives/C043UU89KKQ/p1702456394353949
- https://github.com/nf-core/modules/pull/4571
- should it be a subworkflows?
- we already have some bwa mem|samtools
- we could have some extra samtools "super" modules on a case base
- good for reducing data footprint and resource usage
- not a minimal re-usable module
- Changing guidelines on what a module is?
- @maxulysse will start a focus group for conventions and guidelines
- Simon will provide a couple of examples
- Be strict about failures -> https://nfcore.slack.com/archives/C043UU89KKQ/p1704904281357509
- :clap: no :clap: merging :clap: failing :clap: tests :clap:
- Immutable test data (and refactoring test-datasets)
- mirror of github hosted files on s3 as a POC
- Volunteers wanted: Need group leaders for the hackathon in march
- Maxime will ask people
- Deprecation guidelines: https://github.com/nf-core/website/pull/2217
- Update docs with latest ideas for file name capture or snapshots cf qualimap + @jfy133 PR
- Nicolas to have a look
- Agreeing on a naming convention for the tests
- Nicolas will lead the focus group
- (@jfy133) Spring cleaning?
- Matthias H to lead
- (@jfy133) Maintainers team review?
- Maxime to lead

### Action Points

- @Mahesh will start implementation pattern focus group
- @jfy133 add to approved new pipeline applciations instructions to join #pipeline-maintainers
- @maxulysse (with help from Simon) will start a focus group for conventions and guidelines