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

«‹›1024 »

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

fletcher posted:

What about things like a build server that devs would need to access over https? Should I just give it an elastic IP & security group or is that something you should have behind a VPN and have buildserver.mydomain.com point to the 10.x.x.x private IP?

It depends. One of the really annoying things about security groups in EC2 is that anything other than an instance is treated as a second-class citizen. In a normal firewall like a PIX/ASA or Palo Alto or whatever, you can create named aliases like "here's the IP for the office NAT, here's the IP for Jeremy's workstation" and reference them in your configs in ways that are maintainable and make sense. In EC2, you end up with a clusterfuck of random IP addresses poo poo all over your configurations if you try to manage them through the GUI. That leaves you using stuff like CloudFormation or the EC2 command line in a bunch of heavily-commented scripts if you want to have any idea what's happening a month after you create the rule. (The same problem exists on S3 bucket policies; gently caress JSON for not supporting comments in any way, shape or form.)

For this reason alone, sometimes VPNs are easier in the long run, but it depends a lot on your tooling and the amount of investment you want to make into doing things "the AWS way." If you're building a whole scalable web operations infrastructure in the cloud, none of this stuff is a big deal, but it's a huge pain in the rear end if you just want to host a few servers and keep your sanity in the process.

# ? Oct 11, 2014 21:33

Adbot: ADBOT LOVES YOU

# ? Apr 19, 2024 22:08

fuf: Sep 12, 2004; haha

Does anyone know of a nifty little script that'll scan a directory of mp3s and print out a track list with length and quality etc?

# ? Oct 13, 2014 18:25

fletcher: Jun 27, 2003; ken park is my favorite movie; Cybernetic Crumb

fuf posted:

Does anyone know of a nifty little script that'll scan a directory of mp3s and print out a track list with length and quality etc?

You could probably whip something up in python pretty quickly using one of the ID3 packages

# ? Oct 13, 2014 18:47

spankmeister: Jun 15, 2008

yeah or mediainfo and some bash

# ? Oct 13, 2014 18:50

Quixzlizx: Jan 7, 2007

OK, I tried getting smbnetfs to work on my latest model Acer C720 chromebook, using this guide, but I'm getting stuck on step 11. I have 14.04 LTS and xfce4 running on my chroot currently. As you can see, I already tried running that chmod 600 fix... I got the same exact error the first time.

code:

(trusty)quixzlizx@localhost:~$ sudo chmod 600 /home/quixzlizx/.smb/smbnetfs.conf 
(trusty)quixzlizx@localhost:~$ sudo smbnetfs ~/Downloads/Network 
	 WARNING: Unknown libsmbclient version: 4.1.6-Ubuntu
         SMBNetFS may not work as expected.

4192->reconfigure_read_config_file: Error: Insecure config file permission.
Can't apply 'auth' directive.
Run 'chmod 600 /home/quixzlizx/.smb/smbnetfs.conf' to fix it.
4192->reconfigure_read_config_file: Error: Insecure config file permission.
Can't apply 'host' directive.
Run 'chmod 600 /home/quixzlizx/.smb/smbnetfs.conf' to fix it.
fuse: bad mount point `/home/quixzlizx/Downloads/Network': Permission denied

# ? Oct 14, 2014 00:58

YouTuber: Jul 31, 2004; by FactsAreUseless

Apparently Debian TC is trying to divert back off the course they took on systemd adoption and is trying a last ditch effort to make the distro to be init neutral?

https://lists.debian.org/debian-vote/2014/10/msg00001.html
https://lists.debian.org/debian-vote/2014/10/msg00012.html

Why the gently caress are they trying to make more work for themselves in a community that is already strapped for people?

# ? Oct 17, 2014 00:16

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

That's just Ian Jackson being Ian Jackson. It is part of the General Resolution process, because politics. Ignore it.

# ? Oct 17, 2014 01:15

nem: Jan 4, 2003; panel.dev
apnscp: cPanel evolved

Having made the conversion from SysV[init] to systemd, that's an asinine argument to keep some obsoleted shell scripts still functional. systemd is fantastic: automatic restarts, built-in watchdog, parallelized start-up, auditing, coherent interface, and if you so choose, there is backwards-compatibility with service/chkconfig.

