Add import tool and split 2026 notes

meeting_notes/2026/2026_01_jan.md

@@ -0,0 +1,297 @@
+# Matplotlib Weekly Meeting: January 2026
+
+###### tags: `2026 dev call`
+
+---
+
+# January 8, 2026
+
+_attending_: [@efiring](@QWhXj01mSwmTjk5kN1H_qQ), @tacaswell, @QuLogic, @ksunden, @timhoffm, @story645 
+
+## Agenda
+- [ ] RSE updates
+- [x] NASA updates
+- [x] 3.11 status
+- [x] 3.10.8 / build overhaul status - RELEASED
+- [x] [name=@rcomer] Trial [Discourse Chat](https://meta.discourse.org/t/discourse-chat/230881) as replacement for Gitter?
+- [x] [name=@hannah] GSOC: backend API improvements
+    - builds on [versioning](https://github.com/matplotlib/matplotlib/pull/30777)
+    - [GraphicsContext data classes](https://github.com/matplotlib/matplotlib/pull/30811)
+    - [fonts](https://github.com/matplotlib/matplotlib/issues/30890#issuecomment-3690830981)
+- [x]  Docs are getting big on digital ocean 
+
+## Notes
+### RSE updates
+* Melissa
+    - starting this week!
+* Tom
+    - very busy, some grant management
+* Kyle
+    - Purchased some ARM machines with money from pevious grant
+    - linux on ARM is harder than we expected
+* Elliott
+    - not much over break
+    - review a few things, update freetype and harfbuzz
+        - few minor changes that are improvements
+
+### NASA updates
+- Starting to spend money on ROSES 2024
+    - Y2 funds are in-process
+- need to write grant report on ROSES 2020
+
+### 3.11 status
+- still need to review Antony's PR
+- image resampling PR affects some text images
+    - one in, one pending
+- widget canvas handeling PR was not enough
+- aim for RC end of next week
+- Elliott is release manager
+- https://github.com/matplotlib/matplotlib/pull/30777 (reviewed and merged)
+- https://github.com/matplotlib/matplotlib/pull/30871/changes (reviewed)
+- https://github.com/matplotlib/matplotlib/pull/30886 (reviewed and merged)
+
+### chat in discourse?
+- no love of element/gitter expressed
+- proposal:
+    - turn discourse chat on
+    - redirect conversation from gitter -> discourse for a month
+        - posts made and pinned on matplotlib/community/incubator
+    - at end of month either turn it off or document and shutter element channels
+- Turned it on and created 3 channels
+    - @rcomer should weigh in on how to further organize
+
+### 3.10.8
+
+- out and merged up
+- publish removed from the main repo
+- need to double check that the trusted publisher from the main repo removed on pypi side
+- publishing from external repo worked
+
+### docs too big
+- checked out docs are now 32G+
+- running out of disk space on venus (our file server)
+- git seems to be doing a very good job at compression (the `.git` folder is 6-7G)
+- to get out of the jam we will prune old micro docs for non-current releases and replace with permenant redirects in caddy.
+    - can get back if we _really_ need it
+    - will buy us some more time before we run out of disk again
+```
+163M    1.2.1
+196M    1.3.0
+173M    1.3.1
+249M    1.4.0
+249M    1.4.1
+249M    1.4.2
+189M    1.4.3
+230M    1.5.0
+236M    1.5.1
+233M    1.5.3
+205M    2.0.0
+203M    2.0.1
+204M    2.0.2
+134M    2.1.0
+126M    2.1.1
+143M    2.1.2
+146M    2.2.0
+174M    2.2.2
+169M    2.2.3
+168M    2.2.4
+173M    2.2.5
+183M    3.0.0
+152M    3.0.2
+172M    3.0.3
+169M    3.1.0
+200M    3.1.1
+191M    3.1.3
+194M    3.2.0
+201M    3.2.1
+199M    3.2.2
+285M    3.3.0
+288M    3.3.1
+285M    3.3.2
+290M    3.3.3
+298M    3.3.4
+300M    3.4.0
+301M    3.4.1
+301M    3.4.2
+301M    3.4.3
+579M    3.5.0
+628M    3.5.1
+627M    3.5.2
+628M    3.5.3
+606M    3.6.0
+606M    3.6.1
+588M    3.6.2
+589M    3.6.3
+626M    3.7.0
+631M    3.7.1
+629M    3.7.2
+629M    3.7.3
+573M    3.7.4
+629M    3.7.5
+658M    3.8.0
+659M    3.8.1
+659M    3.8.2
+662M    3.8.3
+662M    3.8.4
+594M    3.9.0
+595M    3.9.1
+602M    3.9.2
+555M    3.9.3
+571M    3.10.0
+572M    3.10.1
+572M    3.10.3
+572M    3.10.5
+573M    3.10.6
+573M    3.10.7
+573M    3.10.8
+```
+- Dropping 1.x and 2.x will save 2.7G
+- Dropping 3.0.x is 334M
+- Dropping 3.1.x is 368M
+- Dropping 3.2.x is 395M
+- Dropping 3.3.x is 1.2G
+- Dropping 3.4.x is 901M
+- Dropping 3.5.x is 1.8G
+- Dropping 3.6.x is 1.8G
+- Dropping 3.7.x is 3.1G
+- Dropping 3.8.x is 2.6G
+- Dropping 3.9.x is 1.8G
+- Dropping 3.10.x is 3.4G
+
+### GSOC backend project
+- hard part is the design phase - how 
+    - moving backend function signatures from parameter list to data class of backend parameters
+        - in fonts, all the new feautures only available indirectly
+            - drawing text as last parameter takes text object
+            - measurement doesn't take text object so can't compute directly
+    - is likely a medium sized GSOC project, w/ a lot of testing work 
+    - @tacaswell has some concerns about this not having enough concrete upside for the cost of the code churn
+- alternative idea
+    - PR 30516 / issue [30515](https://github.com/matplotlib/matplotlib/issues/30515)
+    - how to push more "native" overlays into the backends to allow cross hairs and similar
+    - try out new scheme for passing things into the backend using dataclasses as entry point to using the pattern in the backend
