Pace of the character reading varies and rambles on. The problem here is that it will only read what is on the page and if it is a long drawn out page (without much pause or break) that's just the limitations of this software or programming. Also mispronounces some words is the limitations of text 2 speech engines. Keep in mind that you are on the "cutting edge" of text-2-speech and voice recognition programming/ software and this is not a perfected science as of yet!

  "Skins" do not display well in lower (640x480) resolution or if system has older display hardware. This is a known problem and explained at first startup by your talkster (i.e. oh by the way, best view in ...).

When character tells the time = puts a "point" before the time value.

Talkster just hates  web sites using frames! (I do too! It has caused MANY extra hours of programming and the problem is still not completely solved).