Hi improving 'objective' performance is great, but in the case of the address book one could also think about addressing the 'subjective' performance. Let me try to explain ...
Usually when I enter an address, the number of characters in the To line are sufficient to narrow the number of possible hits considerably. In addition one often uses the same addresses over and over again, at least during a certain time period. The set of 'most' used addresses may change in time, but usually rather slowly for the normal Joe User. One could keep statistics of the characters entered before requiring completion and of the respective address selected. This statistics should obviously be kept from session to session. With this information loaded in memory quite a few nice tricks can be implemented which may avoid using the full regex scan most of the times. For instance the most frequently used address for a given set of entered characters could be given as a first proposal upon entering a TAB. An aditional TAB could show the next in terms of usage frequency or start a full search. I guess this would be simpler to implement and give a great increase of responsiveness ... Just an idea. -- Michele Ravani [EMAIL PROTECTED] "Those who live hoping, die singing" My Gran ------------------------------------------------------- This sf.net email is sponsored by:ThinkGeek Welcome to geek heaven. http://thinkgeek.com/sf _______________________________________________ Mahogany-Developers mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/mahogany-developers
