You solved your problem faster than I was able to answer. :)
The solution is indeed to go into the directory you want to use before calling the "java" command.

If you really want to be ablre to specify your "working directory" (although I can"t think of a scenario where you could not simply go into the directory to launch Ubooquity), you can do it by explicitely setting the Java "user.dir" property:

java -Duser.dir="/my/directory" - jar /path/to/the/jar/file/Ubooquity.jar

But I don't really recommend that solution as it could have unintended side effects (I don't know any but it's safer to just use your method).

I'm glad it's working now. Don't hesitate to let me know if you have other questions.

It helped a lot. I use mostly Firefox, so I dind't notice the shortcuts were not woking in Chrome.
The bug has been fixed. Shortcuts will work for all browser in the next release (1.6.0).

Thanks!

Well, there are already keyboard shortcuts usable when reading a comic online:

• ← and → (arrow keys) to go to previous/next page
• ENTER to display/hide the overlay
• ESC to close the comic book

If these shortcuts do not work for you, could you tell me which browser you use and on which OS so that I can investigate ?

I am not using Ubuntu myself, so I did not test anything, but you should be able to run Ubooquity as a daemon by following this tutorial:

http://zerocool.is-a-geek.net/running-java-program-as-a-daemon-in-ubuntu-linux/

Where would the data come from ? 
Ubooquity relies entirely on what it find inside the comics/books files. Do you have some files containing the informations you mentionned ? (index, summary, authors...)

That could be done. Although I suppose that a consistent enough naming
scheme should be sufficient to organize your content by date.

I've been asked for this feature several times.
Ubooquity is a "content server": its purpose is to allow file access, not to modify them. So a metadata grabber would have to be done in another project (Ubooquity never modifies your files, and storing metadata outside of the files they describe is a bad idea).

I have almost started working on a metadata grabbing software. But after having dug a little into the amount of work required to have something efficient and easy to use, I have realized it was not a realistic task.

At some point in the future, Ubooquity will read and make use of the metadata contained in the "comicinfo.xml" file some comic files contain. But this file will have to be generated by another software.

I'm not sure I understand what you mean by "sub index".
Do you mean a specific way of organizing comics, like by series ?

One thing I have planned for the future is to read and use the "ComicsInfo.xml" metadata file contained in somes comics. It has informations on the current series, the number, the date, the authors etc. So I'll probably add sorting options related to these additional metadata at the same time.