|
operating system in french is "système d’exploitation"
|
# ? Sep 8, 2018 08:46 |
|
|
# ? Apr 24, 2024 23:34 |
|
Docjowles posted:Oh my god how have I never been aware of TempleOS before right now Did that dude really post to YOSPOS for a bit before getting trolled into oblivion? So the story goes. I've been wanting to read those YOSPOS threads for years now (I'm sure they're magnificent), so if anyone has a link, I'd be eternally grateful.
|
# ? Sep 8, 2018 19:45 |
|
Powered Descent posted:So the story goes. I've been wanting to read those YOSPOS threads for years now (I'm sure they're magnificent), so if anyone has a link, I'd be eternally grateful.
|
# ? Sep 8, 2018 19:51 |
|
TempleOS actually has some really clever features that make you question the implementation of things in Linux or Windows. https://tinyurl.com/y76p9fb2 at about 1:45 into the video as an example of what I'm talking about. He has images and videos built directly into the source code. The question really comes up, why can't we have images built directly into Terminals? Sometimes it takes a mentally disturbed person to break convention to allow others to notice can be improved. Throwing a theoretical example such as "ls -EP" to show thumbnails of all images in a directory inside the terminal. Unix/Linux was devised in the 70s to mainly handle math and text documents. 40 years later it's still functionally the same yet the type of files it is encountering has progressed vastly.
|
# ? Sep 8, 2018 21:52 |
|
YouTuber posted:TempleOS actually has some really clever features that make you question the implementation of things in Linux or Windows. https://tinyurl.com/y76p9fb2 at about 1:45 into the video as an example of what I'm talking about. He has images and videos built directly into the source code. The question really comes up, why can't we have images built directly into Terminals? Sometimes it takes a mentally disturbed person to break convention to allow others to notice can be improved. Throwing a theoretical example such as "ls -EP" to show thumbnails of all images in a directory inside the terminal. Unix/Linux was devised in the 70s to mainly handle math and text documents. 40 years later it's still functionally the same yet the type of files it is encountering has progressed vastly.
|
# ? Sep 9, 2018 01:23 |
|
Ranger+w3m has term support for displaying images as well.
|
# ? Sep 9, 2018 01:29 |
|
There's been pictures of Tux in the console boot process in Linux for as long as I can remember, and fbdev has been around for like 20 years?
|
# ? Sep 9, 2018 02:36 |
|
The issue has never been whether it could be done, it's more that people realized they don't really give a poo poo and the feature dies on the vine.
|
# ? Sep 9, 2018 02:37 |
|
When you put it that way... I can't believe there hasn't been at least one person who'd stop at nothing until the output of "ls -l" included a high-res still with correct subs from every anime in their collection.
|
# ? Sep 9, 2018 02:45 |
|
Last year there was a Suse employee who was working on a kernel module to make the boot process be all graphical, no more text lines (even if modules would print them), and was using fbdev. After people berated him for not using DRM/KMS he kinda gave up. No idea how far along was he, but still, it's a bit sad.
|
# ? Sep 9, 2018 03:13 |
|
Images in terminal was one of those “neat but why” things until I saw Rails added inline screenshots of failures while running browser tests
|
# ? Sep 9, 2018 19:44 |
|
YouTuber posted:TempleOS actually has some really clever features that make you question the implementation of things in Linux or Windows. https://tinyurl.com/y76p9fb2 at about 1:45 into the video as an example of what I'm talking about. He has images and videos built directly into the source code. The question really comes up, why can't we have images built directly into Terminals? Sometimes it takes a mentally disturbed person to break convention to allow others to notice can be improved. Throwing a theoretical example such as "ls -EP" to show thumbnails of all images in a directory inside the terminal. Unix/Linux was devised in the 70s to mainly handle math and text documents. 40 years later it's still functionally the same yet the type of files it is encountering has progressed vastly. IPython has been around for 17 years and ELOG has been around for at least 10.
|
# ? Sep 9, 2018 20:40 |
|
I'm trying to get apache 2.4 to insert a header when acting as a proxy for a webapp. The equivalent nginx config is below, the line that does the header injection being "proxy_set_header X-Real-IP $remote_addr".code:
Anyone done something like this before? (edit: just for clarification, the header needs to be visible to the webapp, not just something you can log in the apache logs, which seems to be the most common usage I see in the documentation) My Rhythmic Crotch fucked around with this message at 15:18 on Sep 23, 2018 |
# ? Sep 23, 2018 15:14 |
|
https://httpd.apache.org/docs/2.4/mod/mod_headers.html#requestheader should do it.
|
# ? Sep 23, 2018 16:15 |
|
How does snap work with distro-specific packaging solutions? Say I'm on debian and install gimp via dpkg, then I grab the latest gimp off snap. The latter is more up to date, so will it remove the dpkg version, or install it alongside? Also what about Debian menus, does snap use/respect those?
