Tom

@devioustree
7 Followers
13 Following
31 Posts
A few years ago I designed a way to detect bit-flips in Firefox crash reports and last year we deployed an actual memory tester that runs on user machines after the browser crashes. Today I was looking at the data that comes out of these tests and now I'm 100% positive that the heuristic is sound and a lot of the crashes we see are from users with bad memory or similarly flaky hardware. Here's a few numbers to give you an idea of how large the problem is. 🧵 1/5
Xcode 26 LLM Markdown Summaries Are Actually Useful for Humans

Xcode 26 has a couple of tech docs that are intended to help LLM’s generate modern solutions to coding problems in Swift, but that also contain valuable info for us human readers.

Christian Tietze

@andrew @mhoye @siracusa This reminds me of recently trying to convince Claude Code that iOS 26 exists - necessary for debugging why a crash happened in iOS 18 but not 26.

After finally getting it to accept the online sources, it agreed - for roughly two messages. Then it pivoted to using "the lack of iOS 26 events" as evidence for its original theory, completely ignoring the actual crash I was debugging (iOS 18 crashes, iOS 26 doesn't - seems relevant!).

The unprompted 😊 did not help

Always nice to have options
brb buying a new MacBook in the settings app
@siracusa I'd highly recommend this French sci-fi animated movie. Great animation, darkly funny, and tackling some interesting themes https://m.imdb.com/title/tt26915336/
Mars Express (2023) ⭐ 7.5 | Animation, Action, Mystery

1h 28m | Not Rated

IMDb
How I sometimes feel when I read Apple's documentation:

Are you f* kidding me, Apple?!

After a long time, I filed another bug report using Feedback Assistant because the bug was bad enough that it’s worth the effort of writing it all down.

When uploading a sysdiagnose (or probably any other attachments) you get the usual privacy notice that there is likely a lot of private and other sensitive info in those log files. It’s not a great feeling but it is what it is with diagnostic data and I mostly trust the folks at Apple to treat it with respect and I trust the Logging system to redact the most serious bits.

However, when filing a feedback today a noticed a new addition to the privacy notice:

"By submitting, you […] agree that Apple may use your submission to [train] Apple Intelligence models and other machine learning models."

WTF? No! I don’t want that. It’s extremely shitty behavior to a) even ask me this in this context where I entrust you with *my* sensitive data to help *you* fix your shit to b) hide it in the other privacy messaging stuff and to c) not give me any way to opt out except for not filing a bug report.

Do you really need *more* reasons for developers not to file bug reports? Are the people who decided to do this really this ignorant about the image Appleā€˜s bug reporting process has in the community? How can you even think for a single second that this is an acceptable idea?

So, WTF, Apple?!

The hardest problem in computer science is trying to figure out how any Apple API works by just reading the docs.
A18 should be pronounced ā€œeeeeeightingā€