TwonkyScrobbler V1.0.1

I have released an update to the TwonkyScrobbler. A user on the Twonkyvision forum highlighted a problem when using it with an older version of the server.

TwonkyScrobbler was developed and tested using the latest TwonkyMedia version 4.4.9. The RSS feed facility was introduced in version 4.1 and I had made the incorrect assumption that the RSS feed format had remained unchanged.

In Version 4.4.9 I observed the items in the RSS feed have the following format

<item>
<title>Gemini</title>
<link>http://192.168.0.20:9000/disk/O1$11$207905894$2623831632.mp3</link>
<description>Eye In The Sky</description>
<author>Alan Parsons Project, The</author>
<enclosure url=http://192.168.0.20:9000/disk/O1$11$207905894$2623831632.mp3 type="audio/mpeg" />
<pubDate>Mon, 01 Jan 1981 01:00:00 GMT</pubDate>
<moddatetime>Thu, 20 Apr 2006 20:38:10 GMT</moddatetime>
<playeddatetime>Sat, 08 Nov 2008 07:44:01 GMT</playeddatetime>
<class>object.item</class>
<genre>Rock</genre>
<playcount>1</playcount>
<tracknumber>32148</tracknumber>
<duration>131000</duration>
<bitrate>192</bitrate>
<frequency>44100</frequency>
<audiochannels>2</audiochannels>
</item>

The scrobbler was using the element <playeddatetime> to determine if the song was a new one, storing the value in the program configuration file.

Following the bug report I installed a much older version (4.2.1) on another PC and discovered that items in the RSS feed have much less information.

<item>
<title>Old Ghosts</title>
<link>http://192.168.0.21:9000/disk/O1$11$134217730$2550136852.mp3</link>
<description>Stormwatch</description>
<author>Jethro Tull</author>
<enclosure url="http://192.168.0.21:9000/disk/O1$11$134217730$2550136852.mp3" type="audio/mpeg" />
<pubDate>Mon, 28 Apr 2008 12:59:26 GMT</pubDate>
</item>

This obviously created a problem because the required information is missing, the <pubDate> element isn’t related to the time the song was played.

To attempt to remedy this problem I have modified the scrobbler so it attempts to find the element, however if it is missing the other information title,description and author (i.e. title, album and artist) is compared to the last scrobbled track and any difference is assumed to signal a new track and the new track is scrobbled using the current time as the played time.

Twonkyvision had made reference to this fact in the RevisionHistory.txt file

What's new in Version 4.4.3.1
------------------------------

+ DivX and WMV support for PS3 (4.4.3.1)
+ AAC support for PS3!
+ Time information for pictures on PS3
+ added Album Artist and Album Art for OGG and FLAC
+ added rss feed support for all content types. Start with http://127.0.0.1:9000/rss/feed
+ enhanced generated rss XML feed by more properties

Updating from V1.0 will require a new configuration/state file with the extra data.

TwonkyMedia Scrobbler V1.0 Released!

Tested it again last night with no problems.

So have released it to the world - have even done some documentation

http://www.insanityworks.co.uk/twonkyscrobbler/index.htm

Scrobbler nearly there!

Well the scrobbler is nearly there, done some fiddling with it's look!

Twonkymedia last.fm scrobbler

Well I haven't posted anything for a while, but today I finally completed the first stage of a programming project I hinted back in February.

It is a scrobbler for the last.fm service that sits and monitors the TwonkyMedia uPnP server and sents information of tracks it is servicing!

The TwonkyVision people added an RSS service (think it was on version 4 and primarily to support non-upnp devices such as the SonyPSP) This is available at http://localhost:9000/rss (that link won't work unless you are reading this on a computer running the server!)
By drilling down from this page it is possible to reach a 'last played' rss feed.

It is buggy, the feed isn't updated correctly, but appears the last entry in the rss feed is always the last audio file served up. My scrobbler which is based on the KexpScrobbler by Markus Palme reads this file and extracts the appropriate information from the last entry and scrobbles the data.

I have to do a bit more testing but looks like it works, and I will be releasing it on my website soon!