The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

«‹›1036 »

Longinus00: Dec 29, 2005; Ur-Quan

Serephina posted:

So, uh, I read the Op then skipped down to the last page. Questions/Issues regarding Desktop Enviroments that work better on legacy hardware:

I have a ATI Radeon HD 4650 in an old system. This causes no ends of headaches, as apparently ATI stopped updating their propriatary linux driver ['fglrx'] for older cards. So I'm left with the open source alternative, 'radeon', which sucks balls for any shiny gaming and 3d stuff, or using the last known good propriatary driver, 'fglrx-legacy' which is unmaintained. Oh, and fglrx-legacy doesn't work with newer versions of X, which I understand to be the ONLY window manager in existence.

This leaves me finding older distros such as Ubuntu 12.04.1 (but not 12.04.2, that's too new!) to get a sufficiently outdated version of X. 3d gaming SEEMS to work ok, but (naturally...) this breaks desktop speeds to unusably slow.

This is kinda where I throw my hands up and ask for help -slash- consider Windows.

Suggestions for legacy hardware Desktop Environments?

Technically the answer to your question is something like "XFCE" or "LXDE" but it sounds like you're really asking something different.

The 4000 series cards haven't received performance improvements as nice as the 5000+ series have been receiving over the past year or so. Your particular card is also not very fast to begin with, further hampering things. While it's conceivable that the gallium developers could squeeze more performance out of the 4000 series, I think they're going to be mainly focusing on newer product lines from now on.

If you are set on using cataylst (fglrx) then maybe running debian with xfce would be the smarter choice.

# ? Mar 12, 2014 05:30

Adbot: ADBOT LOVES YOU

# ? May 30, 2024 23:41

Vulture Culture: Jul 14, 2003; I was never enjoying it. I only eat it for the nutrients.

Illusive gently caress Man posted:

My coworker wants to change all our servers from CentOS 6.4 to Ubuntu 12.04. I loving hate Ubuntu (probably irrationally). What are the standard arguments for using CentOS so I can convince my boss to veto this poo poo? They don't need to be pure objective Truth, I just want ammunition.

Right now I have this:
Our poo poo works as it is right now.
It's a hell of a lot easier to create RPM packages of our software compared with debs. (or I'm incompetent. idk)
We have working rpm repositories.
It's slightly simpler to manage rpm repos.
We're using Cobbler and cobbler doesn't handle ubuntu or deb repos well. (or I hosed something up when I attempted/failed to provision ubuntu)
I prefer kickstarts over preseed.
I prefer yum over apt.
Our poo poo works as it is right now.

The email I just got says the 2.6 kernel is causing "significant trouble" and I'm like what?? I kinda wish we could just hire someone who really knew what they were doing to handle this stuff but for now it's all me.

This is the Linux thread and I don't usually put on my management consultant hat here, but the problems you're talking about run a lot deeper than your preference of Linux distribution. I like to see teams run well.

It sounds like you already don't have a very good relationship with your coworker. I'm not sure why you need to go to your boss about this instead of communicating these concerns with your coworker directly, but the utter lack of two-way communication being demonstrated in this situation is going to implode your team sooner or later. I cannot be clearer about this: if you make an end-run around your coworker here instead of just communicating like a human being, you are going to cause larger damage to your team dynamic than any irrational hate of a distro ever could. It also communicates to any competent boss that you are incapable of resolving conflicts on your own, whether that's true or not.

Let me reiterate: your infrastructure isn't a pet project, and it isn't a fiefdom to be defensive over. Talk to your coworker and find out why your current infrastructure isn't working for him or her. Your business will be happier for it.

