You're telling me a point fixed this theorem?

ok I guess this could theoretically make sense if all access went through some privileged daemon, then you basically have a microkernel architecture

allwinner did not do that

Show thread

like what is even the point of kernel drivers then. just have all applications open /dev/mem and go wild

Show thread

wow allwinner's "official driver" really just mmap'd the hardware into userspace huh

the thing can DMA to whatever address you give it....

Show thread

jade todo: fuck around with sunxi-cedrus driver (the driver for hw video encode/decode on allwinner chips, including pinephone)

looks like nobody has implemented h264 encoding yet, even though it's mostly reverse engineered

maybe my real contribution will be making voice/video rooms actually work without needing to be full mesh (and without needing to run webshit)

Show thread

probably not anytime actually soon.

but I will read obsessively about how the protocol works, how webrtc, ice, rtp, etc. work, and what interfaces I'd need to use to get gstreamer to encode/decode webrtc, and hook that into a gtk application (answer: all the hard parts are done, and have gobject bindings, so I can get bindings in haskell or some other nice language)

Show thread

hmmm today I will make a matrix client (<- clueless)

I need to make a collection of mathematicians describing their own work as inflicting pain on readers.

I guess, your constructive definition of a != b might contain some information that's needed to construct an inverse. E.g. for power series, it could tell you the first coefficient for which a and b differ, which you'd need to know to constructively define 1/(a-b).

Show thread

Ok so apparently one way to solve this is to have an *inequality* relation that explicitly comes with your structure. Then you can make the condition "if a != 0, then a is a unit". And I guess inequality might have a more subtle definition than just the negation of equality...

Show thread

saying something like "assuming LEM, the real numbers are a field" just sounds fucking comical to me

Show thread

ok so I think my next procrastination sink will be trying to formalize random bits algebraic geometry in HoTT using agda

first time to read a bunch of shit about how algebra even works constructively. like how do I even define "field". if it's like, "all elements are either 0 or invertible", then the real numbers aren't a field, because checking if a number is 0 isn't decidable????

wow I can't believe I ever thought I knew what "equals" meant

what a fool

Show thread

I do not understand univalence in the slightest but apparently it means I can stop giving a fuck if things are "only isomorphic" and treat them like they're equal? That sounds cool and is basically how everyone already thinks about math I guess

Show thread

oh great finally getting around to reading that HoTT book now that I have other things to procrastinate

I feel like I'm missing a lot of the motivation since I don't know anything about higher homotopy nonsense...

lmao why did I think it was a good idea to put a first name as my username

Ok so I just learned these things are called "capabilities" and lots of systems like this exist, some modern and some from like 1970.

I want to make one...

Show thread

vrms sending me a monthly nag via cron that the docs for emacs, gcc, and gdb are non-free is hilarious.

Show older

cybrespace: the social hub of the information superhighway jack in to the mastodon fediverse today and surf the dataflow through our cybrepunk, slightly glitchy web portal support us on patreon or liberapay!