Matt Massicotte

macOS/iOS ⌨️, outdoors 🏔, justice ⚖️, games 👾

Independent developer. Previously: Crashlytics, Apple

Public Key

npub1xfkfzq9r68c7e5xnt0u4jzexx0nvy239m6as32k8n4d0jqaaseeqskwjhv

NIP-05 Address

[email protected]

Profile Code

nprofile1qqsrymy3qz3aru0v6rf4h72epvnr8ekz9gjaawcg4tre6kheqw7cvuspz4mhxue69uhhyetvv9ujumt0wd68ytnsw43quu0h03

Publishing to

relay.mostr.pub

Author Public Key

npub1xfkfzq9r68c7e5xnt0u4jzexx0nvy239m6as32k8n4d0jqaaseeqskwjhv

Show more details

Last Notes

2024-06-27 00:10:38

Thinking about feedback I've received about stuff I've made. It was true, I agreed with it, but it was delivered unkindly. It had this quality of intentional hurt. Years later, still effective.

2024-06-19 20:51:43

- reply

@npub17cs…fsmn - Determine line range (kind of a pain)
- get for attributes for that range
- check them for the special attachment thingy

2024-06-19 15:03:18

Wow! Fisker Ocean and Lucid Air sighting within 1 minute of each other.

2024-06-18 11:49:33

I find it incredible when I hear “but it’s popular!” or “people want it!” as justification for anything. Have a look back at other things that have been popular throughout human history. Not a great track record.

2024-06-15 00:24:41

- reply

@npub17cs…fsmn Are they submodules?

2024-06-15 00:00:55

- reply

@npub17cs…fsmn I don’t use it, but the answer must be yes right? They are local!

2024-06-13 15:23:30

- reply

@npub15ml…2w4s ahh a great one! I will give you some hints in the gist.

2024-06-13 13:47:53

- reply

@npub17cs…fsmn I have two big outdoor trips planned for this summer! One will be my first multi-day adventure 😅

2024-06-13 13:42:30

After all my writing about Swift concurrency during the week, blogging about it on the weekend, and answering questions about it in-between here, its really nice to take a break and listen to podcast about Swift. Balance just feels right, you know?

2024-06-13 12:26:28

@npub15ml…2w4s I heard on the show that you have had some trouble with Swift concurrency and your apps. I’d love to know about what’s given you problems and if you have any specific questions!

2024-06-12 17:55:05

Checked out https://www.swift.org/migration yet?
Lots of stuff in there about understanding and adopting the Swift 6 language mode. Would love to hear your feedback! You can also check out the DocC sources, see executable example code, and file issues/contribute here:
https://github.com/apple/swift-migration-guide

2024-06-11 15:06:41

Logger is now Sendable 😮‍💨

2024-06-11 13:54:48

- reply

@npub1src…fms4 These were my top two wishes and I would have bet money neither would ever arrive.

2024-06-11 13:53:19

- reply

@npub1src…fms4 I stumbled across it in the Xcode release notes

2024-06-11 13:50:38

I still cannot get over it. Xcode getting EditorConfig support AND directory-based compilation. I could cry.
Everyone using local packages take note. Static libs are far more powerful, and their biggest downside is now gone.

2024-06-10 22:25:35

🫨🫨🫨
friends, it has been 25 years
https://files.mastodon.social/media_attachments/files/112/594/199/078/518/418/original/ebe78005f0ab487d.png

2024-06-10 02:33:53

- reply

@npub1u24…32pl 💯💯💯 Amazing stuff!
And I am deeply honored 🥺

2024-06-09 13:28:50

Concurrency in Swift 6
SE-0423: Dynamic actor isolation enforcement from non-strict-concurrency contexts
You wouldn't know it from the title, but this one will make your code more concise!
https://www.massicotte.org/concurrency-swift-6-se-0423

2024-06-07 12:38:58

Concurrency in Swift 6
SE-0414: Region based Isolation
This is a big one, folks. Region-based isolation is amazing. It makes so many problems just disappear.
https://www.massicotte.org/concurrency-swift-6-se-0414

2024-06-06 20:18:43

- reply

@npub1src…fms4 The internals for sure. I was actually thinking more UI stuff. Do you think that could work?

