Saagar Jha

@harshil Which one is that

Also I feel like more judges need to read Hacker News because people will actively try to present a “quirky, weird project” that someone literally posted there like a month ago and just not get called out on it

Back in my day if you delivered the 20th variation of the same idea at least you could argue you spent the time learning a new tech stack or something. If you are obviously having AI do all the implementation what did you spend the last 48 hours doing exactly

Also I feel like more judges need to read Hacker News because people will actively try to present a “quirky, weird project” that someone literally posted there like a month ago and just not get called out on it

Not to sound like a thought leader and all but I was reviewing a bunch of hackathon projects and the fact that everyone is slopping them out now (code, frontend, slides, literally everything honestly) really highlights how ideas limited people are

Back in my day if you delivered the 20th variation of the same idea at least you could argue you spent the time learning a new tech stack or something. If you are obviously having AI do all the implementation what did you spend the last 48 hours doing exactly

Also I feel like more judges need to read Hacker News because people will actively try to present a “quirky, weird project” that someone literally posted there like a month ago and just not get called out on it

What I don’t get is how Mail is able to make this San Francisco but Passwords somehow ends up with Helvetica for both its collapsed and expanded state subtitles

@saagar wouldn't be surprised if this were hardcoded based on bundle ID

Of course if you read the code that is exactly what the method does before calling -[UINavigationItem setAttributedSubtitle:] but surely this wacky behavior is intentional and must have gone through some sort of review, right?

What I don’t get is how Mail is able to make this San Francisco but Passwords somehow ends up with Helvetica for both its collapsed and expanded state subtitles

@saagar wouldn't be surprised if this were hardcoded based on bundle ID

Usually I’d point at this and go “haha obviously someone just made an NSAttributedString without configuring it” but like it’s been this way for a while and this is the most basic use of -[UINavigationItem setSubtitle:] so now I’m not so sure this is a bug or not

Of course if you read the code that is exactly what the method does before calling -[UINavigationItem setAttributedSubtitle:] but surely this wacky behavior is intentional and must have gone through some sort of review, right?

What I don’t get is how Mail is able to make this San Francisco but Passwords somehow ends up with Helvetica for both its collapsed and expanded state subtitles

@saagar wouldn't be surprised if this were hardcoded based on bundle ID

Hey CoreSimulator team? Can you please get some CI for status bar overrides to make sure they actually work? This is like the third time you’ve broken it

FB22539899 (STStatusBarData.applyDictionary(dict:) does not copy over date information for status bar override)

Can you imagine how embarrassing it must be to be a Siri engineer right now? Your SVP asks for more GPUs to make the thing actually work and gets fired for it. All your coworkers hate you for making 10% more than them and then the guy who made The Pact sends you to slop school

So far I think I keep hitting it on updates but I want to test this in a VM or something and I don’t really know if they even have a keychain

@saagar I think they'd have to have a keychain, but what I'm skeptical about is whether they have a concept of "data that is not available because the device is locked".

Ok so I did some digging around for https://federated.saagarjha.com/notice/AxUQzSbs3YIHd1HEK8 and I strongly suspect it’s Safari coming up while the device is locked, which means the broker does not load its “SecureDefaults” (keychain-backed storage). Does anyone know how I might be able to simulate this

So far I think I keep hitting it on updates but I want to test this in a VM or something and I don’t really know if they even have a keychain

@saagar I think they'd have to have a keychain, but what I'm skeptical about is whether they have a concept of "data that is not available because the device is locked".