# ? Oct 17, 2014 04:54

kujeger: Feb 19, 2004; OH YES HA HA

nem posted:

Having made the conversion from SysV[init] to systemd, that's an asinine argument to keep some obsoleted shell scripts still functional. systemd is fantastic: automatic restarts, built-in watchdog, parallelized start-up, auditing, coherent interface, and if you so choose, there is backwards-compatibility with service/chkconfig.

It's not an argument to keep old shell scripts working, it's an argument to keep new stuff working without needing systemd.

# ? Oct 17, 2014 11:29

SYSV Fanfic: Sep 9, 2003; by Pragmatica

nem posted:

Having made the conversion from SysV[init] to systemd, that's an asinine argument to keep some obsoleted shell scripts still functional. systemd is fantastic: automatic restarts, built-in watchdog, parallelized start-up, auditing, coherent interface, and if you so choose, there is backwards-compatibility with service/chkconfig.

But... But... Unix...

# ? Oct 17, 2014 12:15

RFC2324: Jun 7, 2012; http 418

keyvin posted:

But... But... Unix...

Must not deviate from standards developed before I was born!!!!

# ? Oct 17, 2014 12:17

fuf: Sep 12, 2004; haha

Why can't I get Apache's server status page to work? What am I missing?

code:

$ sudo a2enmod status
Module status already enabled

/etc/apache2/mods-available/status.conf:

code:

<Location /server-status>
    SetHandler server-status
    Order deny,allow
    Deny from all
    Allow from 127.0.0.1 ::1
    Allow from localhost
</Location>

code:

$ sudo service apache2 restart
[...] done
$ lynx [url]http://localhost/server-status[/url]
(lynx shows 403 forbidden)
$ tail /var/log/apache2/error.log
[Fri Oct 17 12:51:27 2014] [error] [client 127.0.0.1] client denied by server configuration: /home/fuf/default/www/server-status

# ? Oct 17, 2014 13:54

spankmeister: Jun 15, 2008

Are you using apache 2.4? In that case you need the Require directive

# ? Oct 17, 2014 15:02

fuf: Sep 12, 2004; haha

It's apache 2.2.22. I tried adding the require directive but still no joy.

# ? Oct 17, 2014 15:31

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

fuf posted:

Why can't I get Apache's server status page to work? What am I missing?

code:

$ sudo a2enmod status
Module status already enabled

/etc/apache2/mods-available/status.conf:

code:

<Location /server-status>
    SetHandler server-status
    Order deny,allow
    Deny from all
    Allow from 127.0.0.1 ::1
    Allow from localhost
</Location>

code:

$ sudo service apache2 restart
[...] done
$ lynx [url]http://localhost/server-status[/url]
(lynx shows 403 forbidden)
$ tail /var/log/apache2/error.log
[Fri Oct 17 12:51:27 2014] [error] [client 127.0.0.1] client denied by server configuration: /home/fuf/default/www/server-status

Apache is confusing. Read this

kujeger posted:

It's not an argument to keep old shell scripts working, it's an argument to keep new stuff working without needing systemd.

This is beyond what Debian can do, and they probably don't have the engineering manpower even if they wanted try to maintain large patchsets for a million packages. Even the BSDs are looking at implementing compat layers for some stuff (logind in particular). Debian could try to use some of those, but it's time to move forward.

Sysv is dead. "Modern" UNIXes (Solaris, OSX) don't use it either. Ubuntu doesn't use it. Move on.

# ? Oct 17, 2014 16:08

telcoM: Mar 21, 2009; Fallen Rib

fuf posted:

Why can't I get Apache's server status page to work? What am I missing?
code:
$ tail /var/log/apache2/error.log
[Fri Oct 17 12:51:27 2014] [error] [client 127.0.0.1] client denied by server configuration: /home/fuf/default/www/server-status

What is your DocumentRoot setting, I wonder?

If it isn't /home/fuf/default/www, you've done something more complicated to map the root of your web space there.

Are you using any RewriteRules?

Alternatively, the Location block in the module configuration will apply to the "main server" configuration only. If your VirtualHost settings catch your request, it might not apply and you might have to place a similar Location block within the applicable VirtualHost.

# ? Oct 17, 2014 16:30

kujeger: Feb 19, 2004; OH YES HA HA

evol262 posted:

This is beyond what Debian can do, and they probably don't have the engineering manpower even if they wanted try to maintain large patchsets for a million packages. Even the BSDs are looking at implementing compat layers for some stuff (logind in particular). Debian could try to use some of those, but it's time to move forward.

