adempewolff's questions

While fighting with my discrete graphics card, I reinstalled Ubuntu several times. The first couple times I was very careful to move the .priv and .der key I was using to sign my dkms kernel modules to another partition. Unfortunately, as I got more frustrated I became less careful and apparently accidentally cut and pasted from the backup (rather than copying and pasting) and then reformatted over the sole remaining key in the next re-install. I used photorec to try and recover the files but unfortunately no *.der or *.priv files were recovered.

It seems that the intended way to un-enroll machine owner keys (MOKs) is

mokutil --delete keyname.der

Given that I no longer have the public key, is there another way for me to un-enroll it?

I just got a new computer and installed version 3.2 of Astrill's VPN client for debian based linux OSes.

It runs fine and created an icon on the desktop, but:

a) there is no panel icon when it starts (and ergo I can't lock it to the panel using the right click menu)

b) I cannot find it when searching using the unity menu

c) each time it starts it acts as if it's freshly installed (asked me to install browser plugin, and enter username and password regardless of whether I'd clicked remember password before)

Disclaimer: I'm documenting this problem and solution so I don't have to re-discover it each time I reinstall Ubuntu. If anyone has a better solution though I'd be happy to see a better answer.

I am using a Windows 7 (64 bit) guest with an Ubuntu 14.04 (64 bit) host. I just recently upgraded to Virtualbox 4.3.16 r95972, although this problem appeared before upgrade--at the very least in the previous version. The Virtualbox extension pack is the latest edition and guest additions is installed on the guest OS.

My problem is that when I put the guest OS into full-screen mode, the Ubuntu system bar remains on top, regardless of where the mouse focus is. There is also a unity-style shadow behind where the Virtualbox mini menu is hiding in the bottom center.

I can't remember exactly when this started happening unfortunately.

I also noticed that on the unity sidebar context menu for Virtualbox, there is now a 3rd item called Qt-subapplication, which I suspect is the Virtualbox mini-menu, and unity treating it as a separate window is why it has the shadow, and potentially why the guest OS window can't capture the fullscreen.

I'm guessing this might be a bug of some sort, but, if it is, I suspect it's a bug in the form of some cool new feature of unity that Canonical is unwilling to change, such that it falls to the Virtualbox developers to comply with the new feature. As such, I don't see it being fixed anytime soon, so my question would be about possible workarounds.

So, anybody have any ideas about how to keep unity from capturing the mini-menu as a separate window and preventing the guest OS from capturing fullscreen?

Thanks!

Banshee's Amazon MP3 Import extension until recently allowed seamless importing of songs purchased from Amazon MP3. It did this by a)opening .amz files and using them to connect to and download the purchased files from Amazon's servers, and b) using hooks in Banshee's built-in browser to automatically recognize and open the .amz files when clicked on in the browser.

However, recently this functionality stopped working. Banshee will display Contacting Server in the lower left hand corner for a little while and then stop.

Furthermore opening the Amazon Cloud Player in the Banshee browser or any other browser on a Linux system to manually download the .amz file now results in the message:

On Linux systems, Cloud Player only supports downloading songs one at a time. To download your music, deselect all checkboxes, select the checkbox for the song you want to download, then click the "Download" button.

How can I get around this and import my purchased music into Banshee as I used to?

I was reviewing the updates I was about to download today and I

1. noticed that a lot of them had to do with gesture support,

2. noticed that many of these were new installs rather than upgrades.

Has 12.04 just added multi-touch gesture support mid-release? If so, what are the capabilities that this adds? Which applications already support these capabilities and can I expect others to add support in the near future?

Here are the packages that were installed:

Install: libframe6:amd64 (2.2.4-0ubuntu0.12.04.1), libgeis1:amd64 (2.2.9.2-0ubuntu1), libgrail5:amd64 (3.0.6-0ubuntu0.12.04.01, automatic)

And here are those that were upgraded (also including many with touch support):

Upgrade: libgrip0:amd64 (0.3.4-0ubuntu2~ubuntu12.04.1, 0.3.5-0ubuntu1~12.04.1), eog:amd64 (3.4.2-0ubuntu1, 3.4.2-0ubuntu1.1), ginn:amd64 (0.2.4-0ubuntu1, 0.2.4.1-0ubuntu1)

Of which the descriptions for the new installs are,

libgeis1:

Gesture engine interface support
 A common API for clients of a systemwide gesture recognition and propagation
 engine.

libframe6:

Touch Frame Library
 This library handles the buildup and synchronization of a set of simultaneous
 touches. The library is input agnostic, with bindings for mtdev, frame and
 XI2.1.

libgrail5:

Gesture Recognition And Instantiation Library
 This library consists of an interface and tools for handling gesture
 recognition and gesture instantiation. Applications can use the grail callbacks
 to receive gesture primitives and raw input events from the underlying kernel
 device.

