The old way is specifying Devices and Screens and using Xinerama and using directives like RightOf and LeftOf in your xorg.conf.  The new way moves all of this stuff out of xorg.conf and shifts the responsibility to xrandr which will configure screens and devices at runtime and can do so completely dynamically.

So the basic steps are just to get xorg up and running, which should work by default, tweak your xorg.conf slightly, and then run xrandr at the beginning of your X session to configure your displays to your liking.

The tweak to xorg.conf, at least for the i915, is to add the keyword Virtual to the Display Subsection of the Screen section.  Don’t worry, I’ll include my entire xorg.conf so you can see what I’m struggling to say in a clear way.  When you do this you must also remove the Modes lines.  This sets up a display window that the X server will use to place your output.  Don’t worry, it does it in a smart way.  You can have different resolutions on each display and it will be displayed correctly.  You need to specify a virtual size that is large enough to hold all of your displays.  In my case one display is 1280×1024 and the other is 1400×1050.  The virtual display is 2680×1050 (with a side-by-side setup).  Unfortunately the i915 driver disables DRI if either dimension exceeds 2048.  I hope this is a hardware limitation and not a software limitation.

The final step is to run xrandr to configure your displays.  I use the following command to setup an external monitor on the DVI output running to the left of my laptop display. I run this out of the Startup Programs under System/Preferences/Sessions.

/usr/bin/xrandr --output TMDS-1 --left-of LVDS

Here is my xorg.conf (scrollbar at the bottom):

Section "Files"
	FontPath	"/usr/share/fonts/X11/misc"
	FontPath	"/usr/X11R6/lib/X11/fonts/misc"
	FontPath	"/usr/share/fonts/X11/cyrillic"
	FontPath	"/usr/X11R6/lib/X11/fonts/cyrillic"
	FontPath	"/usr/share/fonts/X11/100dpi/:unscaled"
	FontPath	"/usr/X11R6/lib/X11/fonts/100dpi/:unscaled"
	FontPath	"/usr/share/fonts/X11/75dpi/:unscaled"
	FontPath	"/usr/X11R6/lib/X11/fonts/75dpi/:unscaled"
	FontPath	"/usr/share/fonts/X11/Type1"
	FontPath	"/usr/X11R6/lib/X11/fonts/Type1"
	FontPath	"/usr/share/fonts/X11/100dpi"
	FontPath	"/usr/X11R6/lib/X11/fonts/100dpi"
	FontPath	"/usr/share/fonts/X11/75dpi"
	FontPath	"/usr/X11R6/lib/X11/fonts/75dpi"
	# path to defoma fonts
	FontPath	"/var/lib/defoma/x-ttcidfont-conf.d/dirs/TrueType"
EndSection

Section "Module"
	Load	"i2c"
	Load	"bitmap"
	Load	"ddc"
	Load	"dri"
	Load	"extmod"
	Load	"freetype"
	Load	"glx"
	Load	"int10"
	Load	"vbe"
EndSection

Section "InputDevice"
	Identifier	"Generic Keyboard"
	Driver		"kbd"
	Option		"CoreKeyboard"
	Option		"XkbRules"	"xorg"
	Option		"XkbModel"	"pc104"
	Option		"XkbLayout"	"us"
# Make capslock become another control key
	Option  "XkbOptions"    "ctrl:nocaps"
EndSection

Section "InputDevice"
	Identifier	"Configured Mouse"
	Driver		"mouse"
	Option		"CorePointer"
	Option		"Device"		"/dev/input/mice"
	Option		"Protocol"		"ImPS/2"
	Option		"Emulate3Buttons"	"true"
EndSection

Section "InputDevice"
	Identifier	"Synaptics Touchpad"
	Driver		"synaptics"
	Option		"SendCoreEvents"	"true"
	Option		"Device"		"/dev/psaux"
	Option		"Protocol"		"auto-dev"
	Option		"HorizScrollDelta"	"0"
EndSection

Section "Device"
	Identifier	"intel0"
	Driver		"i810"
	BusID		"PCI:0:2:0"
EndSection