Sysv is dead. "Modern" UNIXes (Solaris, OSX) don't use it either. Ubuntu doesn't use it. Move on.

I agree with you (and so it appears the majority of the technical commitee does), but that is the argument of the pro loose-coupling camp in Debian.

# ? Oct 17, 2014 16:35

fuf: Sep 12, 2004; haha

evol262 posted:

Apache is confusing. Read this

On the basis of this I changed status.conf to:

code:

<Location /server-status>
    SetHandler server-status
    Order allow,deny
    Allow from 127.0.0.1 ::1
    Allow from localhost
</Location>

But still no luck.

telcoM posted:

Alternatively, the Location block in the module configuration will apply to the "main server" configuration only. If your VirtualHost settings catch your request, it might not apply and you might have to place a similar Location block within the applicable VirtualHost.

Aha, this must have been it. I added the same Location block to /sites-available/default and now it works. But aren't requests for localhost always caught by the "default" VirtualHost? Where is the "main server" configuration? :shobon:

Thanks both.

# ? Oct 17, 2014 16:54

telcoM: Mar 21, 2009; Fallen Rib

fuf posted:

Aha, this must have been it. I added the same Location block to /sites-available/default and now it works. But aren't requests for localhost always caught by the "default" VirtualHost? Where is the "main server" configuration?

Main server configuration is everything that has not been enclosed within any <VirtualHost> block.

As to your first question... well, evol262 said it best:

evol262 posted:

Apache is confusing.

You could run "apache2ctl -S" to get a synopsis of your VirtualHost configuration, which might help.

This part of Apache documentation might also help... or at least give you an idea of the complexity involved.

Enabling the "info" module (which behaves similar to the "status" module) and accessing http(s)://<server-name>/server-info?config would allow you to see the entire configuration in one go, with references to each file used in building up the whole.

Here is another important part of Apache documentation. Note that the <Location> blocks are processed last in the overall order, so if you have something else that mutates the request URI path, it might make the Location block no longer match, even though you would expect it to.

# ? Oct 17, 2014 21:28

covener: Jan 10, 2004; You know, for kids!

fuf posted:

On the basis of this I changed status.conf to:

My moneys is on the status.conf you're looking at never actually being included in your configuration. The actual context change should not have helped. The apache layout for debian encourages this sort of thing.

quote:

But aren't requests for localhost always caught by the "default" VirtualHost?

No, the local interface and port of the incoming connection either matches an explicit virtualhost or is handled by the "base server configuration" or "main server [configuration]". But most things defined in the main server configuration are merged into the virtualhosts anyway.

covener fucked around with this message at 04:06 on Oct 18, 2014

# ? Oct 18, 2014 03:48

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

I've got a super-weird problem with my Chef server and a NAT gateway that I'm trying to wrap my brain around.

Communications between client and server occur over HTTPS on port 443. Server is on a completely separate network; traffic must traverse the Internet to reach it. My Google Compute Engine systems that are directly connected to the Internet function just fine. Systems on Google Compute Engine that are connected over a NAT gateway send a bunch of successful REST API calls, then send one specific REST API call, receive an RST from the server, drop the connection, and blow up with an EOFError in Chef/OpenSSL/Ruby.

The NAT server on GCE is itself managed through Chef and works fine. It's also using the exact same iptables NAT configuration on AWS, where it works fine. The server's logs indicate that it's serving a 200 response as normal.

Any ideas what might be going on?

Vulture Culture fucked around with this message at 05:51 on Oct 18, 2014

# ? Oct 18, 2014 05:38

jaegerx: Sep 10, 2012; Maybe this post will get me on your ignore list!

Misogynist posted:

I've got a super-weird problem with my Chef server and a NAT gateway that I'm trying to wrap my brain around.

Communications between client and server occur over HTTPS on port 443. Server is on a completely separate network; traffic must traverse the Internet to reach it. My Google Compute Engine systems that are directly connected to the Internet function just fine. Systems on Google Compute Engine that are connected over a NAT gateway send a bunch of successful REST API calls, then send one specific REST API call, receive an RST from the server, drop the connection, and blow up with an EOFError in Chef/OpenSSL/Ruby.

The NAT server on GCE is itself managed through Chef and works fine. It's also using the exact same iptables NAT configuration on AWS, where it works fine. The server's logs indicate that it's serving a 200 response.

Any ideas what might be going on?

Tcpdump the traffic.

