That’s me.

Your results:
You are Spider-Man

Spider-Man 70% Green Lantern 60% Robin 60% Hulk 50% Superman 50% Supergirl 45% The Flash 45% Iron Man 45% Wonder Woman 35% Batman 20% Catwoman 15%

You are intelligent, witty, a bit geeky and have great power and responsibility.

Click here to take the Superhero Personality Quiz

This is a text, which I’ve had sitting as a draft in my blog for quite a while. It was only today, that I took the time to recheck something, that I wanted to have checked before publishing this entry.

It’s basically all about attaching Apaches LDAP authentication module to the Windows Active Directory. It took quite a bit of nerves to get it actually working. The problem was, that some tools (ldapsearch or Java JNDI) appeared to be working against the AD, while the actual authentication against the AD didn’t appear to be working.

The bit of information, that got it working in the end is, that the AD provides the standard LDAP port 389 and a so called “Global Catalog” on port 3268. While they both provide the same information for my purposes, they behave somewhat differently. This can be best seen by what is displayed by ldapsearch.

Here is a query against the standard port number 389 (actual query output deleted):

# ldapsearch -x -p 389 -h ip-addr -D user@XXX.ZZZ -W cn=ridderbusch*
....
# search reference
ref: ldap://ForestDnsZones.xxx.zzz/DC=ForestDnsZones,DC=xxx,DC=zzz

# search reference
ref: ldap://DomainDnsZones.xxx.zzz/DC=DomainDnsZones,DC=xxx,DC=zzz

# search reference
ref: ldap://de001.fsc.net/DC=de001,DC=fsc,DC=net

# search reference
ref: ldap://fsc.net/CN=Configuration,DC=fsc,DC=net

# search result
search: 2
result: 0 Success

# numResponses: 6
# numEntries: 1
# numReferences: 4

As can be seen this query entry displays for additional ref: lines. I think, this what throws the Apache LDAP module off the road. If you compare this against the query to the 3268 port you don’t see the ref: lines.

ldapsearch -x -p 3268 -h ip-addr -D user@XXX.ZZZ -W cn=ridderbusch*
# search result
search: 2
result: 0 Success

# numResponses: 2
# numEntries: 1

It appears, that some tools, that can be used to query the AD LDAP have a problem with this behaviour, while other work perfectly.

Since our company AD doesn’t allow anonymous binding against the AD, it’s best to contact the AD administration to create a “service account”. The passwords of these accounts don’t expire. In the end, this entry in a .htaccess works for me (for the auth_ldap module in Apache 1.3.xx).

AuthLDAPBindDN "service account@XXX.ZZZ"
AuthLDAPBindPassword "abcdef"
AuthLDAPURL ldap://wdc.xxx.zzz:3268/dc=xxx,dc=zzz?sAMAccountName

Another interesting bit of information I learned is, that there are different arguments, which can be used as arguments for the -D, when using ldapsearch for binding (which might be a matter of configuration inside the AD, I don’t know).

• The standard (?) domain login: domainlogin@XXX.ZZZ This can also be used to initiate single sign on through Kerberos.
• An (your) email address: frank.ridderbusch@company.com
• A standard LDAP DN: CN=Ridderbusch\, Frank,OU=Users,OU=DE-Paderborn,DC=xxx,DC=zzz

Here are some additional links, which I found through Google and which offered some interesting bits of information.

• http://www.petri.co.il/ports_used_by_exchange.htm
• http://support.microsoft.com/?kbid=278339
• http://www.stbernard.com/products/docs/ip_technotes/ldapwin2k.pdf

Some time ago I bought a used Hauppauge WinPVR 150 TV card. I briefly tested it under Linux and it basically worked fine. However all the recipes I found on Web sounded somewhat intimidating and it appeared to me, that quite a bit of learning would be involved. Therefore I moved the card to my Windows box and used the supplied software.

However, in the end I’m back to the Linux for creating DVDs from the recorded programs. The included editor for manipulating the recorded program under Windows is total crap. The Ulead software for producing the DVD is better, but with my current work-flow, process of editing and creating the DVD is much faster under Linux.

I’m still recording the TV program under Windows simply for the fact, that I can very quickly “program” the recording with the Hauppauge WinScheduler. It would work under Linux, but there is only a command line tool as part of the ivtv-driver and it is somewhat tidious to “program” the VCR with the at and record-v4l2.pl commands. Then the recording is transfered to the Linux box over the network.

Editing is then done with Avidemux, which is very good. Much better than the program included with the Hauppauge software.

The toolkit of choice for actually creating the DVDs is Tovid. It can’t be any simpler as with this set of programs. It typically involves there steps:

• tovid -dvd -pal -in recording.mpg -out file-1
  This step would do any convert the source file into the appropriate format for the DVD. But since the recording originally was in the correct form (MPEG Program-Stream Audio+Video) and was saved in the same format by avidemux no processing is required, the output file is replaced with a symbolic link to the input file. So, if you are always following the same procedure and record with the same parameters, you can actually save this step.
• makemenu -dvd -pal -fontsize 30 -background ~/graphics/scans/canada.jpg -align center "Title 1" "Title 2" -out menu
  This will create a menu for the DVD. The menus on a DVD are actually MPEG files as well. In this example I used a custom background and increased the font size. If you’re only putting one recording on the DVD and don’t care too much about a finished look, you can save this step as well. In this example I created a menu with two entries.
• makexml -dvd -overwrite -menu menu.mpg -chapters 3 file-1.mpg file-2.mpg out-file
  This step will create a XML file out-file.xml, which is used in the final production step. To allow navigation when you’re watching the movie, the chapters option creates chapters in 3 minute intervals.
• makedvd -burn -device /dev/hdd out-file.xml
  This will create the DVD file structure and burn the DVD.