2024-06-06 19:22:37

- reply

@npub1src…fms4 This is also something I think could make for an amazing piece of shared infrastructure!

2024-06-05 12:51:40

- reply

@npub17cs…fsmn You can buy more, and they are cheap. Don’t even hesitate.

2024-05-30 02:21:13

Here’s an open source phenomenon I cannot stand.
You propose a change, and maintainers agree it makes sense. But! This now makes them realize some other, related-but-independent stuff also could be better. So now they want you to expand the PR to fix that other stuff too before they accept.
PR author gives up. Everyone loses.

2024-05-29 23:48:13

- reply

@npub1src…fms4 Wow you are making progress! I saw that completion view too 👀

2024-05-24 14:27:28

- reply

@npub17cs…fsmn Absolutely not! Have you never used it? It’s amazingly good!

2024-05-24 13:55:36

On the topic of great apps, the Canada Post app is an absolute delight.
I would love to understand how, exactly, such a large government organization managed to pull this off. Was it subcontracted? If you know anything please tell me! I want to point all my representatives to it.

2024-05-21 01:50:00

Wrapped up my first official day at Apple again*, working on the forthcoming Swift concurrency migration guide. Gotta admit, so many parts of this I did not see this coming.
* as a contractor this time, relax

2024-05-10 18:20:44

- reply

@npub17cs…fsmn You used a $ in an unexpected place I bet

2024-05-04 14:48:27

Ah Saturday. A time to relax, recharge, and frantically invent a presentation you for some reason agreed to do in 36 hours.

2024-04-23 23:45:53

- reply

@npub1kvt…0vtn I haven’t played with Text much at all and SwiftUI. But I think it’d be interesting to try to make a text editor component with it. Just to see how it would work out.

2024-04-17 12:01:05

MainActor.run is almost *never* the right solution. You want to define your isolation statically on the type, via a global actor annotation.
Dynamic isolation is an escape hatch, and it should set off alarm bells every time.

2024-04-12 12:01:50

I hope Xcode 16 does include some new editor-level features to help with concurrency.
But I bet per-line actor annotation will not be one of them. It will never change for synchronous spans. More than one actor can be involved in an await. That’s just off the top of my head.
I deeply sympathize with the struggle, but I don’t think this is a viable or even desirable solution.

2024-04-09 17:26:40

Concurrency in Swift 6, part 1
SE-401: Remove Actor Isolation Inference caused by Property Wrappers
I mentioned this one the other day. It's the only Swift 6 feature you can turn on now.
https://www.massicotte.org/concurrency-swift-6-se-401

2024-04-08 12:04:34

You have written about swift concurrency. But it was years ago, and now it has problems. Totally understandable!
But high rankings in search engines. People trust you. They are getting confused. They are *introducing crashes* into their apps. You have to go and fix these posts. Concurrency is hard enough.
And posts *about concurrency* in 2024 without strict checking? This is not understandable.

2024-04-06 12:00:15

I have found myself frequently swapping between TextKit 1 and 2 for debugging. It’s a huge pain. So I made a thing!
Small (but growing!) set of tools for working with TextKit that are convenient, efficient, and work for both versions.
https://github.com/ChimeHQ/Glyph

2024-04-05 14:34:05

It can be very hard to make classes Sendable. But when it’s hard, it’s also a sign it is not the right solution.
The alternative, find out why you want this to be Sendable in the first place and change *that* might require rethinking how you are using concurrency. And that could be harder, but it’s where you should start.

2024-04-02 00:57:08

- reply

@npub1src…fms4 Go all platforms!

2024-04-01 19:01:15

- reply

@npub1src…fms4 I am will try to track you down there!

2024-04-01 02:55:24

- reply

@npub1src…fms4 Right here

2024-03-31 03:35:34

- reply

@npub1src…fms4 We may also be able to cook something up for tree-sitter languages. I’ve been SPM-ing them for a while now, and can be a better match when you only need a few. I can take a look at GDScript for this too.