Section "Monitor"
	Identifier	"Internal Monitor"
	Option		"DPMS"
EndSection

Section "Screen"
	Identifier	"Internal Screen"
	Device		"intel0"
	Monitor		"Internal Monitor"
	DefaultDepth	24
	SubSection "Display"
		Depth		24
# This is required to get the proper size for one display of 1280x1024 and another of 1400x1050
		Virtual		2680 1050
	EndSubSection
EndSection

Section "ServerLayout"

	Identifier	"Default Layout"
	Screen 	"Internal Screen"
	InputDevice	"Generic Keyboard"
	InputDevice	"Configured Mouse"
	InputDevice	"Synaptics Touchpad"
EndSection

References:

http://bgoglin.livejournal.com/9846.html

I also found that holding shift while clicking on Reply or Reply to All also opens the composition window in HTML editing mode.

References:

1

Long ago I found this set of instructions for installing “RALUS” on Debian. Since their install scripts were so terrible I was happy to find good alternative. I’m reproducing the instructions here for my own reference. These instructions are based off of installing version 11d.7170.

The first step is to convert the RPM files to debs. The RPMs are repackaged as .tar.gz files in the directory cdimg/pkgs/Linux.

$ tar xzvf VRTSralus.tar.gz
$ tar xzvf VRTSvxmsa.tar.gz

You will now have two RPM files. They must be converted to .debs using alien.

$ alien VRTSralus-11.00.7170-0.i386.rpm
vrtsralus_11.00.7170-1_i386.deb generated
$ alien VRTSvxmsa-4.4-021.i686.rpm
Warning: Skipping conversion of scripts in package VRTSvxmsa: postinst prerm
Warning: Use the --scripts parameter to include the scripts.
vrtsvxmsa_4.4-22_i386.deb generated

The next step is to install the .debs.

# dpkg --install vrtsralus_11.00.7170-1_i386.deb
(Reading database ... 45374 files and directories currently installed.)
Preparing to replace vrtsralus 11.00.6235-1 (using vrtsralus_11.00.7170-1_i386.deb) ...
Unpacking replacement vrtsralus ...
Setting up vrtsralus (11.00.7170-1) ...

# dpkg --install vrtsvxmsa_4.4-22_i386.deb
(Reading database ... 45374 files and directories currently installed.)
Preparing to replace vrtsvxmsa 4.4-18 (using vrtsvxmsa_4.4-22_i386.deb) ...
Unpacking replacement vrtsvxmsa ...
Setting up vrtsvxmsa (4.4-22) ...

Be sure to install libstdc++2.10-glibc2.2 if you haven’t already:

# apt-get install libstdc++2.10-glibc2.2

Link to the startup script:

# ln -s /opt/VRTSralus/bin/VRTSralus.init /etc/init.d/

Create beoper group and add root to that group:

# groupadd beoper
# adduser root beoper

Note Well:

The source of this post provides more information. You should also read their page as I have omitted creating configuration files and setting up the Windows side.

References:

http://newvibes.net/index.php/veritas-backup-exec-agent-for-unix-linux-on-debian

Using ffmpeg you can split out the audio and video components of any video stream it understands. In fact you can tell ffmpeg to transcode the audio as well. In my case the audio was already in mp3 format so I didn’t want to transcode. Ffmpeg allows you to specify “copy” as an output format in which case it doesn’t modify the audio stream in any way.

To extract the audio I ran the following. The -vn switch instructs ffmpeg to omit the video stream. The default behavior would be to output a video stream as well as audio streams. The -acodev switch tells ffmpeg what format you want the audio stream to be output as. In my case I didn’t want it transcoded so I specified “copy.”

$ ffmpeg -i slayer-rock-am-ring-2007.avi -vn -acodec copy slayer-rock-am-ring-2007.mp3

Now I just need to find a tool to break mp3s into audio tracks without re-encoding.

References:
http://howto-pages.org/ffmpeg/#strip

Thanks go to everyone that contributed code.

The latest version is 0.8.1.

I’ve been using Sync Kolab for months now but I’ve always been disappointed. Thunderbird has an option to choose the format of messages, either plain text or HTML, for each contact. This is great because I like the richer formatting available in HTML but HTML isn’t appropriate for many situations. Mailing lists are a prime example of when to use plain text. I’d finally grown annoyed enough that I started peering into the source code to add the feature myself.

