lofty

@lofty@fieldprogrammable.gay

logic synthesis abyss domain expert - frangible lut slut - caprotti valve gear and kylchap exhaust enjoyer - gives people tail dysphoria - profile art by twitter.com/mikifeuille - eggsite: https://cohost.org/lft

endorsements:
"some days, your posts terrify me" - @ChiralTheAlien

feel free to say hi; I like making friends.

This profile is from a federated server and may be incomplete. Browse more on the original instance.

eniko, to random
@eniko@peoplemaking.games avatar

me, having 5 mins ago publicly said im taking it easy for the next two days: feeling guilty about not working hard enough

lofty,

@eniko see, I read this, and my immediate reaction is "you're gonna have to spend the money you get from kitsune tails on therapy to recover from developing kitsune tails".

lofty, to random

I've been having a lot of fun playing Rabbit and Steel.

Too bad I suck at it though.

danderson, to random
@danderson@hachyderm.io avatar

Back on my bullshit in a limited capacity, doing some therapeutic stupid stuff. Today: which of the 7400 series chips are still being made at a vaguely reasonable price? Because I have Plans, and they involve quite a bit more than just logic gates and flip-flops.

lofty,

@danderson @petrillic if only someone wrote a Yosys script to convert logic to those weird chips.

lofty,

@danderson That's the skeleton of it; I was very new to Yosys when writing it, and I think I'd do things very differently these days.

GENLIB is so much easier to use than Liberty

Maybe if I have a moment I can write an example GENLIB file for this.

lofty, to random

Seems like my goal at work is to add a bunch of metrics to nextpnr.

So, is there any information people would like to have available?

lofty,

@azonenberg

It's on the list, and it shouldn't be too hard to add. (Currently it has a heat map of routing by wire type)

azonenberg, to random
@azonenberg@ioc.exchange avatar

Anybody else ever look at a distributed-memory architecture diagram and start humming "Dragostea Din Tei" to yourself?

lofty,

@azonenberg MMIO hii, MMIO huu, MMIO hoo, MMIO ha-ha

gsuberland, to random
@gsuberland@chaos.social avatar

holy shit I can just see the aurora in Nottingham! that's wild.

lofty,

@gsuberland sorry, I've been living a Red Arrow away from you for how long?

lofty, to random

I am doing my best to be too professional to vaguepost, but I really, really want to.

lofty, to random

at some point I'm sure I will accept that other people perceive the things I do as witchcraft just as much as I perceive the things others do as witchcraft.

lofty, to random

where's a set theoretician when you need them.

lofty,

@ChiralTheAlien oh, you have no idea.

lofty,

@ChiralTheAlien I should add this as an endorsement :p

lofty, to random

yawns
Morning fedi

lofty,

@ChiralTheAlien morning chiral, and thank you

lofty, to random

taxes done.

now to curl up in a corner and pretend the numbers are fine.

lofty,

@azonenberg not so much when you realise you included some figures from this FY in the tax return for last FY and have to file an amendment.

danderson, (edited ) to random
@danderson@hachyderm.io avatar

TIL there is an "abstraction of electricity" offense in the UK. It's just stealing electricity. I choose to disregard this and interpret it literally. London bobby, at the police station:

> It is alleged that you sir, on or about the third of this month, did willfully and with malice aforethought, model the behavior of an electrical carrier, to wit this wire 'ere, using Maxwell's equations. Thus BLATANTLY creatin' an abstract mathematical model of reality. This is a very serious matter...

lofty,

@danderson there was no law against phreaking at one point, so in a desperate attempt to discourage it, phreakers got charged with abstraction of electricity instead.

astrid, to random
@astrid@fedi.astrid.tech avatar

evil hatsune miku: aa ii aa

lofty,

@eris @astrid and mikumiku_ebooks would be about actual ebooks

lofty, to random

I'm an abandonware radicalist; not only should abandoned software have its source released, abandoned hardware should be opened up to development to stop it being ewaste.

danderson, to random
@danderson@hachyderm.io avatar

As I write more rust, it's interesting which ingrained habits are harder to break, after many years of writing go.

I keep camelCasing stuff and making the linters mad at me.

When reading about unsafe and pointers, I had a default wariness of doing things that might conceal the pointer from the GC. In particular reading about packing marker bits into pointers gave me strange feelings until I unpacked why.

lofty,

@danderson eeeh, modifying pointers like that still isn't exactly kosher in rust, but for different reasons than a GC.

lofty,

@danderson I came to Rust from OCaml. Rust has Unsafe Rust. OCaml has Deadly OCaml.

For example, std::mem::transmute is dangerous, but the OCaml equivalent of Obj.magic might as well be called segfault because the GC might get creative ideas about your data.

"oh, you have an even integer? let's treat that as a pointer"

lofty,

@danderson (this is why OCaml has 31 and 63-bit integers, which are fine, up until they're not)

lofty,

@danderson

The compiler handles a Pointer converted to a uintptr in the argument list of a call to a function implemented in assembly by arranging that the referenced allocated object, if any, is retained and not moved until the call completes, even though from the types alone it would appear that the object is no longer needed during the call.

I sure do love me some compiler magic

danderson, to random
@danderson@hachyderm.io avatar

Hmm, in rust, with a generic type Foo<T>, is it possible to provide two implementations of PartialEq, one for if T doesn't implement PartialEq and another if it does? Can I just write out two impl blocks, one with a bound and one without, and have the compiler pick the right one?...

Arguably that behavior would be too weird, and being an internal helper type the non-bounded implementation actually does what I want at all times, I think... But I'm still curious if that's possible!

lofty,

@danderson the others mentioned specialisation, but I'm going to raise a different point.

what does it even mean to test for equality if T does not implement PartialEq? the only meaningful implementation would be to always return None, which has little practical use.

  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • JUstTest
  • thenastyranch
  • magazineikmin
  • ethstaker
  • khanakhh
  • rosin
  • Youngstown
  • everett
  • slotface
  • ngwrru68w68
  • mdbf
  • GTA5RPClips
  • kavyap
  • DreamBathrooms
  • provamag3
  • cisconetworking
  • cubers
  • Leos
  • InstantRegret
  • Durango
  • tacticalgear
  • tester
  • osvaldo12
  • normalnudes
  • anitta
  • modclub
  • megavids
  • lostlight
  • All magazines
    •