I know the combo box performance issue is widely known and many have been working towards a solution. I've posted an application (link below) which I developed for benchmarking combo box performance. It may be of help to anyone working on combo box performance.
The application times a process whereby a cascading series of combo boxes query automobile data from a server. The app allows the user to dynamically add or remove dummy combo boxes (in blocks of 10) and then time the data look-up process again. Here it is: http://eio-online.com/solocombo/comboperformance.lzx.html This automobile data combo box setup is a component of larger application that we have developed. The user selects the vehicle year in the top combo box and the 4-5 boxes below it are repopulated. Here are results of a test I just ran (time elapsed in mins : secs : millisecs): 0 combo boxes added, select vehicle year 2006 - time elapsed 0:1:344 10 combo boxes added, select vehicle year 1990 - time elapsed 0:2:609 20 combo boxes added, select vehicle year 2006 - time elapsed 0:7:47 30 combo boxes added, select vehicle year 1990 - time elapsed 0:8:984 40 combo boxes added, select vehicle year 2006 - time elapsed 0:23:609 50 combo boxes added, select vehicle year 1990 - time elapsed 0:20:391 60 combo boxes added, select vehicle year 2006 - time elapsed 0:43:937 70 combo boxes added, select vehicle year 1990 - time elapsed 0:42:157 80 combo boxes added, select vehicle year 2006 - time elapsed 1:20:547 As you can see the response times rapidly get longer as you increase the number of combos in the app until the app boggs down. The larger application we've developed contains in the area of 70-80 combo boxes (no way out of this. Check boxes or other components won't fit our layout). The program's load time and performance are currently unworkable. They are keeping us from being able to deploy so we are very hot to try to solve this combo box issue. We're currently looking at the combo box implementation in hopes of working towards a vastly stripped down alternative but we are not well versed in it's construction or inner workings. Pointers are welcome :) Our hope is that others will also continue the pursuit of a solution and post further developments as they occur. Thank you. -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.1.385 / Virus Database: 268.3.5/300 - Release Date: 4/3/2006 _______________________________________________ Laszlo-user mailing list [email protected] http://www.openlaszlo.org/mailman/listinfo/laszlo-user
