Your twitter archive

after three separate times I've finally managed to get dracut to resume for me

turns out, the module is only enabled if there's a swap /partition/ which is a PROBLEM if you're resuming from a fucking swap /file/.

Not sure what to make of the dev responses though?

 Tue Feb 04 21:50:04 +0000 2020

This patch (or removing the check entirely) – https://github.com/nabijaczleweli/dracut/blob/8c1064e1b5dc20004e065e69b578c7d6836e02a7/debian/patches/resume-file – fixes the issue by also including the module when swap files are active.

Granted, seeing as this check is only performed in host-only mode, I can't see why one wouldn't check if /sys/power/resume != 0:0, but

 Tue Feb 04 22:54:21 +0000 2020

update: they just got back from putting her down

beats a slow renal-failure-induced death ig but still :(

 Wed Feb 05 14:07:42 +0000 2020

well, that's uhhh not gonna work all that well, now is it?

 Sat Feb 08 16:45:31 +0000 2020

baby (2x)

 Sun Feb 09 22:21:07 +0000 2020

baby (2x)-ng

 Thu Feb 13 15:39:46 +0000 2020

replacement wlan antenna covers since the manual said they are not to be shipped unprotected apparently?

 Thu Feb 13 17:27:58 +0000 2020

I made a Debian package from scratch today, went exactly how you'd think – impenetrable from the outside since nothing actually explains what the build process is supposed to be, but now that i get the gist, I'm forgetting what the roadblocks actually were, so I can't fix that :/

 Fri Feb 14 04:31:55 +0000 2020

anyway, paclage https://foreign.nabijaczleweli.xyz/debian/README :)

 Fri Feb 14 16:16:38 +0000 2020

volatile binds like const, so it'd be `(volatile uint32_t) *` and `(uint32_t *) volatile`

 Fri Feb 14 23:22:10 +0000 2020

the useful one

 Fri Feb 14 23:40:19 +0000 2020

PoC (implemented the other way around because only stripping cv-qualifiers is an error): http://coliru.stacked-crooked.com/a/64221e475fe7926e

 Fri Feb 14 23:49:55 +0000 2020

RT @thecoshman: I still find it so odd that I accidentally designed a good tool and been lucky enough that @nabijaczleweli implemented it.…

 Mon Feb 17 00:55:50 +0000 2020

pir-8-emu (and BearLibTerminal, which was a bigger adventure) now available as a Debian package: https://foreign.nabijaczleweli.xyz/debian/README

 Wed Feb 19 17:00:00 +0000 2020

im honestly blown away by how good of a tool gbp-pq is, and not just usable, it's actually *helpful*??

granted, the documentation is dogshit (expectedly, but how much it misses the point is quite something) so it dissuaded me from trying to use it for a big long time, but still

 Thu Feb 20 23:09:18 +0000 2020

reasonable idea? horrible hack?

 Sun Feb 23 20:47:30 +0000 2020

well.. hm

 Mon Feb 24 19:11:19 +0000 2020

good news: it's reproducible
bad news: i3-wm on sid is 4.17.1, which is too old to report a bug against upstream

 Mon Feb 24 19:44:34 +0000 2020

ive sent the maintainer a nice email on the bugtracker lets ho[e he doesn't yell at me >_<

 Mon Feb 24 22:09:21 +0000 2020

new ISA, featuring more bad decisions, but at least you can actually implement strcpy() to port (left; WIP, missing ADDC to do more than 255) without having to self-modify the code area with the LOAD IND argument (right; current) so that's an... improvement? i guess??

 Tue Feb 25 20:27:31 +0000 2020

idk the more i think about this the stupider it sounds but like.. i feel like a little bit of the "soul" or whatever of the architecture was lost today.. now it reads closer to a 6502 with its 80%-LDA/STAe code than a.. Different Thing™?

 Tue Feb 25 21:19:16 +0000 2020

it's ALIVE

 Tue Feb 25 21:46:35 +0000 2020

that's 655 bytes, with minimal modifications to the assembly! couldn't do *this* before! (well you could I guess but it was too much of a garbage-fire even for me)

 Tue Feb 25 22:59:37 +0000 2020