# ? Mar 12, 2014 08:07

babies havin rabies: Feb 24, 2006

I've been using an Ubuntu VPS for a while now for basic stuff. Prosody XMPP, Mumble-server, LAMP development environment, and the occasional Minecraft or Starbound game. This is really the only Internet-facing Linux machine that I've dealt with aside from the gateways at work (which are reviewed by two outside consultants and very locked down - I know what I don't know). How important is it to have restrictive iptables on a VPS like this? Right now I have fail2ban configured on ssh with a permaban which has kept botnet attempts to a minimum. I also monitor the open ports and goings-on daily using netstat and regularly review activity in auth.log (not hard as I am the only person with ssh access). I know that it's probably important and I'm going to put up restrictive rules anyway, but as a more technical question, how is it important? I would think that anything which gained root access to a system would have the capability to just change the iptables on its own, correct?

babies havin rabies fucked around with this message at 15:42 on Mar 12, 2014

# ? Mar 12, 2014 13:30

Vulture Culture: Jul 14, 2003; I was never enjoying it. I only eat it for the nutrients.

babies havin rabies posted:

I've been using an Ubuntu VPS for a while now for basic stuff. Prosody XMPP, Mumble-server, LAMP development environment, and the occasional Minecraft or Starbound game. This is really the only Internet-facing Linux machine that I've dealt with aside from the gateways at work (which are reviewed by two outside consultants and very locked down - I know what I don't know). How important is it to have restrictive iptables on a VPS like this? Right now I have fail2ban configured on ssh with a permaban which has kept botnet attempts to a minimum. I also monitor the open ports and goings-on daily using netstat and regularly review activity in auth.log (not hard as I am the only person with ssh access). I know that it's probably important and I'm going to put up restrictive rules anyway, but as a more technical question, how is it important? I would think that anything which gained root access to a system would have the capability to just change the iptables on its own, correct?

On an unimportant and typically-configured system, root access isn't nearly as important as an attacker having some kind of unprivileged access to execute arbitrary code on the server. If you run vulnerable software, you're very likely to end up sending hundreds of thousands of spam emails or unwittingly participating in DDoS attacks, and somewhat less likely to be proxying somebody's child porn downloads. Root just gives them the ability to cover their tracks.

If you keep your software up-to-date and avoid dumb misconfigurations like public-facing DNS recursors, that's like 99.5% of the way there, though.

# ? Mar 12, 2014 15:47

Experto Crede: Aug 19, 2008; Keep on Truckin'

I'm trying to use SCP to copy a file to a remote server. When I do this I get the following output:

code:

stdin: is not a tty
 ________________________

and the file doesn't copy.

Apparently this is because I'm using cowsay with fortune, and in your .bashrc you need to make it not be used when you're using an interactive shell with [[ $- == *i* ]] || return

But I'm using fish at the moment, and adding that line in ~/.config/fish/config.fish causes a syntax error. How do I specify this in fish?

EDIT: Never mind, it was serverside configuration issue!

Experto Crede fucked around with this message at 17:56 on Mar 13, 2014

# ? Mar 13, 2014 17:39

reading: Jul 27, 2013

evol262 posted:

In .bashrc

shopt -s histappend
PROMPT_COMMAND="$PROMPT_COMMAND;history -a; history -n"

Of course, you'll have to press enter once to refresh the history if you're trying to get to a command from another session, but...

What exactly will this PROMPT_COMMAND line do? I can't understand a self-referencing variable (PROMPT_COMMAND is not already in my .bashrc somewhere else) and the man page for history shows nothing relevant for -a or -n.

# ? Mar 14, 2014 04:37

Polygynous: Dec 13, 2006; welp

reading posted:

What exactly will this PROMPT_COMMAND line do? I can't understand a self-referencing variable (PROMPT_COMMAND is not already in my .bashrc somewhere else) and the man page for history shows nothing relevant for -a or -n.

PROMPT_COMMAND: "If set, the value is executed as a command prior to issuing each primary prompt."

If it looks confusing it's just setting PROMPT_COMMAND to itself and adding the "history" bits. (If it's not set before, it'll just be ";history -a; history -n" afterwards.)

Also 'history' in this case is a bash builtin, that and PROMPT_COMMAND are in 'man bash' (or 'help history' since 'man bash' is lol huge).

All together: the 'shopt' tells bash to append rather than overwrite the history file (on shell exit), and (the interesting part) the PROMPT_COMMAND adds lines from the current shell then loads any new lines from the history file. So even if you've got multiple shells open from different hosts, if you run 'ls' in one, after your next command or just hitting enter 'ls' will be in the history on the others.

