> I will say that I’ve seen a lot more vehement trash talking about mypy and gushing about pyright than vice versa for quite a few years. It doesn’t quite add up in my mind.
agreed! mypy's been good to us over the years.
The biggest problem we're looking to solve now is raw speed, type checking is by far the slowest part of our precommit stack which is what got us interested in Ty.
Thanks Astral team! We use Pydantic heavily, and it looks like first class support from Ty is slated for the stable release, we'd love to try it.
While we wait... what's everyone's type checking setup? We run both Pyright and Mypy... they catch different errors so we've kept both, but it feels redundant.
> But what exactly is the value in having humans grovel through logs to isolate anomalies and create hypotheses for incidents?
Agreed! I think about this using Weakly's own reference to "standing on the shoulders of giants."
To me, building abstractions to handle tedious work is how we do that. We moved from assembly to compilers, and from manual memory management to garbage collectors. That wasn't "deskilling" - it just freed us up to solve more interesting problems at a higher level.
Manually crawling through logs feels like the next thing we should happily give up. It's painful, and I don't know many engineers who enjoy it.
Disclaimer: I'm very biased - working on an agent for this exact use case.
The ‘with evidence’ part is key as simonw said.
One anecdote from evals at Cleric - it’s rare to see a new model do better on our evals vs the current one. The reality is that you’ll optimize prompts etc for the current model.
Instead, if a new model only does marginally worse - that’s a strong signal that the new model is indeed better for our use case.
This website is pure nostalgia <3, pretty much every corporate page in the 90s looked like this! I was half disappointed they're using <div>'s and CSS for layout and not a bunch of <table>'s.
“But drugmakers also faced changes to the Medicaid rebate program that would have likely cost them hundreds of millions of dollars each if they didn’t lower their list prices.“
This was a fascinating read for a totally unexpected reason.
I’ve spent most of my life in countries where the metric system is used for distance, weight, temperature etc.
This year I’ve had to travel a lot to the US for work and found the constant mental conversions a PITA. I kept wondering why people keep holding out against such an obviously easier system.
Then I read this article about 8 hours of sleep would be 3.33 metric hours. How you wake up at 9:50 after sleeping at 2:75 and I notice real-time at the absolute recoil I feel reading this. Maybe I’m getting older , but I completely get how familiarity to numbers being represented a certain way is hard to let go of.
you're right, its a few scrolls down. It used to be right at the top, so it became a learned behaviour for me to go their site and quickly check the cost of a transfer versus what my bank offers me.
I noticed an immediate loss of functionality via this redesign. The previous home page let me calculate quickly how much it would cost to do a transfer - arguably the PRIMARY value of the service.
now it’s replaced with two CTA buttons to sign up etc.
Chile has a great system which guarantees a free bank account linked to your national ID called Cuenta RUT. It has some limits like only a debit card and a max value you can store there but I think it’s a fantastic idea.
You just need to walk in to any branch with your ID and you’re all set with an account you can receive and send payments from. If you need something more from your bank account - it stands to reason you have the necessary documentation to apply for a ‘regular’ bank account which most do.
Even foreigners with any kind of work permit get this ID called a RUT and are eligible.
I was wondering who else uses FDB beyond Apple and Snowflake. Found that CouchDB is rewriting their storage layer to use FoundationDB for their v4 release. That said I couldn’t find much info on progress or release dates.
fair. just to clarify though - i find the expectation to be always online / checking work email at all hours also a sign of a toxic workplace. definitely could have phrased my original comment better.
I know this isn’t the spirit of your comment but this is one of the use cases of VR that I’m most looking forward to.
We’ve invested a LOT into immersing people into past worlds - think Jurassic park for dinosaurs, plenty of TV and movies about Roman times etc. VR would make all that seem so dated and boring.
Work up a sweat running from a T-Rex - quick shower and off to work!
To most people, it's not Nomad vs Kubernetes - it's a choice between Nomad vs Managed Kubernetes.
All major cloud providers offer a managed kubernetes service at minimal added cost to running the worker VMs yourself.
With managed Kubernetes, the simplicity question is no longer obviously in Nomad's favour. As other comments allude to, Kubernetes as a user is pretty easy to master once you get used to its mental model.
I don't think this is a great example. The Objective-C to Swift transition going smoothly is only because of Apple's almost total control of the ecosystem (I mean this positively).
Apple is automatically the loudest voice in the room for iOS development. If they embrace Swift, the writing is on the wall for Objective-C. It's not just sticks, I'm sure they also put a lot of effort into making the transition as easy as possible.
In SQL - there is no equivalent. I think a better example is x86.
It took a generational new form of computing (mobile) to give ARM the momentum to seriously challenge x86. It's been almost 15 years since the original iPhone and we're only now seeing ARM based processors in computers.
SQL IMO is EVEN harder to displace than x86. x86 has a massive ecosystem but only only two serious manufacturers. SQL has a similarly massive ecosystem AND is the de facto language of the vast majority of major databases. Going to be very hard to unseat.
agreed! mypy's been good to us over the years.
The biggest problem we're looking to solve now is raw speed, type checking is by far the slowest part of our precommit stack which is what got us interested in Ty.