Thanks to Matthew Zimmerman for sending me his modified version of MythPyWii, it now has power-saving – after 35 minutes of inactivity the Wiimote turns off. You can download the latest version, as always, here; or you can get this specific version (v17) here.

I love open source!

Before we start, just to be clear – this is not for crackers (people who illegally break in to computer systems) but for hackers (people who modify the software or hardware of their computer system including building, rebuilding, modifying and creating software and electronic hardware either to make it better, faster, give added features or to make it do something it was never intended to do).

UPDATE: we have a google group: http://groups.google.com/group/southackton

I’m intending on setting up a “Hackerspace” in Southampton targeted around software programming (for fun) and hardware hacking (e.g. robotics). I anticipate that people would attend the hackerspace either with their own personal projects (such as controlling their television using a Wii remote), or looking to join in on a cool project. The idea is that we can all share knowledge and help each other out in a fun and innovative environment.

I am provisionally calling this project Southackton, though I expect (hope!) a better name will surface once more people are interested! If you or anyone you know is interested, please contact me via email (my name at benjiegillam.com) or by leaving a comment on this post.

The ultimate goal would be having an “office” somewhere in Southampton where members can drop in any time (within reason) and get to work on their project/just hang out with fellow hackers. If you work from home you could even work form the Hackerspace directly – it wouldn’t be so lonely! I am, however, expecting that we would run it as a “club” until we have enough interested members to make it feasible. Whether this club be weekly, fortnightly, monthly, or bi-monthly is a matter for interested persons to discuss – get in contact and let me know what you think!

As you may know if you read this blog (doubtful!), I am a great fan of open source, and have been a dedicated GNU/Linux user for almost 10 years now, so Linux users would be cool to have as they tend to be interested in tweaking hardware/software, however everyone would be welcome independent of their choice of operating system. Yes, that’s right – even Windows users!

Further, there is no required skill level. I’m happy to have people attend who want to acheive something (e.g. making a webcam recognise who is sat at the PC and change the computers background to a related picture/have their favourite teddy walk towards them) but have no idea how to go about it, so long as they are patient and willing to learn!

I hope you’re interested – let me know!

UPDATE: if you’re interested, sign up to our google group: http://groups.google.com/group/southackton.

Here’s a couple of bash scripts I’ve written recently that might be of use to someone.They work well under Ubuntu, and should work under any GNU/Linux or Unix system with the suitable software installed.

Improved “svn diff” command (ignores whitespace, colour highlights output, requires colordiff) – I put it in /usr/local/bin/svndiff:

#!/bin/bash svn diff -x -w $@ | colordiff

Improved recursive grep command (greps recursively, ignores files in the .svn folders, ignores temporary files, highlights output in colour) – I put it in /usr/local/bin/grepr:

#!/bin/sh SEARCHTERM=$1; shift; find . -type f -a -not -iname "*~" -a -not -iwholename "*/.svn/*" -exec grep -H --color=auto $@ "$SEARCHTERM" "{}" \;

You might also be interested in my previous “whoops I deleted files without telling svn” post here.

When deleting files in a working copy of an SVN repository you should do it on the command line: svn rm [filename]. If, however, you don’t do this (e.g. delete through a gui, or just do “rm” without the “svn“) then SVN gets confused, and puts a “!” in it’s status before all the deleted files. If you svn update, all the files will be recovered, rendering all your time spent deleting them wasted. Really you should use svn rm, but if it’s already too late for that, you can use this bash fragment to delete the files from SVN:

svn status | grep "^\!" | sed 's/^\! *//g' | xargs svn rm

This command does a status command, finds all lines starting with “!“, and then extracts the filename and runs it through “svn rm” – really deleting the file.

Caveats:

1. Manually deleted files are not the only things that makes svn use “!” – so beware of this! Ensure you do really want to delete all those files!
2. This works for filenames which are all_one-word/without/any.spaces but I am not sure if it will work or not for filenames with spaces in.
3. Use at your own risk. The code is simple enough so you should be able to grok what it does.

Enjoy!

Related articles by Zemanta

• Save Your Skin by Customizing Your Bash Prompt (ghacks.net)
• Moving Directories in a SVN Repository using Tortoise SVN (modhul.com)

Image via Wikipedia

Under Ubuntu Intrepid Ibex (should also work for Hardy Herron) with pulseaudio.