# ? Oct 18, 2014 05:51

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

jaegerx posted:

Tcpdump the traffic.

I did, which is where I see the RST coming from. I'm not able to correlate anything else meaningful from the dumps, though, even after decrypting the SSL payloads. The question is about why the server might be sending an RST to clients behind this particular NAT gateway.

# ? Oct 18, 2014 05:53

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

e: bad data, removed, see above

Vulture Culture fucked around with this message at 06:53 on Oct 18, 2014

# ? Oct 18, 2014 06:05

other people: Jun 27, 2004; Associate Christ

Misogynist posted:

I did, which is where I see the RST coming from. I'm not able to correlate anything else meaningful from the dumps, though, even after decrypting the SSL payloads. The question is about why the server might be sending an RST to clients behind this particular NAT gateway.

Did you capture on the server? Perhaps it isn't actually the one sending the reset.

# ? Oct 18, 2014 13:09

Blowjob Overtime: Apr 6, 2008; Steeeeriiiiiiiiike twooooooo!

DrKennethNoisewater posted:

My mother-in-law needed a computer for internet browsing and storing pictures of grandchildren, so I assembled one for her and put Mint 17 on it, because gently caress paying for an OS for that. My overall command of Linux is probably about a 2 out of 10, but I have run a few different flavors on my laptop over the last few years, and will be changing it to Mint 17 as well to help build familiarity with the OS for when she has questions. My goal is that being her tech support will give me the kick in the rear end I need to start to really put more effort into learning Linux.

The one hiccup with her computer requirements is that she is also an avid sewer, and the machine that she has accepts a proprietary USB storage card that uses its own software to interface with the PC. The software is Brother PED-Basic (version 1.07, if you're curious). WINE handles the software no problem because it is basically babbys first file-transfer program; however, the instructions for installing the USB dongle is to plug it in on your Windows system and wait for it to recognize the device, then point it towards the driver on the CD. The problem is that WINE doesn't care to acknowledge the newly plugged in USB device and installing the proper driver.

My skill set may be limited, but it doesn't seem like it should be too big of a stretch to have WINE recognize that a USB device is connected and needs a driver associated with it (it is a glorified thumbdrive). Am I assuming that this is much more simple than really will be, or is it achievable with a little bit of goon assistance?

Quoting myself for background on the situation. Crazysim recommended a VM, so I installed VirtualBox and put an XP VM on the Linux box. The USB driver and software installed, and the program opens, but when I try to select the item on the computer that I want to transfer to the card the program becomes stops responding.

Here is a picture of the program itself:

Inside of the program I can browse the PC, change the filepath, move and resize the window, but as soon as I click one of the little images on the left half in order to move it to the card the program freezes up. The software is installed on my Windows PC (Windows 7) and runs fine (as pictured above). The software advertises compatibility with Windows 98 through 7, so I figured XP was the safer bet to work well. Would the best use of my time be to put on a Windows 7 VM and try that, or does this raise any of your various spider senses as a VM/Linux issue?

# ? Oct 18, 2014 16:45

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

Kaluza-Klein posted:

Did you capture on the server? Perhaps it isn't actually the one sending the reset.

The server's sending the RST -- three of them for the same TCP connection (?!) -- but I've had some more time to analyze dumps of good runs from the NAT server as well, and those are also receiving RSTs at certain points without exploding. But I got tired of not being able to see meaningful JSON payloads in my packet dumps, so I disabled gzip compression on the server (along with sendfile, for good measure). Now I get this more meaningful error out of the clients behind the NAT:

code:

Unexpected Error:
-----------------
Chef::Exceptions::ContentLengthMismatch: Response body length 737072 does not match HTTP Content-Length header 754030.

Wireshark doesn't seem to want to decrypt partial streams, so I can't see where it's choking, but one interesting thing I just noticed is that the continuations on the TLSv1.2 stream are arbitrarily large when seen from the NAT host before translation, and fragmented near MTU size (1410/1474 size vs. 1460 MTU) when seen from the host behind the NAT.

EDIT!
The fragmentation seems to be the key thing I needed to find more on this issue: https://code.google.com/p/google-co...20Summary%20Log

Disabled tx offload on eth0, everything is fine. Welcome to our software-defined networking hell.

Vulture Culture fucked around with this message at 17:30 on Oct 18, 2014

# ? Oct 18, 2014 17:22

Ashex: Jun 25, 2007; These pipes are cleeeean!!!

