#0.15 Release Crew

1087 messages · Page 2 of 2 (latest)

warm tinsel
#

nop, failing because some of the commits do have typos

spark brook
#

extra ]

image.png
uneven snow
#

The list contains only my old handle 🥲 (irate-devil)

main sparrow
#

looks like a bunch of the listed PRs in the big ol' list are also not actually 0.15 PRs, I think

tight crystal
#

Is the migration guide meant to still be a draft?

uneven snow
#

No..

weak tinsel
#

"Contributors" is noisy as it is populated by anyone that contributes any change to a PR (including type fixes). It is intentionally excluded. If someone deserves credit, add them to the authors list

spark brook
#

ah...

cinder cave
#

i didnt know this was autogenerated

spark brook
#

Yeah

cinder cave
#

how are PRs selected and grouped?

spark brook
#

It's based on the milestone iirg, and also a C-needs-release-notes tag or smth

cinder cave
#

yeah but some entries have like several PRs

#

like required components has 3 PRs listed

#

(i feel like it was more)

warm tinsel
#

it's grouped manually

cinder cave
#

if its grouped manually, how are authors and contributors combined

#

is it by PR?

#

so like if a PR was made by a person, and its grouped in, then they will be an author

warm tinsel
#

I suppose, but I didn't do it myself

cinder cave
#

whats the point of the contributors list if its not displayed

#

might as well not have it

warm tinsel
#

I don't know, I haven't fully followed that conversation, but I think it was a kinda last minute change

cinder cave
#

im not on the contributors list either

green kayak
cinder cave
#

oh its by my old handle for some reason

#

thats annoying

main sparrow
#

Yeah I was also pretty sure I listed pcwalton and Jan as contributors for BRP as well but only my handle shows up there 🤔

cinder cave
pine breach
#

in 0.14 contributors and changelog sections came from the toml files (which came from the PRs), the 0.15 post doesn't use the toml data, it's already rendered in the markdown 🤷‍♂️

rigid oak
#

In general, I feel like we should prefer trivial false-positives to false-negatives

#

it feels worse to be excluded by accident then to have perhaps less central authors share a bit of the credit

muted spear
#

imo, if one person made 10 line changes, they should be labelled a contributer

rigid oak
#

several very important people seem to have been missed.

muted spear
#

and included in the blog post

main sparrow
#

Yeah I was just assuming that the toml with the sections would be used to generate the contributors/authors for each section, but maybe that wasn't the case 🤔

gusty flint
#

I was correctly listed as an author with custom cursors but I'm not in the contributors list

shell jasper
pine breach
#

the post is not using the zola shortcodes though (unless you meant that there was something borked and we opted out of the shortcodes)

shell jasper
#

Don't be shy about claiming it

#

Yeah, I think we should list contributors in the headers too

shell jasper
#

We can change that convention, but it seems silly to have metadata that we totally hide

cinder cave
#

wheres the long contributor list

#

it uses my old handle

#

idk why it does that, last couple releases used the right handle

shell jasper
glossy willow
weak tinsel
shell jasper
weak tinsel
#

We already sort authors by size of their contribution. I don't think we need categories here, and it increases the "informational noise" to introduce them.

weak tinsel
quartz cosmos
#

I fixed the typos, but I definitely don't want credit 😆

weak tinsel
#

FYI I used the wrong changelog, contributors list, and contributor count in my rush to get the release notes out :/

#

Sorry yall!

#

Peoples' concerns that the number was too low were well founded

#

294 instead of 195!

#

(fix is deploying now)

quartz cosmos
#

nice! Cue the 300 memes

rich osprey
real swift
#

Congratulations on the 0.15 release!

#

Suggestion for the migration guides etc. in the future:

  • Migration guide is mandatory to be written so it contains all the necessary information, before the merge of a pull request.
  • When a pull is merged, the migration guide part from the pull description is automatically copied as a .md in the right directory on the website.
  • If the pull is tagged as needing a release note, when the pull is merged, the pull description (sans migration guide) gets copied as a release note .md automatically as well. This is mostly a placeholder, but it's easy to start writing the real thing from the description.
  • A build of the website with drafts included is automatically created and updated and distributed online with some trivial authentication.
    Should reduce the pain at release time?
languid pelican
real swift
green pike
crystal rock
# real swift Suggestion for the migration guides etc. in the future: - Migration guide is man...

There's a problem with writing migration guides at the time of the PR - by the time that the release actually happens, the guide may be out of date or completely wrong. Take this example from the 0.15 notes:

The logical_rect and physical_rect methods have been removed from Node. Use Rect::from_center_size with the translation and node size instead.

When this was written, Node component had a size() method, but that was later changed to be on ComputedNode.

shell jasper
real swift
#

Yeah, my first proposal was that the pull description is the place to edit implementation guides as well... but that has the problem that only pull author and maintainers may edit it. So converting it to a file at merge time allows tweaking the contents by anybody (via pull requests) until release.

#

Another alternative would be to not directly add the file, but to just add a pull request with that file only. Then most migration guide pulls might remain open until close to release possibly, one pull per migration guide entry. However, metadata needs to be handled differently then and overall it's probably a worse way of doing it.

shell jasper
#

Did you read my discussion about merging the repos? Or the 0.15 retrospective discussion? 🙂

real swift
#

I glanced at the discussion about merging the repos. Personally, I'm not a fan of that idea, but you know better than me. And I haven't read through 0.15 retrospective yet.

finite plank
#

Many thanks to all of you for getting 0.15 out the door 🙏

#

Just wanted you to know that your effort is appreciated

shell jasper
#

Same from me! The small fixes and help with the notes is incredibly valuable

acoustic hedge
# shell jasper I disagree with this TBH: I've been using this for minor authors and excluding t...

The contributors metadata for the notes is supposed to be used by editors of the release notes to add significant contributors to the authors list manually, this was implemented after false positives where folks whose contributions were considered too minor for the author field, like singular typo corrections or what not, were added to the authors and considered an issue.

We should probably add comments above each contributors field saying something like add major contributors to authors field.

Or just rework it so the templates use it as a separate list of different level to authors.

shell jasper
acoustic hedge
main sparrow
#

Yeah, I think that true coauthorship is relatively uncommon, but it would be good to acknowledge when it actually happens (e.g. BRP)

shell jasper
acoustic hedge
#

How useful would it be for me to go thru the Migration Guides and polish / edit them at this point? (Currently going thru the process of migrating, and finding things in the guides that could be improved).

warm tinsel
#

A lot of people probably didn’t update yet, so probably useful. And from past experience migration guides still get looked at when they’re not the latest one

acoustic hedge
#

Okay, cool. I'll probably start working on that after work then (and a lot more on the weekend).

green kayak
shell jasper