And the descriptions for the upgraded packages are,

ligrip0:

provides multitouch gestures to GTK+ apps

 Libgrip hooks gesture recognition into GTK+ applications.

ginn:

Gesture Injector: No-GEIS, No-Toolkits

 A daemon with jinn-like wish-granting capabilities: it gives applications the
 ability to support a subset of multi-touch gestures without having to integrate
 GEIS or multi-touch GTK/Qt libs.

Adding in a ton of new libraries and upgrading the existing components makes me wonder if 12.04 is meant to start natively supporting gestures other than two finger scroll in the near future. I expected these capabilities to be introduced soon but I thought that they would only be rolled out in a new release, not as upgrades for an existing release.

Anyone have any info about this?

I have a server running Ubuntu 10.04 (it's on the other side of the world and I haven't built up the wherewithal to upgrade it remotely yet) and I have been told that there is a CD in one of it's two CD drives. I want to rip an image of the cd and then download it to my local computer (I don't need help with either of these steps).

However, I cannot seem to confirm whether or not there actually is a CD in the drive as I was told. It did not automatically mount anywhere (which I'm thinking might just be a result of it being a headless server not running X, nautilus, or any of the other nice user friendly things).

There are two CD drives connected via SCSI:

austin@austinvpn:/proc/scsi$ cat /proc/scsi/scsi
Attached devices:
Host: scsi0 Channel: 00 Id: 00 Lun: 00
  Vendor: ATA      Model: WDC WD400EB-75CP Rev: 06.0
  Type:   Direct-Access                    ANSI  SCSI revision: 05
Host: scsi1 Channel: 00 Id: 00 Lun: 00
  Vendor: Lite-On  Model: LTN486S 48x Max  Rev: YDS6
  Type:   CD-ROM                           ANSI  SCSI revision: 05
Host: scsi1 Channel: 00 Id: 01 Lun: 00
  Vendor: SAMSUNG  Model: CD-R/RW SW-248F  Rev: R602
  Type:   CD-ROM                           ANSI  SCSI revision: 05

However when I try mounting either of these devices (and every other device that could possibly be the cd-drive), it says no medium found:

austin@austinvpn:/proc/scsi$ sudo mount -t iso9660 /dev/scd1 /cdrom
mount: no medium found on /dev/sr1
austin@austinvpn:/proc/scsi$ sudo mount -t iso9660 /dev/scd0 /cdrom
mount: no medium found on /dev/sr0
austin@austinvpn:/proc/scsi$ sudo mount -t iso9660 /dev/cdrom /cdrom
mount: no medium found on /dev/sr1
austin@austinvpn:/proc/scsi$ sudo mount -t iso9660 /dev/cdrom1 /cdrom
mount: no medium found on /dev/sr0
austin@austinvpn:/proc/scsi$ sudo mount -t iso9660 /dev/cdrw /cdrom
mount: no medium found on /dev/sr1

Here are the contents of my /dev folder:

austin@austinvpn:/proc/scsi$ ls /dev
agpgart          loop6               ram6      tty10  tty38  tty8
austinvpn        loop7               ram7      tty11  tty39  tty9
block            lp0                 ram8      tty12  tty4   ttyS0
bsg              mapper              ram9      tty13  tty40  ttyS1
btrfs-control    mcelog              random    tty14  tty41  ttyS2
bus              mem                 rfkill    tty15  tty42  ttyS3
cdrom            net                 root      tty16  tty43  urandom
cdrom1           network_latency     rtc       tty17  tty44  usbmon0
cdrw             network_throughput  rtc0      tty18  tty45  usbmon1
char             null                scd0      tty19  tty46  usbmon2
console          oldmem              scd1      tty2   tty47  usbmon3
core             parport0            sda       tty20  tty48  usbmon4
cpu_dma_latency  pktcdvd             sda1      tty21  tty49  vcs
disk             port                sda2      tty22  tty5   vcs1
dri              ppp                 sda5      tty23  tty50  vcs2
ecryptfs         psaux               sg0       tty24  tty51  vcs3
fb0              ptmx                sg1       tty25  tty52  vcs4
fd               pts                 sg2       tty26  tty53  vcs5
full             ram0                shm       tty27  tty54  vcs6
fuse             ram1                snapshot  tty28  tty55  vcs7
hpet             ram10               snd       tty29  tty56  vcsa
input            ram11               sndstat   tty3   tty57  vcsa1
kmsg             ram12               sr0       tty30  tty58  vcsa2
log              ram13               sr1       tty31  tty59  vcsa3
loop0            ram14               stderr    tty32  tty6   vcsa4
loop1            ram15               stdin     tty33  tty60  vcsa5
loop2            ram2                stdout    tty34  tty61  vcsa6
loop3            ram3                tty       tty35  tty62  vcsa7
loop4            ram4                tty0      tty36  tty63  vga_arbiter
loop5            ram5                tty1      tty37  tty7   zero