This issue has been plaguing me for months and I always assumed it was the terminal I was using but today I replicated it with putty.

On Archlinux occasionally when I'm installing a package with pacman and I enter Y to confirm installation and hit enter, it inserts ^M instead of a newline (or whatever it is).

Why is this happening? It's been driving me nuts as I can't reproduce it reliably but when it does happen it's only in pacman (doesn't matter if it's run through yaourt or direct).

I'm running zsh as my default user shell if that helps any.

Edit: also happens when rm prompts me.

Ashex fucked around with this message at 18:50 on Oct 18, 2014

# ? Oct 18, 2014 18:42

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

Ashex posted:

This issue has been plaguing me for months and I always assumed it was the terminal I was using but today I replicated it with putty.

On Archlinux occasionally when I'm installing a package with pacman and I enter Y to confirm installation and hit enter, it inserts ^M instead of a newline (or whatever it is).

Why is this happening? It's been driving me nuts as I can't reproduce it reliably but when it does happen it's only in pacman (doesn't matter if it's run through yaourt or direct).

I'm running zsh as my default user shell if that helps any.

Edit: also happens when rm prompts me.

What's the output of locale and the contents of $TERM? Can you also post the relevant encoding settings from your PuTTY profile?

# ? Oct 18, 2014 19:56

telcoM: Mar 21, 2009; Fallen Rib

DrKennethNoisewater posted:

Quoting myself for background on the situation. Crazysim recommended a VM, so I installed VirtualBox and put an XP VM on the Linux box. The USB driver and software installed, and the program opens, but when I try to select the item on the computer that I want to transfer to the card the program becomes stops responding.

[...] or does this raise any of your various spider senses as a VM/Linux issue?

Most Windows USB drivers happily install even if they don't yet see the device they are supposed to control. Does the program stop responding the same way if you start it in a real Windows system without the card plugged in?

Did you tell VirtualBox to pass through USB traffic between the VM and the USB storage card device?

If you didn't:
Start the VM without the USB storage card plugged in, and right-click the USB plug icon at the bottom of the VM window. Note the list of USB devices, but don't select anything.
Now plug in the USB storage card, wait a few seconds and right-click the icon again. You should see one or more new USB devices in that list. You can now select the storage card device(s) to allow the VM direct USB access to it/them.

(Although the storage card is physically just one device, it is possible it might appear as several devices electrically on the USB bus.)

Once you know the device(s) you need, open the VirtualBox Manager window, select the VM and click on Settings.
Select "USB" and click on the icon with the USB plug and a green plus sign. Here you can tell Virtual box to always allow the VM to access selected USB devices directly: select the device(s) associated with the storage card.

# ? Oct 19, 2014 14:01

telcoM: Mar 21, 2009; Fallen Rib

Ashex posted:

This issue has been plaguing me for months and I always assumed it was the terminal I was using but today I replicated it with putty.

On Archlinux occasionally when I'm installing a package with pacman and I enter Y to confirm installation and hit enter, it inserts ^M instead of a newline (or whatever it is).

Why is this happening? It's been driving me nuts as I can't reproduce it reliably but when it does happen it's only in pacman (doesn't matter if it's run through yaourt or direct).

I'm running zsh as my default user shell if that helps any.

Edit: also happens when rm prompts me.

The problems with Enter key processing cannot usually be caused by locale settings. A wrong $TERM might do it, but if there are no other visible issues, that's unlikely. Most likely it is a bug in pacman's handling of stty settings.

In order to do character-based menus efficiently and/or to accept keyboard input without waiting for the user to hit Enter, some programs will switch the (pseudo)TTY they're using for the session to "raw" mode. The raw mode disables various functions in the tty driver: one of them is the "icrnl" conversion, which translates the ^M produced by the Enter key to the real Unix-style newline character, which is ASCII #10, also known as LF or ^J.

Obviously, if pacman expects you to enter Y and hit Enter, it should re-enable the icrnl conversion before waiting for user input, but apparently that does not happen. It might be because $TERM is selecting a wrong terminal type with wrong terminal control codes, or it might be because pacman doesn't do an appropriate tcsetattr() function call after displaying that prompt for some reason.

As a workaround, you can try pressing Control-J instead of hitting the Enter key.

If you want to know more about this, be warned that this rabbit hole is ancient and goes pretty deep. Many things won't have any practical purpose on modern systems, other than the all-important "compatibility with earlier systems".

