Pavel Machek
Posts
1765Following
90Followers
167
Pavel Machek
pavel
"If it's max, not a pax". It looks like 737MAX saga is continuing, this time it is not computers. It looks like door flew of a plane at 14000 feet, vas-av has a good video on youtube.
0
0
0
@drewdevault Yep, that's ugly. Would like to see that one fixed. (But if you can open /proc/self... you can usually also go to the original file and open that; and that's why it survived for so long).
0
0
0
Pavel Machek
pavel
@jiska First, thanks for nice talk at 37c3. I don't have an iPhone, but catching IMSI catchers is cool. OTOH it is also too late -- by the time you analyze your logs, IMSI catcher already has your position logged.
Early gsm phones (like Nokia 5110) had service menus with capabilities such as 'lock to specific cell tower' which could be used to prevent that -- just lock your phone to "known good" cells. Have you seen similar functionality in modern modems?
Early gsm phones (like Nokia 5110) had service menus with capabilities such as 'lock to specific cell tower' which could be used to prevent that -- just lock your phone to "known good" cells. Have you seen similar functionality in modern modems?
1
0
2
@drewdevault NixOS may not be for you, but I believe it has its uses. Would not use that emoticon...
0
0
1
Pavel Machek
pavel
@martijnbraam @linmob I never seen machine with ARM SBBR. Yes, that's a step in right direction, I wish PinePhone (etc) would be using that. (And it would be nice to have something similar for R-V, preferably without complexity of ACPI.)
1
0
0
Pavel Machek
pavel
@martijnbraam @linmob Not sure, and have not met those uefi-only laptops, fortuantely. But I'm sure they for example power down when battery goes low? Don't require vendor-patched u-boot of particular version to init their DRAM? Can charge battery without vendor-patched Linux kernel successfully booted? I really wish there was standard arm64 platform (and standard risc-v platform).
1
0
1
Pavel Machek
pavel
@thatprivacyguy One possible solution would be to start a local proxy and make chromium use that...
1
0
0
Pavel Machek
pavel
@martijnbraam @linmob While x86 hardware can be pain, it usually does not have problem booting bootloaders or charging its batteries. I expect RISC-V to be improvement over x86 CPUs, and it would be cool to have some kind of next-generation platform. We should not really need 100 drivers for RTC.
1
0
1
Pavel Machek
pavel
@slightlyoff It is somehow important small&light phones keep existing. (And I don't mind "cheap"). Usually I'm either on horseback (where 180gram smartphone breaks easily) or carry notebook (where I already have decent screen I want to use, thank you). I played with Geeksphone Keon, and that one is almost right size and non-evil. I wonder if I should get one of those kaiOS things.
0
0
0
Pavel Machek
pavel
@AsteroidOS So... There likely is place for Linux smartwatch, but do you really want big-enough watch so that they can run Linux for weeks? Smartwatch is good reason for good & small system that is smaller than Linux (in particular does not need MMU and can run in less than 256KB RAM). See https://www.espruino.com/ , I like that one.
0
0
1
Pavel Machek
pavel
Trying to do debayering on CPU using vector extension. So far targeting x86-64 and pinephone... and so far it is few times slower than naive approach. Is there good high-level overview of avx and neon somewhere? Any hints? Anyone willing to help?
0
0
0
Pavel Machek
pavel
@hrw @haeckerfelix I'm afraid there way too little video clips with permissive-enough licensing...
1
0
1
Pavel Machek
pavel
"Recently" terminals started beeping when mouse enters/exits the window. It seems like running emacs remotely then terminating the connection does that. How is this ... annoying glitch called and how do I get rid of it? Debian 11.8 with mate.
0
0
0
Pavel Machek
repeated
repeated
c3voc releases
c3voc_releases@chaos.socialBreaking "DRM" in Polish trains has been released on media.ccc.de https://media.ccc.de/v/37c3-12142-breaking_drm_in_polish_trains
1
8
2
@drewdevault "Perfect mainline Linux support" for a SoC is cool. But even better would be create compatible hardware, so that Linux, Windows, DOS, Xenix, ... can be ran on the hardware without needing any support. PCs in 486 era used to be like that, everything compatible with everything else, and it would be good to get there again with Risc-V. Asking not for compatible instruction sets, asking for compatible platform.
0
0
1