Tovid is only a collection of bash scripts, which connect the various required tools in a meaningful way. Here is an extract from the configure run.

checking for grep... ok
checking for sed... ok
checking for md5sum... ok
checking for mplayer... ok
checking for mencoder... ok
checking for mplex... ok
checking for mpeg2enc... ok
checking for yuvfps... ok
checking for yuvdenoise... ok
checking for ppmtoy4m... ok
checking for mp2enc... ok
checking for ffmpeg... ok
checking for sox... ok
....
Required dependencies   ALL FOUND

Other dependencies      ALL FOUND

  magick (needed to make menus and slideshows)
    composite:    ok
    convert:      ok

  dvd (needed to make DVD menus and burn DVDs)
    spumux:       ok
    dvdauthor:    ok
    growisofs:    ok
    mkisofs:      ok

  vcd (needed to burn (S)VCDs)
    vcdxbuild:    ok
    cdrdao:       ok

  transcode (needed to compress encode videos)
    tcprobe:      ok
    tcrequant:    ok

The RawShooter Essentials 1.1.3 raw converter works flawlessly under Wine. I used it to convert about 60 pictures from my Canon 350D, as I already said, without any problems.

The Wine version 0.9.5 is installed on my Gentoo system and additionally the WineTools 0.9, which create the fake Windows environment into which the Windows applications are installed and operate in. By default the fake environment is created in the users home directory in .wine/drive_c. Therefore I moved the RawShooterEssentials installation archive into the directory .wine/drive_c/temp and executed:

. findwine
$WINE "C:\\temp\\RawShooterEssentials2005-113.exe" &

This installed the raw converter into the fake Windows environment. After that, I created this little shell script (modelled after the IE6 start-up script from the WineTools):

#!/bin/bash
. findwine
$WINE "c:\\Program Files\\Pixmantec\\RawShooter essentials 2005 1.0\\RawShooter.exe" \
&>/dev/null &

Now I can convert my Canon raw files quite comfortably and make some initial adjustments and improvements to the photos.

Heute sind meine Tocher Meret (7 Jahre) und ich in dem Film “Eine zauberhafte Nanny” (“Nanny McPhee” im Original) gewesen. Der Film ist aus meiner Sicht empfehlenswert. Im Vorfeld hatte ich einige Reviews im Web und im Spiegel gelesen, die den Film zwar nicht total verrissen, aber auch nicht hoch gelobt haben. Nach diesen Bemerkungen war ich schon fast geneigt, mir den Film nicht anzuschauen. Den Auschlag dann jedenfalls ein Kommentar gegeben, den ich irgendwo gelesen habe, das dem Betrachter der Film gut gefallen und im besonderen besser als Narnia gefallen hätte. Narnia habe ich auch gesehen und den fand ich bestenfalls “ganz nett”.

Ich froh, das wir heute gegangen sind. Meiner Tocher und mir hat er jedenfalls gut gefallen und nach Reaktionen zu urteilen den übrigen Kinobesuchern auch, besonders den Kindern. Ein schönes Märchen.

Wir hätte nur vielleicht doch eine ode zwei Wochen später gehen sollen. Das Kino war recht gut gefüllt und dank Popkorn und sonstiger Knistereien von Kindern, die es mit der Ruhe nicht so genau nehmen, war es insgesamt sehr unruhig. Mit etwas mehr Ruhe hätte man den Film noch besser geniesen können.

Today I gave the Oracle XE database edition a quick spin on a Gentoo Linux system. The download is about 200Mb in size and is installed through the RPM package manager. Not every Gentoo user might have it installed, since Portage is the standard, but I had it installed. Additionally I needed to unmask libaio in /etc/portage/package.keywords and install it. Executing “rpm -i --nodeps” should then have installed the Oracle package, but the preinstallation procedure insists, that enough swap space must be available (1Gb) and exited the command. However, this could be easily fixed by adding some swap space through an ordinary file created by “dd if=/dev/zero of=swapfile bs=1M count=128; mkswap swapfile; swapon swapfile“. After this the rpm invocation succeeded.

After installation Oracle XE needs to be configured by executing “/etc/init.d/oracle-xe configure. You will see an error message, that the file /etc/init.d/functions can’t be sourced. This doesn’t stop the configuration phase to finish successfully. However the /etc/init.d/oracle-xe can’t be used to start the Oracle processes on Gentoo. It apparently only works for Redhat or SuSE. This is no real problem for me at this point, since I’m not planing to have Oracle constantly started on each boot. It can be easily enough started on demand by executing

su - oraclexe -c "sh /usr/lib/oracle/xe/app/oracle/pro...
duct/10.2.0/server/config/scripts/startdb.sh

(very long path) or stopped with stopdb.sh. I guess someone from the Gentoo folks will create a proper Gentoo startup script, once Orace XE leaves beta status.

So after that, you’ll experience the “real” Oracle feeling. My main experience comes from working with 8.1.7, so this 10g is quite a bit newer. I will see how it goes to import my old 8.1.7 development data into 10g.