<quiliro>i was used to doing everything on firefox. then i discovered that firefox included some non free stuff. so i went with iceweasel, then i found that guix had emacs in which you could do [almost] everything. so i tryed but could not do some things like check email and browse the web and watch videos and listen to audios and open pdfs and odt files
<reepca>Hm, it occurs to me that the reason that it's fine for the C++ daemon to unconditionally remount everything as MS_PRIVATE is because it's in a separate mount namespace (so anything that assumes its mounts have a certain propagation type outside of the container is unaffected). But run-container makes that optional. Perhaps it'd be better to just copy what useful parts I can get from call-with-container instead of trying to refactor it.
<ng0>in other words, you have to package python-runtime
<elreydelaswasas1>trying to install the guixsd operative system i firstable intorduced the command: loadkeys la-latin1 , the i introduced the comand ifconfig -a to see all my networks and i choosed my wired network eno1 with LINKENCAP Ethernet hwaddr 08:2ESF:02:EC:06 and i choosed it using the command ifconfig eno1 up then i entered the command dhclient eno1 then i used the command cfdisk and i created a Linux filesystem in /dev/sda1 with 20G,
<elreydelaswasas1>continuation of what i said:"inux swap with 2 G in /dev/sda3, i have 8 gb of ram, a BIOS boot in /dev/sda4 with 513M, a Linux home partition with 493M in /dev/sda5, a Linux home partiton with 908.5Gb in /dev/sda6 in /dev/sda6 then i executed the command herd start ssh-daemon, the i ejecuted the command mkfs.ext4 -L my-root in /dev/sda1 then i ejecuted the command cryptsetup luksFormat /dev/sda1 , then the command cryptsetup op
<ng0>it should be cryptsetup luksOpen /dev/sdX foo in case that wasn't a typo and your next message will show the error :)
<quiliro>how can i search for a log on this channel but i do not remember what day it was? it was recommended to me to install the appropriate packages for installing that program
<quiliro>i want to check which they were...i remember mentioning openmolar
<quiliro>i think it was rekado which made the suggestion
<quiliro>i want to install the deps for the source of openmolar1
<elreydelaswasas1>continuation of what i wrote:"hen i executed the command mkfs.ext4 -L my-root /dev/mapper/my-partition, then i ehecuted the command mount LABEL=my-root /mnt, when i was partitioning the disk with cfdisk isaved the changes with the Write option before i quit the program with the quit option , after i ejecuted themount LABEL=my-root /mnt command i executed the mkswap /dev/sda3 swapon /dev/sda3 then i executed the command herd s
<janneke>elreydelaswasas1: it is more helpful if you start with your question
<elreydelaswasas1>mkdir /mnt/etc, then the command cp /etc/cnofiguration/desktop.scm /mnt/etc/config.scm, then i executed the command nano /mnt/etc/config.scm, and in the archive of config.scm i made the next changes , in in the script that started with (use-package-modules i changed what it said with (use-package-modules xfce ratpoison enlightenment certs), and in the in the script that begined with (packages (cons* i changed it to (packages (
<elreydelaswasas1>top environments and in the script below to this below the first o letter of the word rat poison i wrote nss-certs and below the last letter of the first enlightenment word i wrote;and continue writing for HTTPS access in the below script after i wrote the firstcerts word next to certs but below it i wrote %base-packages)), and in the script that said with (bootloader (grub-configuration (device "/dev/sdx")) i wrote instead of
<elreydelaswasas1>forget that i really mean i wrote instead of it (bootloader (grub-configuration (device "/dev/sda"))), then i executed the command guix system init /mnt/etc/config.scm /mnt and then appeared the next errorguix substitute: error: TLS error in procedure ´read_from_session_record_port´: The TLS connection was non-properly terminated, in the next line i wrote: substitute: updating list of substitutes from ´https://mirror.hydra.g
<quiliro>elreydelaswasas1: it is still confusing because it is incomplete
<jsierles>i think he's pasting from google translate
<wigust>%load-path must be the same for guix offload to work `user@host$ guile` AND `ssh user@host guile`, doesn't it? `user@host$ guile` has much more PATHS. `ssh user@host guile` has unexisting PATHS.
<wigust>Is it because I need to offload to root@host?
<wingo>well, it is bedtime for me. however i think i might have the issue. there are some times in which we use the stack in an O(n) way. this should be fine; guile can make a stack like this, no problem, if your problem is O(n) anyway then it is better on the stack than on the heap.
<wingo>we mark the stack precisely. that means that we have a mark function that goes frame by frame to see what values are actually gc values in that frame (and not, e.g. values which have a slot allocated but are actually dead, etc)
<wingo>however with libgc, they expect that mark functions do a constant amount of work, not an O(n) amount of work
<wingo>i.e. that calling a mark function will push a constant number of objects on the mark stack
<wingo>but with an O(n) stack, that's not the case. i think what happens is that when a mark phase starts, it starts with a certain stack size; then if too much marking happens, the mark stack "overflows" then actually libgc will fail the whole thing, expand the mark stack, and try again
<jlicht>quiliro: guix environment creates a new shell, so you either need to feed the rest of your commands via stdin (complicated, not really useful), OR you just first type `guix environment --pure guix', and then when you see the shell with [env], run the rest of the commands