mike12345 fucked around with this message at 11:19 on Sep 24, 2018 |
# ? Sep 24, 2018 10:24 |
|
I don't about menus but it integrates with the operating system probably via a desktop file, so maybe? The snap is completely isolated from the OS (which is kinda the point) and it won't remove anything you have installed.
|
# ? Sep 24, 2018 12:59 |
|
lordfrikk posted:I don't about menus but it integrates with the operating system probably via a desktop file, so maybe? The snap is completely isolated from the OS (which is kinda the point) and it won't remove anything you have installed. Oh ok. I thought about how libraries are handled, but that explains it.
|
# ? Sep 24, 2018 15:19 |
|
Is Ubuntu 18.04 DNS configuration broken or what? I'm coming from a RHEL background and it took me like an hour to figure out how to set the loving default nameservers correctly on a machine at my buddy's work. You'd think setting the nameservers in the GUI would suffice, but apparently not. I eventually realized I could hardcode everything into /etc/systemd/resolved.conf but that can't possibly be the right way to do this, can it? I'm trying not to let my personal dislike of Ubuntu creep in here so I'll just assume that I'm doing something wrong.
|
# ? Sep 24, 2018 15:24 |
|
I ended up going with this, and then manually picking the injected data apart. For some reason I'm seeing something like "(null), 1.2.3.4" as the injected string. I think the built-in request handling in my stack was having problems with that... who knows.
|
# ? Sep 24, 2018 18:53 |
|
If I'm using network storage that's RAID 10, does it make sense to use a sync software like Borg or Attic to backup, or should I just go with rsync? I'm not sure I understand it fully, does it put the files in a regular file structure or something else?
|
# ? Sep 25, 2018 00:56 |
|
The XKCD Larper posted:If I'm using network storage that's RAID 10, does it make sense to use a sync software like Borg or Attic to backup, or should I just go with rsync? I'm not sure I understand it fully, does it put the files in a regular file structure or something else? RAID is not a backup so yes, always make sure you create backups. The complexity of what you’re trying to backup usually determines the strategy/tools. If it’s small/simple than rsync is usually good enough.
|
# ? Sep 25, 2018 11:52 |
|
LochNessMonster posted:RAID is not a backup so yes, always make sure you create backups. Will Borg create a full copy of the data on the same volume? Or will it have a folder structure that I can browse from SMB?
|
# ? Sep 25, 2018 17:15 |
|
The filesystem exists on top of the RAID10, which presents a block device. Whatever backup you take will look identical to what's on the filesystem/mount point
|
# ? Sep 25, 2018 17:58 |
|
Borg itself though creates archives. These can be mounted with a fuse driver if you want to browse them.
|
# ? Sep 25, 2018 21:29 |
|
We've recently upgraded a flaky PPTP VPN at work to OpenVPN using a dedicated pfSense box. It's working superbly with Windows clients and the pfSense admin interface allows you to create individual users and download a .zip with three files in it for passing out to co-workers. The .zip for user bob, for example contains: bob.ovpn bob.p12 bob-tls.key The ovpn and -tls.key files are text and the p12 is binary. It's no problem importing the OpenVPN config into a Windows machine and I've also managed it on my Android phone. We each have a unique password for our configs. I'm trying to put together a guide to importing the config into a Linux client and it seems that I have to convert some of these files (or one of them??) into .pem files. So far I've tried the following: code:
Mac verify error: invalid password? I also get this error if I try an arbitrarily made up password, like 'letmein'. Any ideas where I go from here?
|
# ? Sep 25, 2018 23:50 |
|
I got it sorted in the end. The solution was to use openssl to convert the p12 file into various pem files and then create an OpenVPN connection manually, taking settings from the ovpn setup file.
|
# ? Sep 26, 2018 15:59 |
|
apropos man posted:I got it sorted in the end. The solution was to use openssl to convert the p12 file into various pem files and then create an OpenVPN connection manually, taking settings from the ovpn setup file. A bit late to the party I see, but the p12 file is a storage container which contains the user cert, user oey and ca cert. openvpn should be able to habdle that (assuming you’re not running an older version of it).
|
# ? Sep 27, 2018 09:01 |
So I'm running into an issue where when I try to run some software, I get an error message saying that it can't load a shared library because it doesn't exist, however it definitely does exist.code:
|
|
# ? Sep 28, 2018 00:43 |
|
The library maybe doesn't like in your shared library path. I'm on mobile so it's hard to type code but check your LD_LIBRARY_PATH and maybe include that. Some binaries will check for a required library in their own folder, but if your path has the same one but a different noncompatible version it may not work.
|
# ? Sep 28, 2018 03:59 |
|
Shared objects can have dependencies themselves. Run ldd on that library and see what comes back.