And here is my fstab file:

austin@austinvpn:/proc/scsi$ cat /etc/fstab
# /etc/fstab: static file system information.
#
# Use 'blkid -o value -s UUID' to print the universally unique identifier
# for a device; this may be used with UUID= as a more robust way to name
# devices that works even if disks are added and removed. See fstab(5).
#
# <file system> <mount point>   <type>  <options>       <dump>  <pass>
proc            /proc           proc    nodev,noexec,nosuid 0       0
/dev/mapper/austinvpn-root /               ext4    errors=remount-ro 0       1
# /boot was on /dev/sda1 during installation
UUID=ed5520ae-c690-4ce6-881e-3598f299be06 /boot           ext2    defaults        0       2
/dev/mapper/austinvpn-swap_1 none            swap    sw              0       0

Am I missing something/doing something wrong, or is there just no CD in the drive or is the drive possibly broken? Is there any nice command to list devices with mountable media?

Thanks in advance for any help!

I want to replace the folders in my home/ directory with symlinks to folders in another location, but I want to retain the folder emblems without having to go searching for them. How do I do this?

To illustrate what I'm asking, the directories in the home folder have nice emblems that look like this:

If I just delete this folder and replace it with a symlink the music emblem is lost and replaced by a link emblem.

I want it to have have both the music emblem and the symlink emblem, like this:

Preferably I'd like to be able to do this without having to manually hunt down and set the icon for the folder (but a good answer explaining how to do so would be upvoted as a good alternative).

I just installed StackApplet StackApplet 1.5 using the debian package from the launchpad site. When I first clicked on the preferences button in the menu it tried opening http://localhost:8150/ in Chromium, and I got this error:

This webpage is not available The webpage at http://localhost:8150/ might be temporarily down or it may have moved permanently to a new web address.
Error 120 (net::ERR_SOCKS_CONNECTION_FAILED): Unknown error.

My Chromium is launched with a switch to make it use a socks5 proxy, using this command:

chromium-browser  --proxy-server="socks5://localhost:1080" %U

I remembered that I wasn't using a bypass proxy URL list so I added the switch: --proxy-bypass-list="localhost,127.0.0.1" to the above command and I was then able to access the preference page without any problems.

However, after adding my account information I see the message: Network Error. in the StackApplet menu. This makes me think that StackApplet might be trying to use the socks5 proxy--even though it is not set globally, only Chromium uses this proxy. I don't require StacksApplet to use my proxy, in fact it would probably be better if it didn't. So my questions are:

1. What does the Network Error. message mean?
   
   and
2. What should I do to fix/debug this error?

PPA's are normally disabled when you upgrade, and must be re-enabled manually. I upgraded to 12.04 about a month ago and I just noticed that while my other PPAs had all been disabled, the Google PPAs were not disabled. Why is this?

I believe that my Chinese ISP has started detecting and bandwidth-limiting SSH connections. I'm not using the default SSH port so I suspect that they are using deep packet filtering to detect the unencrypted handshake.

I am aware of one solution to this which patches openSSH to encrypt the handshake. However, I have two problems:

1. This project is three years old so if I were to just compile it as is, I would be forsaking three years of security updates for openSSH.
2. My server is in the US and my only way of accessing it is through SSH--so downtime or messing up isn't an option.

As such I have two questions:

1. Can I somehow patch a current version of openSSH to use the obfuscated handshake? If so, how would I do this? Which files would I need to add/modify in the source?
2. How could I avoid any downtime for the ssh server while doing this patch? Can I install two ssh servers side by side?

If anyone could help me with this--or knew of a better way to obfuscate ssh handshakes--my appreciation would be boundless!

I'm looking at buying a new laptop and considering buying one with a hybrid HDD/SSD drive. Specifically I am looking at the new second generation Seagate Momentus XT. For some reviews look here and here.

The caching is all done on the firmware level so there is no question that this basic functionality will work with Ubuntu.

However, the second generation drives are supposed to now reserve space on the SSD for OS boot files.

