Before updating my Infobright patches, I decided to to a closer a look at what actually changed from 3.3.0 to 3.3.1, and to my surprisem, a new very big feature has been implemented, namely UTF-8 support.
Why would they implement something like this, even partially, without telling the public?
Even a tool called CharSetMigrationTool have been included in the source distribution, don't know about the binary distributions. The tool seems to update the Knowledge Grid to support UTF-8.
Just by looking at the source code, it seems like UTF-8 support has been fully implemented with a few exceptions, like CMAPs for UTF8. CMAPs are explained in more detail in this forum post.
On the ICE forum I posted the question, has it been implemented or not, the post can be found here.
Other than that, no untold changes seems to be hiding in 3.3.1 release.
UPDATE: One of the Infobright developers replied to my forum post, and he confirms that UTF-8 support has been partially implemented in ICE 3.3.1. The next release of ICE should contain UTF-8 support for the ICE loader process. See the reply here.