+
+---
+
+# January 15, 2026
+
+_attending_: [efiring](@QWhXj01mSwmTjk5kN1H_qQ), @ksunden, @QuLogic, @melissawm, @story645, @ayshih, @timhoffm 
+
+## Agenda
+- [x] RSE updates
+- [x] NASA updates
+- [x] 3.11 status
+- [x] review Albert's PRs
+    - [x] [Accuracy bugs with image resampling ](https://github.com/matplotlib/matplotlib/pull/30184) - sig changes post review
+- [ ] [name=timhoffm]review [blitting errors in {check, radio} button widgets](https://github.com/matplotlib/matplotlib/pull/30945)
+
+## Notes
+### RSE updates
+- Melissa
+    - getting re-onboarded
+    - discovering build issues
+    - thinking about triage team and how to get recruitment going
+- Elliott
+    - writing statement of work
+    - waiting on responses from @tacaswell to start contract process
+    - alt-text is almost done, but needs some finishing touches
+- Kyle
+    - still waiting on contract
+    - little bit of issue review
+    - setting up hi-dpi monitor
+- Tom
+    - little mpl work, swamped with other things
+
+### NASA
+- contracts are moving 
+
+### 3.11
+- still holding on Antony's PR
+    > https://github.com/matplotlib/matplotlib/pull/30059?    
+    - need to check one last thing and merge
+    - need to check do a scan thourgh one final time to make sure that nothing changes by too much
+- https://github.com/matplotlib/matplotlib/pull/30945
+- 
+### image resampling
+- grid lines and pixel lines not lining up w/ non-affine transforms
+- this PR loosens tolerances to avoid updating baseline images, gets resolved by font prs 
+    - text changes b/c of change in computation of kernal used in text rotation
+    - 20 images had to be changed, about 40-45 loosened for text & noted with TODO
+- [30824](https://github.com/matplotlib/matplotlib/pull/30824) b/c bivariate was using agg interpolator for non-viz purposes, agg takes shortcuts
+
+### radio/check buttons
+- merged 
+- work left in whats new & todos: not all widgets work properly after swapping canvas
+    - blitting should be better encapsulated, hindered by how state is stored in widgets
+        - store if user wants blitting and if canvas supports blitting 
+    - use case is standalone figure that might get attached to canvas
+
+---
+
+# January 22, 2026
+
+_attending_: [efiring](@QWhXj01mSwmTjk5kN1H_qQ), @melissawm, [@anntzer](@pXw4hSgTQF2--OciPYwa1w), @story645, @ksunden, @tacaswell, @scottshambaugh, @timhoffm , @QuLogic 
+
+## Agenda
+- [x] RSE updates
+- [x] NASA updates
+- [ ] 3.11 status
+
+## Notes 
+
+### RSE updates
+- Tom:
+    - minimal work, in operation/project
+- Melissa: Created a doc for discussion around triage team: https://hackmd.io/@matplotlib/r1g8K31Ubx
+    - [ ] Melissa to create a subcategory on discourse for maintainers to add/nominate people to the triage team
+    - [ ] Melissa to figure out a way to add AI use disclosure/information on docs
+    - [ ] Caswell will investigate GH group permissions / external contributor for triage
+- Kyle:
+    - still waiting on contract
+    - mostly review/issue
+
+### NASA
+- slow progress
+- final report for previous roses is due in April
+
+### 3.11
+- [#30059 drop intermediate buffer](https://github.com/matplotlib/matplotlib/pull/30059)
+    - blockers: 
+        - design decision is wrong, properly would be to have intermediate buffer in different position 
+        - slight issue around positioning fraction bars when drawing math
+            - want to draw fraction bars exactly on pixel, requires controlling snapping better 
+                - accept snapping b/c improvement over fuzzyness
+                - bars are slightly too low/high/wrong position - fix is figure out if floor or ciel rounding
+                - bar positioning issue is independent of buffer 
+            - is here to avoid regenerating all the fraction images
+                - maybe kick fraction image regenerating down the road
+        - ok w/ edge case of overlapping glyphys? 
+            - yes
+        - consensus: should go in, put buffer positioning and fraction bar on "nice to have" stack
+
+- [#30974 widget blitting](https://github.com/matplotlib/matplotlib/pull/30974)
+    - not sure if problen is in MNE tests, widget blitting code, or base canvas API (vs Gui canvas API) 
+        - fire events where handler forces a draw that asks for a renderer that the base canvas doesn't have
+        - MNE is holding a canvas reference that calls for a non-existent canvas
+
+## other PRs
+- [#30974 right click context menu](https://github.com/matplotlib/matplotlib/pull/30976) 
+    - turn off able b/c other tools provide context menu
+        - general API figure.add_context_menu + callback
+    - squeeze into toolbar? 
+        - add hamburger to toolbar w/ additional tools likes context menu
+ -[#30980 log axes on 3d plot](https://github.com/matplotlib/matplotlib/pull/30980)
+    - needs reviews, would fix oldest open bug!
+
+---
+
+# January 29, 2026
+
+_attending_: [efiring](@QWhXj01mSwmTjk5kN1H_qQ) @ksunden @QuLogic @tacaswell 
+
+## Agenda
+- [x] RSE updates
+- [x] font discussions
+
+## Notes
+All PR review