(Gfx card: ATI HD 4850)

It is a simple fix that took me ages to work out. Basically, PulseAudio does not recognize the graphics card’s sound card. To fix, we simply force it to recognize it, and remove detection  of alternatives to make it default. There is almost definitely a better way than removing the alternatives,  but *meh* it works for me.

First, work out which device is your correct HDMI device (in this case, 1,3 as it is card 1, output 3):

$ aplay -l
**** List of PLAYBACK Hardware Devices ****
card 0: Intel [HDA Intel], device 0: ALC883 Analog [ALC883 Analog]
Subdevices: 1/1
Subdevice #0: subdevice #0
card 0: Intel [HDA Intel], device 1: ALC883 Digital [ALC883 Digital]
Subdevices: 1/1
Subdevice #0: subdevice #0
card 1: HDMI [HDA ATI HDMI], device 3: ATI HDMI [ATI HDMI]
Subdevices: 0/1
Subdevice #0: subdevice #0

Open up /etc/pulse/default.pa in your favourite text editor

Add this line just before all the #load_module lines (around line 32), replacing 1,3 with whatever you found above):

add-autoload-sink output module-alsa-sink device=plughw:1,3 rate=48000 sink_name=output

Comment out the module-hal-detect section (around lines 40-47):

#### Automatically load driver modules depending on the hardware available
#.ifexists module-hal-detect.so
#load-module module-hal-detect
#.else
#### Alternatively use the static hardware detection module (for systems that
#### lack HAL support)
#load-module module-detect
#.endif

Uncomment this line at the bottom of the file:

set-default-sink output

Save and close the file, then log out and back in again (or kilall pulseaudio && pulseaudio) and, hopefully, your sound should work!

Related articles by Zemanta

• Why you should care about PulseAudio (and how to start doing it)

Image by Blyzz via Flickr

If you choose to mount over NFS just one of your user’s home directories (e.g. /home/jem) under Ubuntu, then you may come accross issues such as failure to log in, the screen freezing (but mouse still moving), loss of configuration data (e.g. icons in your panels), being told that your login session lasted under 10 seconds, and just general instability.

The reason for this is, in Ubuntu’s rush to get you to the Desktop quickly, it loads up GDM (and possibly auto-logs you in) *before* your home directory is mounted over NFS. This is a simple issue of priorities. However, if you log in before the home directory has been mounted, then gconfd-2 and other similar apps will load (or save) settings to your (supposedly empty) /home/jem on your hard drive. When you give up and log out (e.g. Control-Alt-Backspace, or a proper logout), and log back in again, these programs will still be accessing the wrong settings, because they continue to run in the background.

The solution is to abort Ubuntu’s Windows-like behaviour of allowing you to log in before everything has started running at boot time – change GDMs priority from 14 to 80 (or some other number). I chose to do this the lazy way, using “bum“. BUM, the Boot Up Manager, is a simple way to change all things related to booting. It is easy to use, though it does take quite a while (a few minutes!) to start the first time you run it. It must be run in a graphical environment. Simply tick the advanced box, go to the third tab, find gdm, and change its priority up to 80. Save, exit and reboot, and all is well again in the world… though you may have to restore your settings from a backup, or go through the long process of re-configuring your desktop the way you like it.

Good luck!

It has bothered me a little for a while that the MythWeb (part of the fantastic MythTV package for Linux) aspect ratio is hard-coded to 4:3. Most of the TV that I watch (received over Freeview (DVB-T) in the UK) is in 16:9, so watching it back on 4:3 is a bit of a pain. Thus I was motivated to change the hard coding to 16:9. The process is quite simple:

1. Modify line 102 of mythweb/modules/stream/handler.pl – change “3/4” to “9/16“.
2. Modify lines 35 and 37 of mythweb/modules/mythweb/tmpl/default/set_flvplayer.php – change “3/4” to “9/16” and “4:3” to “16:9” respectively.
3. Modify line 505 of mythweb/modules/tv/tmpl/default/detail.php – change “3/4” to “9/16“.
4. Optional: I also added to the end of line 165 of mythweb/modules/stream/handler.pl (which detailed the ${width}x$height) – adding
   .' -aspect '.shell_escape("16:9")
   (make sure you get the fullstop at the beginning!), though I am not sure if this modification is necessary or even beneficial!

There is a minor bug now where the player does not show the control bar at the bottom properly initially, but a click on the preview picture solves this.