(Not sure how useful that is, but it turns out it can be done. :v:

)

Polygynous fucked around with this message at 04:55 on Mar 14, 2014

# ? Mar 14, 2014 04:53

Elias_Maluco: Aug 23, 2007; I need to sleep

So I boot into my Mint 15 (KDE edition) and for some reason KDE messed my monitor settings (again) and my second monitor is disabled.

Now I cant open the settings to enable it because KDE is apparently opening the settings window on the disabled monitor.

What can I do?

Why is KDE still so stupid about multiple monitors

Elias_Maluco fucked around with this message at 14:01 on Mar 14, 2014

# ? Mar 14, 2014 13:33

Riso: Oct 11, 2008; by merry exmarx

Because KDE still relies on X.

Wayland, where art thou?

# ? Mar 14, 2014 13:49

Lysidas: Jul 26, 2002; John Diefenbaker is a madman who thinks he's John Diefenbaker.; Pillbug

Elias_Maluco posted:

So I boot into my Mint 15 (KDE edition) and for some reason KDE messed my monitor settings (again) and my second monitor is disabled.

Now I cant open the settings to enable it because KDE is apparently opening the settings window on the disabled monitor.

What can I do?

Why is KDE still so stupid about multiple monitors

Have you tried Alt-F3 for the window operations menu, using arrow keys to select the 'Move' action, and then arrow keys (or moving the mouse) to reposition the window?

# ? Mar 14, 2014 14:56

Hollow Talk: Feb 2, 2014

Elias_Maluco posted:

So I boot into my Mint 15 (KDE edition) and for some reason KDE messed my monitor settings (again) and my second monitor is disabled.

Now I cant open the settings to enable it because KDE is apparently opening the settings window on the disabled monitor.

What can I do?

Why is KDE still so stupid about multiple monitors

Alternatively, fire up a console window, and type:

code:

xrandr --query

That will give you your monitor names, the first thing of "<NAME> connected <RESOLUTION> ...". Then do:

code:

xrandr --output <FIRST MONITOR NAME> --auto --output <SECOND MONITOR NAME> --auto

code:

xrandr --output <FIRST MONITOR NAME> --auto --output <SECOND MONITOR NAME> --off

The former sets both to their automatic best mode, the latter disables the second monitor entirely, after which you should be able to set things up as you are used to.

# ? Mar 14, 2014 15:18

Sauer: Sep 13, 2005; Socialize Everything!

Is there any way to try Wayland now? Wouldn't mind tossing it in a VM if there's a testing version out.

# ? Mar 14, 2014 19:13

Elias_Maluco: Aug 23, 2007; I need to sleep

I solved it by killing the system settings process after it was open on the wrong monitor and then, when I opened it again, it opened on the active monitor.

# ? Mar 14, 2014 19:18

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

DSauer posted:

Is there any way to try Wayland now? Wouldn't mind tossing it in a VM if there's a testing version out.

Depends on what you want to try. There are two major Wayland implementations: Weston, developed by Wayland upstream, and mutter-wayland, developed by GNOME.

I'm the tech lead on mutter-wayland and Wayland integration in GNOME, so I'd be happy to answer any questions you have about it.

We're trying to get an updated mutter-wayland preview available for trying out in 3.12 again, and you should be able to try that in any distro that ships GNOME 3.12.

You can try Weston by using Rebecca Black Linux, or by building it from scratch.

# ? Mar 14, 2014 20:08

Sauer: Sep 13, 2005; Socialize Everything!

Suspicious Dish posted:

...using Rebecca Black Linux...

Looks like Arch has an up to date build of Weston in the official repositories so I'll give that a shot. Thanks mate.

# ? Mar 14, 2014 21:59

madkapitolist: Feb 5, 2006

edit: NVM, easily google-able issue.

madkapitolist fucked around with this message at 08:43 on Mar 15, 2014

# ? Mar 15, 2014 08:11

spankmeister: Jun 15, 2008

Suspicious Dish posted:

Depends on what you want to try. There are two major Wayland implementations: Weston, developed by Wayland upstream, and mutter-wayland, developed by GNOME.

