Skip to content

Commit 09ba361

Browse files
Create 2023-03-09.md
1 parent 0e0a3c6 commit 09ba361

File tree

1 file changed

+71
-0
lines changed

1 file changed

+71
-0
lines changed

meetings/2023-03-09.md

Lines changed: 71 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,71 @@
1+
# TWG 2023-03-09
2+
3+
[Last meeting's notes](https://github.com/haskellfoundation/tech-proposals/blob/main/meetings/2023-02-09.md)
4+
5+
## Present
6+
7+
* David Thrane Christiansen
8+
* Laurent P. René de Cotret
9+
* Hécate
10+
* John Ericson
11+
* davean
12+
* gershom
13+
14+
## Apologies from David
15+
16+
## Updates on Projects
17+
18+
### Error message index
19+
20+
* Dark mode
21+
* New contributors
22+
* David has not yet created the SSH keys
23+
24+
### Advisory DB
25+
26+
* Waiting for Fraser
27+
* Feedback: we should be doing this with a mail forwarder on haskell.foundation rather than an individual's mail for anything official
28+
* FWIW Google makes this seem harder than it is
29+
30+
## Open Proposals
31+
32+
### Standard Library Reform
33+
https://github.com/haskellfoundation/tech-proposals/pull/47
34+
35+
- Need buy-in from CLC and GHC folks before we can formally vote on this proposal.
36+
37+
- Not sure how to get 'approval' from stakeholders.
38+
39+
- David likes that the success criteria is bite-sized, rather than all-encompassing - having one portable definition moved to `base` from `ghc-base` and one non-portable definition hidden from `base` shows a template for incremental improvement.
40+
41+
- 1 person-month? John sees it as an upper bound. The PR exists and needs finishing.
42+
43+
- Gershom: If there is time left over, can do follow up work like moving portable code back to base.
44+
45+
- Laurent asked whether we could plausibly fund this. David says "yes".
46+
47+
- Gershom complimented John on the resultant document - he feels comfortable sending it on.
48+
49+
- Hécate will contact CLC
50+
51+
- Laurent will e-mail GHC-devs about the proposal to get buy-in from Matthew Pickering, Ben Gamari, and Simon Peyton Jones.
52+
53+
### IDE/Test Integration
54+
https://github.com/haskellfoundation/tech-proposals/pull/46
55+
56+
### Maximally decoupling Haddock and GHC
57+
https://github.com/haskellfoundation/tech-proposals/pull/44
58+
59+
Students are interested as part of HSoC. No updates to discuss, but we're leaving it open due to student interest.
60+
61+
## Other Business
62+
63+
[GHC and MTL version discussion](https://github.com/haskell/core-libraries-committee/issues/136)
64+
65+
It would be nice to do just one breaking change at a time. John wants to hear our thoughts about having multiple bindists for different build configurations.
66+
67+
The big problem comes about when packaging systems like Nix and Stackage pin a specific global version.
68+
69+
Having bindists for lots of different build configurations will lead to difficulty diagnosing bug reports.
70+
71+
John will attempt to write a draft of a specific, narrow proposal for discussion on tick-tocking breaking changes to libraries and compilers. In other words, a major GHC release would (as much as possible) not change any library dependencies at all with respect to the prior release, and then the next major release would have library changes but no compiler changes.

0 commit comments

Comments
 (0)