I’m currently working on modifications to stream the video in 3gp format to my mobile (a Nokia 6120 Classic), however this seems a lot harder as I have to implement a RTSP server, and have to re-encode all jobs in advance (by using a MythTV User Job) which is not quite what I am after. There is a page about it in the MythTV wiki. I wonder if I can find a cunning way around it…

If this helps you, please let me know in the comments!

I told you about moving my data over to a RAID1+0 and RAID5 system in this previous post, but, as expected, I never got round to finishing it. Until now, that is…

I went to London on business for a few days, and came back to find MythTV had stopped working. The disk was full, so it had given up, and wouldn’t even let me get in to view the recordings – so I couldn’t even delete some to get it working again! (I wonder why Auto-Expire wasn’t working.) Anyway, this spurred me on to finally finish partitioning/setting up my drives. The process was quite simple, really. The following steps generally assume you are root already (sudo su) because I am a “bad” sysadmin and don’t believe in all this constant sudo malarchy. Following my instructions is, as always, at your own risk. I highly recommend that you read the relevant documentation before proceeding (such as this).

1. Reboot into Ubuntu LiveCD, don’t forget to upgrade the software on the LiveCD to prevent any issues!
2. Disable swapspace (swapoff /dev/sd[abcd]1).
3. Use cfdisk to finish formatting the drives (remembering to change partition types to the hexadecimal “fd” – Linux RAID Autodetect). Be careful NOT to modify ANY of the details of the partitions you are already using for data or you will probably lose data!
4. Reboot back into your real system (not LiveCD – minimizes downtime).
5. Optional: add hot spares to current RAID5 devices (mdadm /dev/md1 -a /dev/sdf5).
6. Create the new RAID6 devices (mdadm -C /dev/md3 -l6 -n6 /dev/sd[abcdef]7).
7. Optional: wait for the devices to finish resyncing (watch cat /proc/mdstat).
8. Turn the new RAID devices into LVM physical volumes (pvcreate /dev/md3).
9. Stop any services that depend on /data (/etc/init.d/mythtv-backend stop; /etc/init.d/mpd stop).
10. Unmount the data drive (umount /data).
11. Add the new physical volumes to the current LVM volume group, “raid5” (vgextend raid5 /dev/md3).
12. Expand the logical volume to the full size of the volume group (use pvdisplay to find out the size [411.05GB], and then run lvextend -L+411.05G /dev/raid5/data).
13. Expand the filesystem [ext3] to the full size of the logical volume (e2fsck -f /dev/raid5/data; resize2fs /dev/raid5/data) – running e2fsck on a 600GB drive does take a while… Took about 30 minutes for me with little else running.
14. Remount the data drive (mount /data).
15. Restart the services you stopped eariler (/etc/init.d/mythtv-backend start; /etc/init.d/mpd start).
16. Remember to update (I prefer to re-create) your initramfs and update grub (update-initramfs -k all -c; update-grub)
17. All done!

I chose RAID6 over 6 disks over RAID5 over 5 disk with one hot spare because it has better redundancy and similar performance. It was added to the kernel at the end of 2003 so I think it should be fairly stable by now.

This process was not too complicated, and all of this can be done with very little downtime (if you are clever/daring, you can even re-partition without rebooting, but that was too risky for me!). You can even do the LVM stuff without taking /data offline! I wouldn’t advise it though.

If this post helps you, please let me know in the comments.

Image from Wikipedia

Yesterday I tried to install the latest version of synce in order to get Jem’s Dad’s Windows Mobile 6 phone to share files with Linux (Ubuntu Gutsy Gibbon in this case). After managing to get the software installed, I have been very impressed with it, however actually installing it was a bit of a challenge, though the solution is quite simple and I share it with you now.

1. Uninstall everything synce related before starting.
2. Follow the Synce with Ubuntu instructions.
3. pls should work at this time.
4. Follow the SynceVfs instructions.
   Use ./configure –prefix=/usr
   make; sudo make install
5. Heres the important bit:
   cp /usr/etc/gnome-vfs-2.0/modules/synce-module.conf /etc/gnome-vfs-2.0/modules/
6. killall gnome-vfs-daemon

I think that you can do step 5 alternatively by adding –sysconfdir=/etc to your ./configure command in step 4, however I have not tested this.

Once this is done you should be able to just plug your phone (or other Windows Mobile device) in to the USB, and type synce:/// into Nautilus’ address bar. Simple!