2024-03-31 00:33:13

- reply

@npub1src…fms4 I’d be very into working on some shared UI components for this work if that’s of any interest! I’ve traditionally only done macOS stuff in this area, but I’m trying to branch out.

2024-03-27 13:09:45

I think all SwiftUI Views should be MainActor. If anything, this post has reinforced that position.
But, this is still interesting, refreshingly in-depth, and definitely worth a read!
https://mastodon.social/@fatbobman/112164316465957881

2024-03-14 11:48:09

I'm not great at SwiftUI. But I've been inspired by a conversation with @npub16xq…x72u . Is this any good? Seems to work and is a fairly tidy way to apply arbitrary async-algorithm operators to values from the UI.
https://gist.github.com/mattmassicotte/2737e6de5398d890e6de468fc0f51d76
https://files.mastodon.social/media_attachments/files/112/093/645/231/994/845/original/bf95b7399d04d053.png
https://files.mastodon.social/media_attachments/files/112/093/645/212/526/680/original/4b6fa1e0f8a3be9f.png

2024-03-12 14:24:08

- reply

@npub17cs…fsmn https://github.com/mathialo/bython

2024-03-12 13:06:04

I’ve been thinking more deeply about the upcoming Swift 6 changes. Just a huge amount of additions, tweaks, and improvements to concurrency. It’s going to be almost unrecognizable. So many new patterns will be possible - my head is spinning 😅

2024-03-10 13:12:20

Using actors with NSObject-inheriting protocols is quite tricky. I'm not sure I'd recommend it, but I came up with a solution that at least works in a common case.
https://github.com/mattmassicotte/ConcurrencyRecipes/blob/main/Recipes/Protocols.md#non-isolated-nsobjectprotocol-inheriting-protocol

2024-03-08 14:31:23

- reply

@npub17cs…fsmn Swift 5.8?

2024-02-26 14:35:48

- reply

@npub1svs…p66u Way to go!
I am, though, seeing a new bug with notifications. They all now send a generic message, regardless of content or type.

2024-02-24 12:36:34

News!
I’m helping @npub163h…pflh with an upcoming Swift concurrency migration guide for swift.org!
The goal is to provide documentation around concepts but also practical solutions to problems real apps encounter.
Ultimately this is going to be a community effort. To start, if you have ideas or problems you’ve run into please send them my way!

2024-02-08 19:34:26

The other day I mentioned that protocols have a drawback around isolation requirements that functions can avoid. I took a moment to write up an example:
https://github.com/mattmassicotte/ConcurrencyRecipes/blob/main/Recipes/Protocols.md#solution-4-dont-use-a-protocol

2024-02-06 19:26:33

I think my favorite corporate move is "here is a white paper we commissioned that says we are very good"

2024-02-05 13:41:23

- reply

@npub1svs…p66u Huh interesting - thank you!

2024-02-05 13:36:30

@npub1svs…p66u 👋🏻 found an interesting bug! The following post doesn’t produce a preview for the included link, but for an at-mention (which happens to also not exist)
https://iosdev.space/@hybridcattt/111878638306543652

2024-01-26 14:32:00

- reply

@npub1can…0fm9 …. I don’t think it is!

2024-01-26 14:29:35

- reply

@npub1can…0fm9 Here’s an option that I think, while involved, is at least worth consideration
https://blog.thomasdurand.fr/story/2023-12-14-dont-ship-api-keys/

2024-01-25 19:24:37

- reply

@npub1kvt…0vtn “David excels at hiding all the time he wastes” I kid I kid 🤣

2024-01-18 22:49:36

- reply

@npub17cs…fsmn You define it in Xcode build configurations? I don’t know of a way to do this.

2024-01-18 13:28:26

- reply

@npub17cs…fsmn Yes that would be great! Perhaps next week?
As for this issue, I should clarify the problem more. Self is captured by the Task and that is not allowed *unless* the type is marked with a global actor.

