Hello, On Tue, 02 Feb 2021, Carlos E. R. wrote:
On 02/02/2021 13.50, David Haller wrote:
And BTW2: I have this set up:
==== /etc/init.d/boot.local ==== dbus-uuidgen > /etc/machine-id ====
(just symlinking machine-id to /dev/null sadly makes some apps barf)
Who the F came up with that machine-id thing anyway? deadrat again?
What is that thing for? :-? :-o
It's some identifier for your machine. Ah, yes, as I suspected: https://www.freedesktop.org/software/systemd/man/machine-id.html https://wiki.debian.org/MachineId or generally e.g.: https://www.startpage.com/sp/search?query=/etc/machine-id
What uses it?
Some apps do use it and barf if it's symlinked to /dev/null (i.e. contains nothing).
My "/etc/machine-id" file is dated 2012, and does not contain the same as dbus-uuidgen (I guess I know why)
dbus-uuidgen _generates_ an ID in the required format. Naturally, it'll be (or should be) different! As it seems I need _some_ machine-id (or stuff barfs), I generate a new one on every boot, and as I reboot regularly, that's ok. Otherwise, I'd setup a cron-job to generate one e.g. every hour. Hm. Actually, that's a good idea anyway, fuck the stuff that's using it :) Done: ==== /etc/cron.hourly/gen-machine-id ==== #!/bin/sh dbus-uuidgen > /etc/machine-id ==== *hrhrhr* I'll keep the one in init.d/boot.local, if just for messing with the intention...
HTH, -dnh, who also likes to read print first and foremost, but hasn't printed himself a page in the last ~15+ years, but had others print about 8 pages for him in the first ~7 of those and hasn't even had others print for him in about 8 years ;)
tsk, tsk.
I printed two pages this week.
That's really not much, I guess, so I applaud you :)
The health system here uses electronic prescriptions, but I need to print the PDF so that the pharmacy can scan the barcode and deliver my medicines :-p
What if your printer breaks when your supply has run out? You s'posed die or what? Or your smartphone or PC dies, or whatever you get the barcode with. Etc... And I'd hate to need to get a printer (or smartphone) just for stuff like that, as I do NOT need either. I *HATE* stuff like that without a fallback to "pen and paper" with a vengance! Think more Javier Bardem (you know with what) than Antonio Banderas ;) I think we need to put the "deciders" of stuff like that through situations like a broken printer/smartphone/internet etc. pp. To hammer in the _need_ for an analog, pen and paper, fallback. Repeatedly. Probably. And don't get me started on US elections without a paper trail using _known_ flawed (just broken and/or astonishingly easy hackable) voting machines (there are none other, anywhere) (and no, I'm not and never was a US resident, but it just boggles my mind what is done in some states). Some states do not even save a (anonymous but traceable[1]) record of the votes. Basically the analogue of a printout of a good old fashioned on-paper-vote... Some states fail even on that. It boggles the mind. And lo and behold, back here in good-ole-Germany (I've got rants alot about hereabouts too), we vote Sundays and voting closes at 18:00, and on paper and still manage to get most (officially "preliminary", but they usually turn out to be the same) results by 20:00, or sometimes maybe 23:00 for a few random voting-districts, but final offical votes are usually in on Tuesday after the vote on Sunday but usually don't differ in any meaningful way. So, generally, we know the results by Sunday 20:00 and can go to bed ;) Compare that to e.g. the last US presidential election even in states where _only_ machines were used... *barf* (Yes, I know you're in Spain ;) -dnh [1] I try to say: a legitimate vote for candidate A was cast by a legitimized voter _and that information_ is traceable (not who the voter was). IIRC Pennsylvania, maybe some other state, has failed on this big time in so many ways ... -- if (user_specified) /* Didn't work, but the user is convinced this is the * place. */ linux-2.4.0-test2/drivers/parport/parport_pc.c