I'm the tech lead on mutter-wayland and Wayland integration in GNOME, so I'd be happy to answer any questions you have about it.

We're trying to get an updated mutter-wayland preview available for trying out in 3.12 again, and you should be able to try that in any distro that ships GNOME 3.12.

You can try Weston by using Rebecca Black Linux, or by building it from scratch.

Oh boy, more fragmentation in open source land. :allears:

# ? Mar 15, 2014 09:01

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

spankmeister posted:

Oh boy, more fragmentation in open source land.

Hm? mutter-wayland implements the same exact Wayland protocol, and it should work with the exact same Wayland applications.

# ? Mar 15, 2014 15:18

stray: Jun 28, 2005; "It's a jet pack, Michael. What could possibly go wrong?"

I have a home server I'm putting together and whose configuration I'm managing with (masterless) Puppet and Git. When I push an updated Puppet configuration to the server, I want to check out the configuration files to /etc/puppet (probably through a post-receive hook). However, since that folder is owned by root, what's the best way to do this? The server will have a dedicated Git user.

# ? Mar 15, 2014 17:34

Combat Pretzel: Jun 23, 2004; No, seriously... what kurds?!

So yeah, I figured, I set up a Linux VM to run The Thing System for all my fancy gadgets, and for ease downloaded Ubuntu. I guess that was a big mistake, because there doesn't seem an easy way to start a shell. None of the default shortcuts offers a shell, nor does the desktop context menu, nor that universal search thing. Really?!

--edit: Nevermind. "Terminal".

Combat Pretzel fucked around with this message at 22:20 on Mar 15, 2014

# ? Mar 15, 2014 22:15

RFC2324: Jun 7, 2012; http 418

Combat Pretzel posted:

So yeah, I figured, I set up a Linux VM to run The Thing System for all my fancy gadgets, and for ease downloaded Ubuntu. I guess that was a big mistake, because there doesn't seem an easy way to start a shell. None of the default shortcuts offers a shell, nor does the desktop context menu, nor that universal search thing. Really?!

do a search in the launcher menu for 'terminal'

# ? Mar 15, 2014 22:17

FISHMANPET: Mar 3, 2007; Sweet 'N Sour
Can't
Melt
Steel Beams

RFC2324 posted:

do a search in the launcher menu for 'terminal'

Also, the default terminal program in Ubuntu is gnome-terminal.

# ? Mar 15, 2014 22:29

spankmeister: Jun 15, 2008

Suspicious Dish posted:

Hm? mutter-wayland implements the same exact Wayland protocol, and it should work with the exact same Wayland applications.

Yeah I suppose so but is two projects doing the same thing really a good thing?

# ? Mar 15, 2014 23:08

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

Well, we already had mutter / gnome-shell, our WM / compositor / desktop UI that's been in development for over 12 years. We're not just going to throw it away and start from scratch.

So, there are two options: somehow hack Weston in on the side so that it can support mutter, or add a separate Wayland implementation to our existing codebase. Really, only the latter one makes sense.

# ? Mar 15, 2014 23:14

evol262: Nov 30, 2010; #!/usr/bin/perl

stray posted:

I have a home server I'm putting together and whose configuration I'm managing with (masterless) Puppet and Git. When I push an updated Puppet configuration to the server, I want to check out the configuration files to /etc/puppet (probably through a post-receive hook). However, since that folder is owned by root, what's the best way to do this? The server will have a dedicated Git user.

Why is it owned by root? You should have a puppet user. If not, chown :git /etc/puppet. If you do, put them in the same group and chown it to that.

# ? Mar 15, 2014 23:56

reading: Jul 27, 2013

spoon0042 posted:

PROMPT_COMMAND: "If set, the value is executed as a command prior to issuing each primary prompt."

If it looks confusing it's just setting PROMPT_COMMAND to itself and adding the "history" bits. (If it's not set before, it'll just be ";history -a; history -n" afterwards.)

Also 'history' in this case is a bash builtin, that and PROMPT_COMMAND are in 'man bash' (or 'help history' since 'man bash' is lol huge).