|
# ? Sep 28, 2018 04:02 |
Thanks for the ideas!Jerk McJerkface posted:The library maybe doesn't like in your shared library path. I'm on mobile so it's hard to type code but check your LD_LIBRARY_PATH and maybe include that. libquadmath.so is on my LD_LIBRARY_PATH, and I'm pretty sure it was on my LD_LIBRARY_PATH when I compiled nh. code:
quote:Some binaries will check for a required library in their own folder, but if your path has the same one but a different noncompatible version it may not work. I think libquadmath.so is a standard GCC library, and I'm pretty sure there's no version of it bundled with nh so version compatibility shouldn't be an issue here. Vulture Culture posted:Shared objects can have dependencies themselves. Run ldd on that library and see what comes back. code:
|
|
# ? Sep 28, 2018 06:24 |
|
LochNessMonster posted:A bit late to the party I see, but the p12 file is a storage container which contains the user cert, user oey and ca cert. openvpn should be able to habdle that (assuming you’re not running an older version of it). That's handy to know. I guess it wasn't letting me just import the ovpn file because we've got a more specific setup for the work VPN or maybe it's just that some variants of Linux prefer the pem files. It's good to know about the p12 being a container file for future reference. Cheers.
|
# ? Sep 28, 2018 10:12 |
|
Any recommendations for enterprise backup and restore software that'll run a daily dump of a full OS (including partitions/tables) onto an Lto7 tape drive and then give the possibility of recovering from backup onto a Dell server by booting a recovery ISO over the network? We're trialling Acronis Backup at the moment and getting it to recognise backups on our tape drive is dogshit. I think it's because the majority of our systems are RHEL/CentOS and Acronis is geared towards Windows. Possibly. I've had to slipstream extra drivers into a WindowsPE recovery ISO and even then it wouldn't recognise our Quantum tape drive, so I had to manually load drivers into the PE environment over samba and still it refused to recognise that the 2.2TB of data on our tape is a full OS that was backed up using... Acronis Backup!
|
# ? Sep 28, 2018 15:06 |
|
VikingofRock posted:Thanks for the ideas! Maybe the user running the binary doesn't have permissions to read the library file?
|
# ? Sep 28, 2018 17:27 |
|
Jerk McJerkface posted:Maybe the user running the binary doesn't have permissions to read the library file? He's running the binary directly and it's not a matter of exporting LD_LIBRARY_PATH to a child process? What about putting the library path in /etc/ld.so.conf and running ldconfig?
|
# ? Sep 28, 2018 17:54 |
nem posted:He's running the binary directly and it's not a matter of exporting LD_LIBRARY_PATH to a child process? What about putting the library path in /etc/ld.so.conf and running ldconfig? I don't have root access on this machine so I don't think I can edit /etc/ld.so.conf
|
|
# ? Sep 28, 2018 18:38 |
|
VikingofRock posted:I don't have root access on this machine so I don't think I can edit /etc/ld.so.conf No, you wouldn't. Is /pfs/jeltema/build/heasoft/heasoft-6.21/x86_64-unknown-linux-gnu-libc2.12/bin in your PATH and you're not using a script to load that binary indirectly? In such a case LD_LIBRARY_PATH wouldn't get passed to the child process (x86_64-unknown-linux-gnu-libc2.12/bin/nh) unless you export/declare -x the variable.
|
# ? Sep 28, 2018 18:59 |
|
Jerk McJerkface posted:Maybe the user running the binary doesn't have permissions to read the library file? I don't think ldd would work in that case. Maybe bust out strace?
|
# ? Sep 28, 2018 20:27 |
nem posted:No, you wouldn't. Is /pfs/jeltema/build/heasoft/heasoft-6.21/x86_64-unknown-linux-gnu-libc2.12/bin in your PATH and you're not using a script to load that binary indirectly? In such a case LD_LIBRARY_PATH wouldn't get passed to the child process (x86_64-unknown-linux-gnu-libc2.12/bin/nh) unless you export/declare -x the variable. This got me to the right solution! It turns out nh is actually run from a wrapper script: code:
code:
I'm still a little confused why the nh binary needs to look at LD_LIBARY_PATH at all, since it was compiled with "-L/pfs/sw/gcc/4.9.2/lib/gcc/x86_64-unknown-linux-gnu/4.9.2/../../../../lib64 -lquadmath", but at least it's working now.
|
|
# ? Sep 28, 2018 21:12 |
|
|
# ? Apr 24, 2024 23:34 |
|
I run Home Assistant (Hass) on my home server. Hass is reverse proxied by Apache. I have some other web apps running on the server and I don't want to use a million different ports, so I use VirtualHosts. I want to have Node-Red running on the same VirtualHost as Home Assistant, but I can't seem to figure out the right combo of ProxyPass, ProxyPassReverse, and ReWriteEngine directives to get Apache to send any request starting with "nodered" to Node-Red running on port 1880 on the same machine. No matter what I try, Home Assistant ends up fielding any page request to the domain name in question. The forum software freaked out and told me I was a hacker when I pasted the contents of my VirtualHost file into the post, so anyone willing to take a shot at helping figure this out will have to follow this link to see what I'm working with.
|
# ? Oct 2, 2018 09:00 |