Run "stty -a" on any terminal session. You'll see a lot of cryptic abbreviations, some prefixed with a minus sign (meaning "disabled"), others not (meaning "enabled"). There are also some settings with either a numeric or character value.
This is very ancient Unix stuff, from the time when terminals were literally teletypewriters: basically electronic typewriters with a serial port.

"man stty" will have descriptions for each abbreviation. "man 3 termios" will describe the corresponding programming functions and their flags, which will have a pretty much one-to-one correspondence with the stty settings.

# ? Oct 19, 2014 14:35

Ashex: Jun 25, 2007; These pipes are cleeeean!!!

That's actually really interesting, I'm a bit skeptical it's a pacman bug as I reproduced the issue with rm (when it prompts to delete if I don't specify -f). I'll try using ctrl-j next time I encounter it and see if that works.

As for the output of $TERM, this is the output from when I'm using putty:

xterm

As for locale

quote:

LANG=en_US.UTF-8
LC_CTYPE=en_US.UTF-8
LC_NUMERIC="en_US.UTF-8"
LC_TIME="en_US.UTF-8"
LC_COLLATE="en_US.UTF-8"
LC_MONETARY="en_US.UTF-8"
LC_MESSAGES="en_US.UTF-8"
LC_PAPER="en_US.UTF-8"
LC_NAME="en_US.UTF-8"
LC_ADDRESS="en_US.UTF-8"
LC_TELEPHONE="en_US.UTF-8"
LC_MEASUREMENT="en_US.UTF-8"
LC_IDENTIFICATION="en_US.UTF-8"
LC_ALL=

I use iTerm2 on OSX most of the time and $TERM returns xterm-256color

# ? Oct 20, 2014 12:45

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

Ashex posted:

That's actually really interesting, I'm a bit skeptical it's a pacman bug as I reproduced the issue with rm (when it prompts to delete if I don't specify -f). I'll try using ctrl-j next time I encounter it and see if that works.

As for the output of $TERM, this is the output from when I'm using putty:

xterm

As for locale

I use iTerm2 on OSX most of the time and $TERM returns xterm-256color

It's a problem with terminals in general and unclean exits (or bad code) in applications that use it raw. It's probably not pacman. It probably is tmux or screen or some curses application that you CTRL+C'd out of or kill -9'd because it was unresponsive. Try to figure out which one.

# ? Oct 20, 2014 14:30

reading: Jul 27, 2013

I'm trying to connect to wifi networks using passwords stored in a text file, using bash. But I can't get this line to work:

code:

iwconfig wlan0 essid $NETWORK_ID key s:$WIRELESS_KEY

Even if I put in an incorrect password for WIRELESS_KEY, my Xubuntu install will say that it has connected. But it obviously can't send or receive data since it didn't work. And it stops there. I have to manually disconnect and reconnect and edit the network's information to change the password to the correct one.

tl;dr how can I read passwords from a text file and use that to connect to wifi networks, while being robust against wrong passwords?

# ? Oct 20, 2014 16:43

Hollow Talk: Feb 2, 2014

reading posted:

I'm trying to connect to wifi networks using passwords stored in a text file, using bash. But I can't get this line to work:
code:
iwconfig wlan0 essid $NETWORK_ID key s:$WIRELESS_KEY
Even if I put in an incorrect password for WIRELESS_KEY, my Xubuntu install will say that it has connected. But it obviously can't send or receive data since it didn't work. And it stops there. I have to manually disconnect and reconnect and edit the network's information to change the password to the correct one.

tl;dr how can I read passwords from a text file and use that to connect to wifi networks, while being robust against wrong passwords?

I assume you are not using WEP as wireless encryption, are you? iwconfig cannot connect to WPA networks on its own, which would be why things aren't working. If that is the case, have a look at this link, which will also solve your "read password from file" problem: http://askubuntu.com/a/138476

If you do use WEP, then you will most like have to append the key's index to your iwconfig command. Also, don't use WEP.

# ? Oct 20, 2014 19:43

SYSV Fanfic: Sep 9, 2003; by Pragmatica

Wine question. How do you figure out if crashes are caused by wine regressions or from trying to use an HD 7870?

# ? Oct 20, 2014 21:59

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

keyvin posted:

Wine question. How do you figure out if crashes are caused by wine regressions or from trying to use an HD 7870?

Sending coredumps to the wine devs, probably.

