thealistra

The kerning in clickonme in the example screenshot is unfortunate - I would change the font

Exactly - so far seems like a windows vista “are you sure?” Modal. Are we missing something here?

Seems like somebody is not accepting that every successful project will grow and become unwieldy like this. This is all legacy backwards compatibility of all iterated ideas that now you have to support.

The ocr example says it recognizes Chinese, but output ignores it - maybe just AI bug in generated examples

You should be happy that you have such a financially aware IT dept. The machines are truly expensive and after managing a farm of 150 Mac minis at work I can tell why.

It is like 3 days running the EC2 buys you a Mac mini? And you can only rent the machine in day increments

Can’t really see it nicely on mobile

US only it seems?

My point is can you write a json deserializer with this, where allocation of every child is defacto optional, depending on input JSON?

Isn’t this a very restrictive way to write? Hard for me to imagine as never wrote with such annotations so no idea how viable it is for a large codebase to have this constraint.

I understand how this works in general. I had static analyzers at Uni, I know lattice theory and all this - I am just wondering how Xr0 handles it.

What happens if a function allocates not deterministically, like

if (turing_machine_halts(tm)) return malloc(1); else return NULL;

How is this handled?

For the bootstrap c lexer and parser was hand rolling really necessary? Lex and yacc exist for a reason

Bro no one will use a web service and trust it is 100% private. This is an unprovable claim. Maybe it is private now, but you sell this in 5 years and it stops being so.

Maybe a JavaScript ad jailbreaks an iframe and reads this.

There is no beating a locally running program

But most CIs allow flaky tests :)

Isn’t this like argparse from Python for typescript?