<balrog>
Interesting, one of the variants will have "Functional Safety and Class B Safety Certi cations"
<carl0s>
it seems to have only 12UDBs, but i can't find if those are the same arch than the psoc4 UDBs
<balrog>
it seems like it will have more hard logic blocks
<pointfree>
carl0s: where did you read that?
<carl0s>
there's a video on the second link i sent
<carl0s>
i'm looking for the TRM, but can't find it yet
<balrog>
carl0s: there isn't one yet
<balrog>
or a complete datasheet
<carl0s>
yeh, have to wait then, meanwhile think on a project for it :)
<pointfree>
UDB's are great, but I suppose fewer UDB's and more hard-ip means more product differentiation for Cypress.
carl0s has quit [Quit: Leaving]
<pointfree>
Although the PSoC 6 does have a lot of interesting new hard-ip's such as hardware crypto, wifi, ble... and a lot of space allotted to routing fabric.
carl0s has joined ##openfpga
<pointfree>
I'd be interested to know how the PSoC6 analog measures up to the PSoC5LP analog.
carl0s has quit [Read error: Connection reset by peer]
<whitequark>
in general, any type system with typeclasses is turing-complete
<whitequark>
azonenberg_work: to put it another way, having turing-completeness at compile-time is usually done because it is useful and with an understanding of what you're doing, rather than by accident
<azonenberg_work>
Yeah i just thought it was interesting
<azonenberg_work>
i mean C++ templates are turing complete too
<whitequark>
it's the same for C++ or Haskell or OCaml
<whitequark>
in C++ you can just do computations with constexpr
<whitequark>
in Rust you should soon be able to do the same
<whitequark>
no need for type system hacks
<whitequark>
in C the preprocessor can perform arbitrary computations :)
<azonenberg_work>
yeah
<azonenberg_work>
i saw an IOCCC prgoram once
<azonenberg_work>
that computed the primes up to... 1024, i think?
<azonenberg_work>
with the c preprocessor
<whitequark>
you're limited by the amount of unique tokens in the source
<whitequark>
but that's not fundamentally different than the artificial recursion depth limit in rust or c++
<azonenberg_work>
apparently the guy had to write his own preprocessor to actually test it
<whitequark>
(rustc also has a few natural recursion depth limits, i.e. crashes :P)
<azonenberg_work>
b/c GCC and most other common compilers would run out of stack space before fully evaluating the code
jeandet has joined ##openfpga
<azonenberg_work>
:p
<lain>
I once saw someone write a tesselation algorithm in C++ templates
<lain>
for demoscene magick
carl0s has quit [Read error: Connection reset by peer]
pie_ has quit [Quit: Leaving]
pie_ has joined ##openfpga
jeandet has quit [Ping timeout: 256 seconds]
<rqou>
offtopic: apparently the congressional budget office has evaluated GOPcare
<rqou>
<insert "dis gon be good" meme here>
jn__ has quit [Ping timeout: 264 seconds]
jn__ has joined ##openfpga
azonenberg_work has quit [Ping timeout: 260 seconds]