All together: the 'shopt' tells bash to append rather than overwrite the history file (on shell exit), and (the interesting part) the PROMPT_COMMAND adds lines from the current shell then loads any new lines from the history file. So even if you've got multiple shells open from different hosts, if you run 'ls' in one, after your next command or just hitting enter 'ls' will be in the history on the others.

(Not sure how useful that is, but it turns out it can be done. )

This gave me a cascading recursive error,

code:

<after I do anything in the terminal>
bash: PROMPT_COMMAND: line 12: syntax error near unexpected token `;'
bash: PROMPT_COMMAND: line 12: `;history -a; history -n;history -a; history -n;history -a; history -n;history -a; history -n;history -a; history -n'

I removed the recursive part and just used PROMPT_COMMAND="history -a; history -n;" and added the final semicolon. Is this a good idea? I understand better what PROMPT_COMMAND is but I don't know if this will still work as I want it to.

It's super useful though, I type a lot of huge complex commands as part of coursework and it's great to have those stored in my history, even when I'm working from many terminal windows. That's why I want this feature.

# ? Mar 16, 2014 00:26

Polygynous: Dec 13, 2006; welp

reading posted:

This gave me a cascading recursive error,
code:
<after I do anything in the terminal>
bash: PROMPT_COMMAND: line 12: syntax error near unexpected token `;'
bash: PROMPT_COMMAND: line 12: `;history -a; history -n;history -a; history -n;history -a; history -n;history -a; history -n;history -a; history -n'
I removed the recursive part and just used PROMPT_COMMAND="history -a; history -n;" and added the final semicolon. Is this a good idea? I understand better what PROMPT_COMMAND is but I don't know if this will still work as I want it to.

It's super useful though, I type a lot of huge complex commands as part of coursework and it's great to have those stored in my history, even when I'm working from many terminal windows. That's why I want this feature.

I thought that looked funny, but afaik it *should* only execute once when bash starts so I don't know how what's happening... is happening. But yeah, PROMPT_COMMAND="history -a; history -n;" should be fine. (If you're not using PROMPT_COMMAND already. Otherwise PROMPT_COMMAND="history -a; history -n; $PROMPT_COMMAND" should work, just for reference.)

If you're that interested in keeping your history around you'll probably want to set HISTSIZE and HISTFILESIZE as both default to 500.

# ? Mar 16, 2014 02:38

reading: Jul 27, 2013

How can I draw boxes in the terminal window? I want to get a nice GUI going for a program and I was inspired by how nice Kismet looks, but after looking at their source code it just looks like I'll have to use lots of ncurses and maybe some c++. Is there no terminal program for easily drawing boxes and shapes?

Dungeon Crawl does some amazing graphics in the terminal too. I assume that's also ncurses magic.

# ? Mar 16, 2014 21:30

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

Well, there's some Box drawing characters you can print into the terminal. That's really it.

# ? Mar 16, 2014 21:32

reading: Jul 27, 2013

Suspicious Dish posted:

Well, there's some Box drawing characters you can print into the terminal. That's really it.

I thought these were part of curses. I'm aware of these I was wondering if there's a pre-existing shell program to draw boxes with these or if I need to write that myself.

# ? Mar 16, 2014 21:39

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

curses is just a framework for building terminal UIs. That's it.

I don't really know why you think you need a special program to draw boxes. They'll work with a regular echo. What are you trying to do?

# ? Mar 16, 2014 21:40

reading: Jul 27, 2013

Suspicious Dish posted:

curses is just a framework for building terminal UIs. That's it.

I don't really know why you think you need a special program to draw boxes. They'll work with a regular echo. What are you trying to do?

I want to make a card-playing program that will work in the terminal to let two people place and arrange cards. The cards would be boxes with text inside them. I suppose I could use echo along with tput to place the cursor at a position in the window, draw the box, and repeat elsewhere.

I thought curses was just a group of characters, a subset of ASCII. Is it a c++ library?

# ? Mar 18, 2014 03:24

Suspicious Dish: Sep 24, 2011; 2020 is the year of linux on the desktop, bro; Fun Shoe

curses is a C library. The one most people use is known as "ncurses": http://www.gnu.org/software/ncurses/

# ? Mar 18, 2014 03:48

SurgicalOntologist: Jun 17, 2004

Have you already written the engine? Whatever language you're already using probably has a package with bindings to curses. For example if you're using Python, Urwid is a high-level terminal interface library that allows you to easily make boxes and whatnot.