The Seagate documentation describes it this way:

Boot
Momentus XT drives are known for incredibly fast boot-up times, and now they are even better with FAST Factor boot technology. This system boot-up technique can reduce your boot time to mere seconds—for a fast cold-boot start. It can cut your system start-up time by up to 65% over a traditional HDD. FAST Factor boot technology will deliver the fastest boot possible, whether you boot your system once a day or once a week. Your OS boot-up files are always there, ready to work. (source)

What concerns me is there is no additional information about this in the product manual and the reviews seem to suggest that the firmware might be looking for Windows boot files specifically. See for example:

Dubbed FAST Factor Boot, this new approach aims to accelerate the Windows boot process whether you're firing up the OS for the first time or rebooting after months of uninterrupted activity. With the old Momentus XT, Adaptive Memory needed a few boots to learn which OS-related data to cache. Using the drive for long enough without rebooting had the potential to push that data out of the cache, reversing any previous boot-time benefits. With the new XT, a segment of the NAND has been reserved solely for OS data related to the boot process. That section of the flash is populated as Windows is installed to the drive, which should speed things up starting from the first boot. If Windows is being installed via a drive image, the XT will need a couple of boots to determine what to put in the roped-off section of its cache. As with clean installs, that data won't be kicked out of the cache between boots. (source)

The Seagate documentation mentions nothing about being specific to Windows OS boot files, so I'm wondering if the review is just being ignorant of there being OSes other than Window or if they are actually privy to some information that is not in the official documentation.

Anyone have any ideas about this?

Seagate's Response

(I'm adding this into the question as I don't consider it a satisfactory answer--just more background for the question)

I sent Seagate a message through the presale's email support form on their website asking about this. I talked with two support agents, one via email and one via live chat.

The first agent was professional but not particularly helpful, responding with this email:

Hello Austin,

Thanks for contacting Seagate global support.

We do not test our drives for use with Linux. Sorry we cannot answer your question in depth.

Regards,

Agent (taking names out) Seagate Global Customer Support

I responded saying that I didn't need an answer based on performance proven through testing but rather just the expected behavior. Another agent responded to my email but there was a mix-up and he actually sent me a link for a screen-sharing, which out of curiosity I boot up a Windows VM and followed. The agent realized there was a mix up but was still willing to discuss my question in the screen-sharing softwares chat function.

This second agent was more helpful, but I got the feeling that He wasn't very knowledgeable about the product. He said that Linux is not a supported OS but that it should be compatible with Linux. When I asked for information specifically about the reserved space on the SSD for OS files, I just got another description of the basic features of the Momentus XT line, "It caches frequently accessed files regardless of OS or software." He had no information about the new "FAST Factor Boot" feature--and honestly, I wasn't even sure if he knew about the new feature at all.

I asked him if he could email me a summary of our conversation to post on a website and he sent this:

Dear Austin,

Thank you for contacting Seagate.

Officially Linux is not supported, because it is an open source Operating system.

By selectively tackling data that is frequently used and time-consuming to fetch, the Momentus XT drive will copy this data into the flash and maintain the relevancy. You get the instant response experience you've been searching for.

Momentus XT hybrid drives are designed to work in any standard laptop. These drives are OS-, driver- and software-independent, making them remarkably simple to integrate and easy to use.

If you have any additional questions you may call us during your regional business hours listed below. For your convenience we also have on-line chat assistance.

Live Assistance: Chat: Americas: http://support2.seagate.com/ChatLaunch?rc=1 Europe: http://support2.seagate.com/ChatLaunch?rc=2

Regards, (name removed) Seagate Global Customer Support

So in summary:

1. Confirmation of what we already knew (basic functionality is OS independent)
2. No information at all about the new reserved space for boot files
3. A very weird statement (Officially Linux is not supported, because it is an open source Operating system.) about open source software that I assume/hope is not really Seagate's reason for not supporting Linux.

I think to get a real answer we would need to talk with either a) someone who owns this drive and could test the boot performance, or b) somehow get through to a high-level support agent who is actually familiar with the product's features. I'm considering contacting System 76, who have this drive as an option for their Ubuntu laptops and asking them if they've tested the boot performance (or if their OEM status can get more useful answers out of Seagate). If people have twitter accounts (I don't) they could also start tweeting this question at Seagate (@askseagate) and see if it catches their attention.

I've got a (virtualized) local LAMP server on Xubuntu 12.04 that I am using for Drupal development. I'd like to learn how to use web 2.0 integration Drupal modules (eg. youtube, facebook, etc.) but I live in China so these sites are obviously blocked.

I use SSH tunneling and a local SOCKS proxy for browsing these sites, but I am not sure how to make my virtualized Apache2 server use this proxy. I can see 3 potential solutions but I'm not really sure how to best go about implementing them.

(in order of preference, I prefer limiting the traffic sent through the ssh tunnel to only what needs to get around the firewall)

1. set up ssh tunnel from guest OS get Apache2 server to route external traffic (or even better, only external traffic with specific host names, eg. facebook.com, youtube.com) through the local SOCKS proxy.

2. set up ssh tunnel from guest OS, route all traffic through local SOCKS proxy (easy with Ubuntu but looks like configuring a global SOCKS proxy might be a touch harder with Xubuntu because it doesn't have Ubuntu's Network Proxy GUI)

3. make virtualbox on host OS route all traffic through SOCKS proxy.

Any ideas on which route would be best and how to implement it? Thanks in advance for any help!