@jingibus.bsky.social
Android person at Cash App. http://www.billjings.net
I'm not saying that nobody can ever cowboy a library up and have it be awesome. But if you want to learn from *this* example?
Writing a shitload of code would not have been sufficient by itself. It takes years for an overnight success, as they say β years of collaboration and contribution
Square has/had pathologically complex needs β that's why Anvil came out of that shop. Selling Square on Metro was huge for being an Anvil successor, and it required migration to be seamless in complex integrations.
You can't do that alone. You have to live in each other's heads to ship that.
There are projects like git that just wipe the slate clean and expect people to learn things a new and different way. But a dagger+anvil successor was never going to be like that β Android is so mature that migrations can be truly hellish.
08.11.2025 07:17 β π 1 π 0 π¬ 1 π 0Second, temperamentally Metro is the opposite of a DJB-style one person "I'm going to do this, but right" show. It is jam packed with APIs and decisions that are there to make contributors and consumers happy and successful.
08.11.2025 07:14 β π 1 π 0 π¬ 1 π 0Knowledge comes through work. I learn that every time I start a writing project β you just have to write code and see what works and what doesn't. Ain't no easy way
08.11.2025 07:12 β π 1 π 0 π¬ 1 π 0I say that because I had seriously talked about the idea of working on a dagger2 successor myself. But could I have done it? I am delusional, as a rule, but almost certainly no β b/c I know nothing about the compiler side, I just have API opinions
08.11.2025 07:11 β π 1 π 0 π¬ 1 π 0First, he didn't just waltz in. I was ranting about the need for a dagger2 successor for a while on social media; Zac, otoh, had been doing the grunt work of helping to maintain Anvil for ages
08.11.2025 07:03 β π 1 π 0 π¬ 1 π 0This is fantastic and @zacsweers.dev deserves all the praise but it also is gonna give you a false idea of what Z is like as a maintainer and author, what the project is like, and how you might achieve something like this yourself
08.11.2025 07:01 β π 22 π 7 π¬ 1 π 2Or use dinosaur rules and just eat the smaller car
08.11.2025 03:38 β π 3 π 0 π¬ 1 π 0Same experience here. It was something on the Embarcadero that invited engagement, not just observation.
07.11.2025 20:57 β π 1 π 0 π¬ 0 π 0Excited to share Metro's original design doc. Pulls back the curtain on a lot of the historical community feedback and discussions that went into it in its early days!
www.zacsweers.dev/metros-desig...
This is a valuable document to read in itself, but essentially:
* DHS is using face scanning tech that we technologies know is known bad
* You can't opt out
* DHS claims this supersedes any actual documentation you have as to your citizenship
This week's @unbreaking.org update on immigration is *also* an update on data security because the administration is using immigration enforcement as a way to trial and operationalize a full-on surveillance state.
06.11.2025 22:13 β π 74 π 43 π¬ 2 π 1We have to because this asshole has made himself synonymous with prop 50
05.11.2025 17:12 β π 0 π 0 π¬ 0 π 0If anyone tells Firebender that I'm a Karen, that will NOT stop me from calling management to complain
05.11.2025 01:21 β π 0 π 0 π¬ 0 π 0Semi-related, but this whole thing where people feel like they should be nice to their LLMs is baffling to me
I deserve to be treated better than this, and I am not ashamed to tell them so
I bet I can improve my current results by a fair amount, but β I might not do that.
Having the corpus of averageness at one's ease is.... well, maybe it's bad? Or maybe the things I cared about a lot just aren't as important as they used to be.
Coding with LLMs is exciting and fast in some ways, and profoundly demoralizing in other ways.
I've worked for years to write better coroutines tests (see here! www.youtube.com/watch?v=uk8k...), and it is just so much less work to write tests that I think suck.
Not only is it a series of videos for the month of November in which no effort has been made time
It's also time to play the @techconnectify.bsky.social hits
youtube.com/watch?v=DAX2...
They have an identity politics lens is what it is
30.10.2025 16:58 β π 2 π 0 π¬ 0 π 0The whole "white stripes are huge, seven nation army is played at football games" thing caught me way off guard
They were, like, white blood cells indie rock in my mind
Live from kingston harbor: www.youtube.com/watch?v=jVr7...
28.10.2025 18:17 β π 0 π 0 π¬ 0 π 0From a crewmember on yesterday's Teal 74 mission into now-Category 5 Hurricane #Melissa. As clear of an eye as you will see in the Atlantic basin.
27.10.2025 16:11 β π 2799 π 1056 π¬ 53 π 126Oh lawd, true vibe
27.10.2025 17:15 β π 2 π 0 π¬ 0 π 0It is difficult to maintain AI skepticism in the face of the impact it has on remote junior engineers.
I would rather our juniors learn the ropes in person with senior engineers, but.... they're not, at least not at our company. Having access to genAI tools has wildly improved their learning.
I think martian manhunter is the best, and it is only 6 issues, but the absolute lone running right now is quite good. Absolute Wonder Woman is delightful, and absolute superman is very good (konda like if there were an edgy dark superman who retains his essential goodness)
25.10.2025 13:25 β π 0 π 0 π¬ 0 π 0Have you read absolute martian manhunter
25.10.2025 01:38 β π 0 π 0 π¬ 1 π 0A left wing politics that is oriented around your rights and identity as an individual consumer instead of your responsibilities as a member of a social or political community is basically a left aesthetic pasted onto an internalized neoliberal ontology.
23.10.2025 20:10 β π 1912 π 342 π¬ 15 π 21I've been going down a stats + Android macrobenchmarks rabbit hole
Trying to write a blog post about that, in the meantime I'll share some raw thoughts here as a thread.
I've been focusing on benchmarks that capture frame duration metrics, e.g. developer.android.com/codelabs/jet...
#AndroidDev