# ? Oct 20, 2014 22:06

Blowjob Overtime: Apr 6, 2008; Steeeeriiiiiiiiike twooooooo!

telcoM posted:

Most Windows USB drivers happily install even if they don't yet see the device they are supposed to control. Does the program stop responding the same way if you start it in a real Windows system without the card plugged in?

Did you tell VirtualBox to pass through USB traffic between the VM and the USB storage card device?

If you didn't:
Start the VM without the USB storage card plugged in, and right-click the USB plug icon at the bottom of the VM window. Note the list of USB devices, but don't select anything.
Now plug in the USB storage card, wait a few seconds and right-click the icon again. You should see one or more new USB devices in that list. You can now select the storage card device(s) to allow the VM direct USB access to it/them.

(Although the storage card is physically just one device, it is possible it might appear as several devices electrically on the USB bus.)

Once you know the device(s) you need, open the VirtualBox Manager window, select the VM and click on Settings.
Select "USB" and click on the icon with the USB plug and a green plus sign. Here you can tell Virtual box to always allow the VM to access selected USB devices directly: select the device(s) associated with the storage card.

This is some damned fine troubleshooting, and I like it.

Freezing is, in fact, what happens when there isn't anything connected and one tries to operate the program, even on a Windows machine. I have previously gone through and connected the USB drive through VirtualBox directly to the VM, but it seemed worth a try to give it another go. It seemed to be going smoothly, and the program was running and not freezing while I highlighted different designs that are located on the hard drive. Upon trying to upload one of those designs to the card, however, a message came up that the card was not in the adapter. It was physically in the slot, but not pushed in all the way. As soon as I pushed it in all the way the computer didn't make any acknowledgement that the card was attached, but the program seized up again upon clicking around in the bank of designs on the hard drive. That result is completely repeatable.

As mentioned above, my own PC is running Windows 7 with this same software installed, and that rig has successfully transferred designs from the PC to the card. In order to troubleshoot a little before laying myself at the mercy of this thread again I reinstalled the driver, but it didn't help.

It seems like there is some stupid little thing that I'm missing due to Linux inexperience that this shouldn't be such a pain, but here I am grovelling back to this thread. My first instinct is to just start a new VM with Windows 7 since I have it working on a different PC with that OS, but the Engineer in me knows that potential solution doesn't really solve the problem, just avoid it.

# ? Oct 21, 2014 01:04

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

DrKennethNoisewater posted:

This is some damned fine troubleshooting, and I like it.

Freezing is, in fact, what happens when there isn't anything connected and one tries to operate the program, even on a Windows machine. I have previously gone through and connected the USB drive through VirtualBox directly to the VM, but it seemed worth a try to give it another go. It seemed to be going smoothly, and the program was running and not freezing while I highlighted different designs that are located on the hard drive. Upon trying to upload one of those designs to the card, however, a message came up that the card was not in the adapter. It was physically in the slot, but not pushed in all the way. As soon as I pushed it in all the way the computer didn't make any acknowledgement that the card was attached, but the program seized up again upon clicking around in the bank of designs on the hard drive. That result is completely repeatable.

As mentioned above, my own PC is running Windows 7 with this same software installed, and that rig has successfully transferred designs from the PC to the card. In order to troubleshoot a little before laying myself at the mercy of this thread again I reinstalled the driver, but it didn't help.

It seems like there is some stupid little thing that I'm missing due to Linux inexperience that this shouldn't be such a pain, but here I am grovelling back to this thread. My first instinct is to just start a new VM with Windows 7 since I have it working on a different PC with that OS, but the Engineer in me knows that potential solution doesn't really solve the problem, just avoid it.

dmesg after you insert it. It's possible that the VM doesn't have exclusive access and Linux is seeing the card (probably FAT) and mounting it.

# ? Oct 21, 2014 01:52

Adbot: ADBOT LOVES YOU

# ? Apr 19, 2024 22:08

Polygynous: Dec 13, 2006; welp

keyvin posted:

Wine question. How do you figure out if crashes are caused by wine regressions or from trying to use an HD 7870?

If it worked in a previous version of wine and doesn't work currently independent of graphics driver / whatever then it's a regression. Driver bugs tend to take down the X server though, sometimes anyway. If something is just crashing (bunch of hex and function names and stuff on the terminal) you can file a bug report with that backtrace or post on the forums if you're in doubt.

# ? Oct 21, 2014 02:47

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

«‹›1024 »