2024-01-18 12:56:12

Welp, I just submitted a Swift pitch draft. I really want more flexibility for non-Sendable types, and it just *feels* like it should already work this way.
Internet, you may now begin telling me I am wrong.
https://forums.swift.org/t/isolation-assumptions/69514

2023-12-29 14:21:38

Speaking of open source drama: I was made aware recently of yet another place where code I’ve written was copy-pasted into another repo. Presumably this was to “avoid a dependency”.
For many years, I have used a permissive license. I’m going to be addressing this in 2024. I have yet to find a license that fits my needs, so I’m working on one.
https://github.com/mattmassicotte/cosl

2023-12-06 20:10:35

What is it called when you have a solution to an unimportant problem that works perfectly but you are for some reason just completely philosophically opposed to using it and moving on to the actual work?
Programming. It is called programming.

2023-11-27 16:57:09

If you have any interest in the open source work I do and Discord is your thing, I have a server going now: https://discord.gg/WpEmEyMc
I've found it to be a great way to help people live. Also good for discovering areas that require more documentation, which has been a focus!

2023-10-24 14:29:59

- reply

@npub1can…0fm9 @npub17rh…2g9f I have used this one. It’s ok, but it does feel like the API isn’t quite right.
https://github.com/kishikawakatsumi/KeychainAccess

2023-10-13 03:25:23

Here’s a rule of thumb that has served me well:
When you think a problem is hardware, it’s probably software
When you are certain a problem is hardware, it’s definitely software

2023-10-12 00:21:25

- reply

@npub1kvt…0vtn Admittedly this was a super long time ago, but when I was there I used Xcode and everyone around me was always like “so you’re the guy that uses Xcode huh?”

2023-10-09 16:56:57

- reply

@npub1kvt…0vtn Ahh. In that case maybe it’s just being used under the hood by some other system. If you’re curious, try setting up a breakpoint on its designated initializer!

2023-10-09 16:45:47

- reply

@npub1kvt…0vtn Oh wait, you don’t use core data in this app?
Could be coming from an Apple framework. You all share the same process for better or worse. It is true this might not be your bug, but I still think zombies is a good first try.

2023-10-09 16:43:29

- reply

@npub1kvt…0vtn Yes, so first this is not running a fetch result, it is *deallocating* one. What you are seeing here is a classic overrelease effect. There could be more than one, unfortunately, that's the nature of heap corruption. You want the *cause*, which can be pretty separated in time. But I bet it is related to your core data models somehow.
You are right that dangling pointers are really unusual in Swift. But, regardless, start checking out zombies!

2023-10-09 12:11:14

- reply

@npub1kvt…0vtn Crashes in an auto release pool usually point to an overrelease. Can I see more of the stack trace?

2023-09-22 12:31:33

Does anyone understand why I don't get a warning here? I'm explicitly capturing a non-sendable value in a `@Sendable` closure and I have all the warnings on.
https://gist.github.com/mattmassicotte/d0b2b4eddf7c0c2ca7bdf3d516bb5f86
https://files.mastodon.social/media_attachments/files/111/108/329/266/906/594/original/875d63c15f4fc3ce.png

2023-09-10 18:18:25

- reply

@npub1265…jcry oh man outage analysis must bring up loss aversion all the time! Are there any other biases you talk about?

2023-09-10 16:50:32

I think a lot about dependencies. I'm not saying I love them. But I do *absolutely love* finding a library that does what I need. I also love collaboration via open, shared packages. Strikes me as optimal in so many ways.
But many don't. People just plain hate dependencies!
Got burned once? Just *heard* about *someone else* getting burned once? That's powerful. In fact, it has a name: loss aversion bias. Never thought about it this way before though.

2023-09-01 15:43:02

Can I share with you my new favorite blend: Beaker and Swift concurrency!
My talk from @npub16dy…asrq is now available: https://www.youtube.com/watch?v=HqjqwW12wpw

2023-09-01 10:45:15

- reply

@npub1t5m…9guy often very clever stuff!
But the author is unfortunately… a very problematic person… let’s say.