It works by caching all the results from a JSON Request and feeding them to the autocompleter object. When a item is added as a box, it's removed from the feed array, and when the box is disposed it's added back, so that it becomes available in the list when the user types. It didn't take long before glitches and features were reported, so first of all, i'd like to thank all of you for taking the time to check out the code and report all the bugs you found, it really helped taking the script further, i'm sure that the next build will not be perfect and that's why your help is still needed to keep checking and suggest new cool features for next versions. In this build I tried to include all the suggested features, I hope that nobody will be disappointed. There are also few bug fixes but most features upgrade. I encourage all of you to suggest any features you have in mind, because what might be useful for you, may be useful to others, so please do so.. The new version incorporates few new options: * fetchFile - default: undefined; a JSON file that is being called with the parameter keyword whenever a key is being typed * results - default: 10; the number of results to display before showing a scrollbar, if more results match a scrollbar is displayed * wordMatch - default: false; when set to true will match only the beginning of word, otherwise will match anywhere