# ? Mar 18, 2014 04:16

reading: Jul 27, 2013

SurgicalOntologist posted:

Have you already written the engine? Whatever language you're already using probably has a package with bindings to curses. For example if you're using Python, Urwid is a high-level terminal interface library that allows you to easily make boxes and whatnot.

Awesome! Thank you, this is exactly what I am looking for.

# ? Mar 18, 2014 04:32

evol262: Nov 30, 2010; #!/usr/bin/perl

reading posted:

Awesome! Thank you, this is exactly what I am looking for.

I use urwid daily and can offer lots of advice if you go this way

# ? Mar 18, 2014 05:21

jre: Sep 2, 2011; To the cloud ?

I've hit this bug in gluster 3.3.2 when trying to do a replace-brick command. The node completely locked up and affected some clients so I paused the replace. I had a glusterfsd process still stuck at 100% cpu usage despite being paused and I tried to abort the replace-brick and that failed.

Although the cluster appears to still be working correctly it is still continuously logging this

code:


[2014-03-18 16:20:04.788128] E [name.c:141:client_fill_address_family] 
0-myvol-replace-brick: transport.address-family not specified. 
Could not guess default value from (remote-host:(null) or transport.unix.connect-path:(null)) options

in the log for the brick I attempted to replace. As per the bug attempting to issue an abort command just causes the management interface to hang for 30 minutes.

Anyone know how to get it to stop attempting the replace-brick operation?

replace-brick status shows as "paused".

# ? Mar 18, 2014 17:28

evol262: Nov 30, 2010; #!/usr/bin/perl

jre posted:

I've hit this bug in gluster 3.3.2 when trying to do a replace-brick command. The node completely locked up and affected some clients so I paused the replace. I had a glusterfsd process still stuck at 100% cpu usage despite being paused and I tried to abort the replace-brick and that failed.

Although the cluster appears to still be working correctly it is still continuously logging this
code:
[2014-03-18 16:20:04.788128] E [name.c:141:client_fill_address_family] 
0-myvol-replace-brick: transport.address-family not specified. 
Could not guess default value from (remote-host:(null) or transport.unix.connect-path:(null)) options
in the log for the brick I attempted to replace. As per the bug attempting to issue an abort command just causes the management interface to hang for 30 minutes.

Anyone know how to get it to stop attempting the replace-brick operation?

replace-brick status shows as "paused".

You should probably ask in #gluster. It's very likely in the options in /var/lib/glusterd/, buy ask.

# ? Mar 18, 2014 17:46

jre: Sep 2, 2011; To the cloud ?

evol262 posted:

You should probably ask in #gluster. It's very likely in the options in /var/lib/glusterd/, buy ask.

I have already with no reply :smith:

I suspect it might be a case of removing the rb_dst_brick.vol file but I'm loathe to do that on a production system without knowing if its the right thing to do.

# ? Mar 18, 2014 17:57

Adbot: ADBOT LOVES YOU

# ? May 30, 2024 23:41

Bob Morales: Aug 18, 2006; ~~Just wear the fucking mask, Bob~~

I don't care how many people I probably infected with COVID-19 while refusing to wear a mask, my comfort is far more important than the health and safety of everyone around me!

I have a couple nested directories full of PDF files that need to be TIFF files to be sent somewhere.

SomeClient -
2011
commercial
residential
2012
commercial
residential
...

I know I can just use the 'convert' command on a single file:

code:

convert document001.pdf document001.tiff

What's the best way to loop through these, and rename them all?

A quick bash script with a for loop?

I started doing something like

code:

find * -name "*.pdf" -exec convert $1 xxxxxxxx {} \;

but I wasn't sure how to get just the filename without the extension, to append .tiff to

# ? Mar 20, 2014 13:33

The Something Awful Forums > Discussion > Serious Hardware/Software Crap > The Linux Questions Thread: a bunch of pitfalls, but technically it's possible

«‹›1036 »