What I found in the code was that the feature was already supported — if you are using the VCARD backend. Now that I’ve switched to the VCARD backend all is well!

While it is not possible without TLS extensions to serve different certificates for a single IP (See here on how to setup apache on Debian for TLS extensions.) it is possible to have a single certificate that works with any number of hostnames. I’m not talking about a wildcard certificate but a certificate that allows completely different hostnames to be valid for a single certificate. For example www.foo.com and www.foo.org can share a certificate. This approach would not be appropriate except in certain circumstances. You wouldn’t want to have two different customers using the same certificate but a single customer may wish to use one certificate for all of their domains. Both Internet Explorer and Firefox honor certificates of this type. From what I read some Java SSL libraries do not handle this type of certificate properly but Java was the only exception.

x509 certificates, those that are served in SSL communications, offer a feature known as Subject Altnerative Names. A subject Alternative Name is an attribute that lists an alternate name for the subject of the certificate (that’s oddly fitting isn’t it?). In a web context that subject is the hostname. However it’s not just hostnames that can be an alternative subject. Email is an option as is IP addresses.

The first step is to create a CSR (certificate signing request) that contains the subject alternative names that you desire for your certificate. I will show how to do that using openssl. You will likely need to modify the default openssl.cnf file. In Debian this is located in /etc/ssl/openssl.cnf. Note that you may prefer to make modifications to a local copy and tell openssl to use your locally modified copy using the -config option. For simplicity I will omit -config localopenssl.cnf from my examples.

Config File Settings

You need to tell openssl to create a CSR that includes x509 V3 extensions and you also need to tell openssl to include a list of subject alternative names in your CSR. In my openssl.cnf I have the following:

In the [req] section

[req]
req_extensions = v3_req

In the v3_req section:

[ v3_req ]

# Extensions to add to a certificate request
basicConstraints = CA:FALSE
keyUsage = nonRepudiation, digitalSignature, keyEncipherment

# Some CAs do not yet support subjectAltName in CSRs.
# Instead the additional names are form entries on web
# pages where one requests the certificate...
subjectAltName          = @alt_names

[alt_names]
DNS.1   = www.foo.com
DNS.2   = www.foo.org

Generating the CSR

Then the CSR is generated using:

$ openssl req -new -out $CSR_FILENAME -key $KEY_FILE

To check to see if you got everything correct use:

$ openssl req -text -noout -in $CSR_FILENAME

You should see something similar to this:

        Attributes:
        Requested Extensions:
            X509v3 Basic Constraints:
                CA:FALSE
            X509v3 Key Usage:
                Digital Signature, Non Repudiation, Key Encipherment
            X509v3 Subject Alternative Name:
                DNS:www.foo.com, DNS:www.foo.org

Creating the Certificate

Now you must have a CA (certificate authority) create a signed certificate based on the information provided in your request. Unfortunately most CA software will not honor the subject alternative names in a CSR by default. In the case of the certificate signing tools from Microsoft in Windows 2003 you can tell it to honor subject alternative names using the following:

certutil -setreg policy\EditFlags +EDITF_ATTRIBUTESUBJECTALTNAME2
net stop certsvc
net start certsvc

For openssl you need to use a policy that allows subject alternative names. I believe the policy named policy_anything in the default openssl.cnf file will work. To use that policy:

$ openssl ca -policy policy_anything -in $CSR_FILENAME -out $CERT_FILENAME

Finally to test that your certificate was created correctly use the following:

$ openssl x509 -text -noout -in $CERT_FILENAME

You should see something like this:

 X509v3 extensions:
            X509v3 Basic Constraints: critical
                CA:FALSE
            X509v3 Key Usage:
                Digital Signature, Key Encipherment
            X509v3 Subject Alternative Name:
                DNS:www.foo.org DNS:www.foo.org

The basic setup steps are:

1. Identify the correct jumper settings to enable SPDIF output on the VIA M10000.
2. Test digital audio connection using basic ALSA tools.
3. Setup MythTV to use SPDIF output device.
4. Setup alsa to re-sample audio to 48000Hz.

