As a result of trying to solve MH-6958, a few changes had to be made both to the java services architecture as well as to the 3rd party tools installer.
In a nutshell, we switched text extraction from Ocropus to Tesseract (which is used by Ocropus internally anyway). Tesseract would not work with jpegs, though, so the commandline for the ffmepg image extraction had to be changed, too. So at least we don't have to compile Ocropus anymore. Additionally, we now have two text extraction implementations that can be swapped out (though only one is currently working due to a somewhat broken ocropus installation). This means that people doing QA work need to do the following: - Re-run the 3rd party installer - Update matterhorn-movies.properties from svn - Clear the bundles in <felix.home>/matterhorn - Rebuild matterhorn Please keep an eye on those movies that used to produce the grey preview slides and make sure we get text from those movies. Josh & Tobias _______________________________________________ Matterhorn mailing list [email protected] http://lists.opencastproject.org/mailman/listinfo/matterhorn To unsubscribe please email [email protected] _______________________________________________
