If you want to receive an email whenever a new release is put out (vs. getting the reminder from the program itself), subscribe to this thread. I will post all new changes here.

MediaReader 0.71

Released a new build. It fixes an issue where the content rating was not properly identified. Updated opensubtitles.org api. More importantly MediaReader now uses Traileraddict for the trailers. IMDB kept changing its layout and finally prefixing the content with ads - a constant mess.

MediaReader 0.7

Two enhancements and a few bug fixes (actually, I haven't had much complaints or problems so it seems it is pretty much stabilizing). The two changes are important:

a.) The code no longer does any specific checking whether to use the Lobo browser or Mozilla. It has turned into too much guesswork. Instead, go into your profile and tick 'Use Lobo browser' if you need it (mostly in 64 bit environements). Let me explain the hubbab about Lobo vs. Mozilla quickly: MediaReader needs some sort of browser at least for the 'rescrape function' for movies. Were it a sole Windows app it would use Microsoft's ActiveX control which essentially runs IE. However, it is not. SO it uses the Mozilla/Firefox angine. There is a fairly robust interface between SWT (a UI framework for Java and Mozilla). However, MediaReader is written in Swing which has a very different internal event dispatching mechanism. When interfacing between Java and Mozilla (which is written in C++), Java has to use some glue code to talk to it. Somebody did this and it is fairly complicated business. However, it appears that this project was abandoned at some poitn and the glue code never updated to 64bit. Ergo 64bit Java cannot talk to Gecko. You *can* run 32bit Java under 64bit Vista/Windows7/Linux. You cannot run 32 bit Java under Mac OSX because the guys at Apple are jerks. (I hope this is technical enough of an explanation). So if you run 32 bit Java you can use Mozilla.

For everyone else there is only Lobo, a 100% Java based browser which has many problems. However, it is good enough to sort-of find the movie you are looking for at IMDB. IT has too many problems to be used as a general purpose browser. It also doesn't support plugins.

Now Sun has announce a Swing component called JWebPane a long time ago (years), it has demoed it last summer and there are consisten news about its development status. JWebpane can render web pages and represents the solution to the whole dilemma. However, it appears SUn is asleep at the wheel because this stuff just isn't being release. As soon as it is (and I hope it is soon) MediaReader will have a full rendering engine. SO much for the background.

b.) The code is now sort-of networkable. This means that you can run the databse on a central server, theoretically using different SQL engines (they do not have to be written in Java as long as there is a JDBC interface. However, I have only tested the code with JavaDB (Derby). This is neat as you can now run all your data on a single box (very, very little overhead so feel free to do it on HTPC itself) and look at it from any PC in your household. There are limitations. MediaReader is not a client-server application, so there are caveats. Please read NETWORK.txt for details on how to set it up. You can do it in 5 minutes and not lose any data.

version 0.61

Someone asked me to provide some sort of facility to pick the movie name from the file name in a smarter way than to brute force plug it inot the IMDB search engine. It made sense to me, so I added the Movie Seearch Patter (profile). It is a regular expression describing a capturing group derived from the file name (if you don't know what that is - never mind, by default everything works as before). Let's say all your movies have titles that are followed by the name of the year of the movie in parentheses followed by this or that - essentially the name of the movie is everything before the open parentheses for the year. If you define a capturing group of '(.*?)\(' only the string before the '(' will be used as title. However, a few other things apply as well: First of all during the search the file name's extension is always stripped. Also '_-?*' will be translated to blanks.

Enjoy.

version 0.60

• Mediareader at times was not picking the obvious movie from IMDB but some other obscure information.
• Code comes with a refreshed Mozilla/Swing interface (refreshed from it's svn).
• Added a new function to relocate movies from single movie file into individual directories (drop me a note if you don't quite understand how it works, I'll document it when I find the time).
• Added support for new nfo file format (fileinfo section). Mediareader supports any number of audio streams and any number of subtitle streams. The XBMC Wiki is not clear on what the format tag should look like (AVC or A/VC, AC-3 or AC3 or ac3?). So if there are better specs, let me know and I will change it.

At present the information for the nfo file is *always* read from the physical moviefile and audio/subtitle info cannot be edited. As I am in the middle of writing Mediaserver (unless someone has this title indeed trademarked) I only want to put the absolutely necessary effort into the old code. Mediaserver supports any number of streams including multiple video streams.

I yanked the 'Aeon style studio tag support' as this seems deprecated.

Version 0.58 9/04/09

Makes the movie edit dialog scrollable (for smaller screen resolutions).

Version 0.57 9/04/09

TV scanning hangs on lack of TVDB response or show not found/outage.

Version 0.56 8/29/09

Custom selected movie art is not written to disk.

SQL error when scanning DTS encoded movies.

Version 0.55 7/31/09

SQL error when scanning DTS encoded movies.

Added ability to to create 'dummmy' TV series/seasons/episodes (Check readme).

A few minor bugs scanning TV series.

Version 0.54 7/27/09

A few minor bugs scanning TV series.

Version 0.52 7/27/09

Embarrassing. Two typos causing grief. Grab this instead of 0.51.

Triggered the update alert because the cumulative value of the past few updates probably makes it worth it.

Version 0.51 7/26/09

Changelog:

MediaReader 0.51

Movie trailer not rewritten upon edit (suppose you picked a Youtube trailer).
Movie trailer playback button does not light up correctly. (Well, that wasn't ready when I hastily released 0.50).
Exception editing TV episodes without rating.
TV Series display now highlights incomplete Series/seasons/missing episodes.
Media Reader can now pick up any IMDBIDs in movie titles (as per request).
Nullpointerexception in GetMovies
Semi-automatic thumbnail creation
Error initializing browser under 64 bit Linux
You can now pick the fanart and graphics to write manually.

How to glean IMDB ID from movie title: Go into profile and define a cpaturing group using a regular expression (Sorry, I really could not make this easier, contact me with some sample titles and I will give you the proper expression).

How to use auto-thumbnails:

a.) You will need a working copy of ffmpeg. I have included one built from svn for Windows. Most Linux distros allow you to use the package manager to install one. For Mac OSX follow the instructions here: http://www.haykranen.nl/2007/11/21/howto-install-and-use-ffmpeg-on-mac-os-x-leopard/. It sounds harder than it is. Note that the actual install step will take some time (10-20 minutes as tons of packages are downloaded and built). Also note that you will need the latest version of Xcode from the Apple site. I had no problems building it. If you have issues, please report on the 24hourloop forum. No ffmpeg - no thumbnails. This feature is somewhat experimental so I'd like to hear of problems.

b.) Go into movie art for a movie and click *outside* and existing graphic (do not select one left-clicking) and choose 'Generate thumbnails). This will generate x thumbnails in intervals specified in your profile. It will place these thumbnails into your thumbnail directory (also specified in profile).

c.) You must then import whatever thumbnail you like into your database by right-clicking in the movie art dialog and selecting it via the file dialog (not the browser).

d.) If you want it written for XBMC you must right-click on the fiel in the movie dialog and pick 'fanart' or 'thumbail'.

Right now the thumbnails are captured at the resolution of the actual movie.

I am pondering whether to at some point to do so automatically for movies that don't have thumbnails/fanart.

Sorry for taking so long but I was busy producing this video: http://24hourloop.com/component/content/article/58 - feel free to donate to these folks (my daughter is in the video ;-).

Enjoy.