Jumper Setting

I remember when I originally built the frontend that there was a jumper that allowed you to select between SPDIF and Composite Video for the RCA style jack on the back of the board. I consulted the manual and set the jumper accordingly. Next I hit the net to learn what I needed to do to get the SPDIF output working. Try as I might the receiver would never output a peep. It didn’t even detect a digital connection. In desperation I set the jumper the opposite of what the manual said I should do for SPDIF. That worked. The version of the manual that I have for the M10000 is wrong. Don’t set the jumper based on your manual. Instead look at the printing on the circuit board. On mine one of the jumpers is marked as the SPDIF setting.

Testing Digital Connection

Since Myth, like any app, relies on the OS for sound output it makes sense to get the sound working using simpler tools before attempting to get it working correctly in Myth. ALSA includes a really handy tool called speaker-test that you can use to test your setup. This tool is found in the package alsa-utils on Debian systems. It generates a pink-noise signal alternating between channels. You can tell it the number of channels to generate. When it outputs to a channel it prints to stdout where it is sending audio to so you can be sure that you have everything wired correctly. You can also tell it what audio device to direct output to.

This is the command I used to test my connection:

speaker-test --channels 2 --device ALSA:iec958 --rate 48000

Setting up MythTV for SPDIF Output

Setting up MythTV for SPDIF output is pretty simple. All you have to do is tell MythTV what ALSA device to direct audio to. There’s almost nothing beyond that (well there is, see the next section). This setting for Myth is on the third page of the general setup section. The SPDIF output for the VIA M10000 with a default ALSA configuration is ALSA:cards.pcm.iec958. You can determine what ALSA devices are availble using the command aplay -L. Apparently iec958 is a spec that means SPDIF. You will have to type this in it is not automatically populated in the drop down box.

You should enable pass-through for DTS and AC3 so that if you watch sources with multi-channel audio your receiver will be able to play it. I don’t use Myth to watch DVDs or HDTV content but if I did I’d be extremely happy using this option!

If you use MythVideo with mplayer you may need or want to change your configuration to tell mplayer to use the correct output device, to do AC3 and DTS passthrough, and to resample the audio (possible). I found that mplayer would play videos and audio too quickly. This is because, apparently, it uses the sound card driver for timing. The videos had 44100 Hz sampling but the audio driver in the M10000 runs only at 48000 Hz. This results in a chipmunk effect. You can force mplayer to resample to 48000 Hz by specifying -srate 48000. This step isn’t strictly necessary if you have resampling done by ALSA (see the next section). An example configuration is mplayer -ao alsa:device=cards.pcm.iec958 -ac hwac3 -srate 48000.

Configuring ALSA to Re-sample audio to 48000Hz

The sound hardware in the Epia M10000 board from VIA supports only 48Khz output on the SPDIF output. This can be frustrating as software playing back 44.1 Khz audio or other sampling rates will be played back too fast. Luckily there is a way to have ALSA resample all of the audio. This will result in a loss of fidelity but I can’t tell the difference.

I created a file ~mythtv/.asoundrc. You could also, and probably more appropriately put this in /etc/asound.conf. Inside the file is:

pcm.!default {
        type plug
        slave.pcm "cards.pcm.iec958"
        slave {
                pcm "cards.pcm.iec958"
                rate 48000
        }
}

I believe this instructs ALSA to resample all audio coming into ALSA:default to 48000 Hz and direct it to the ALSA device cards.pcm.iec958 (the SPDIF output). I find the confiuration syntax to be very cryptic. In particular, if my interpretation is correct, then I’m perplexed by “!default” as I would normally interpret that as “not default.”

If you use this technique be sure to change the Myth audio settings to use ALSA:default as the output device. Once you do this playback in Myth will be correct. You can also instruct mplayer and other ALSA capable software to use this device and playback should be correct.

References

I had a lot of help getting this to work.

• MythTV Wiki:Configuring Digital Sound (This is a great resource)
• MythTV Wiki:Configuring Digital Sound with AC3 and SPDIF
• EPIA Linux Howto (specifically the entire section on audio)
• ALSA Wiki:VIA8233