Jezusmina, zoveel bugs dat er nog in ColdFusion zaten: een nieuwe updater zou er iets aan moeten doen, ik zie toch niet te veel known issues meer… Een aantal van de bugs die in Updater 3 gefixed zijn:

• Search engine safe URLs of the form *.cfm/* did not work properly, so a URL such as http:/server/test.cfm/alpha/beta would not execute test.cfm.


• The cfmail tag did not encode the strings in any field except Subject. All fields, including , such as To, From, CC, and Reply-to, are now encoded using the same technique as the Subject field.


• The DateCompare function incorrectly returned 0 if the two parameters were within one hour of each other.


• The QueryAddColumn function generated java.lang.ArrayIndexOutOfBoundsException exceptions when adding a query column to more than 17 rows.


• Under some circumstances, ColdFusion MX set the size of some of its properties and xml files to 0 bytes.


• ColdFusion could cause unexpected crashes in the Microsoft IIS web server.


• In Updater 2, ColdFusion did not properly handle escaped single quotes in queries.


• The driver was returning the euro character as a ? when connecting to SQL Server 7.


• .Net web service clients could not consume any ColdFusion component (CFC) web services that returned a query.


• Access Violations and -12 errors occurred when the connector was subjected to load before JRun has come up completely. This problem affected all web servers.


• A memory leak existed in the connector when multiple HTTP requests were canceled.


• The jrun.xml file was configured such that ColdFusion MX Administrator Simultaneous Requests settings up to 19, including the default value of 10, could cause server instability or reduced performance.


• Updater 2 introduced a memory leak into the IIS connector that could cause performance to degrade over time when running a ColdFusion MX application under load.


• The jrun.xml file did not set mapCheck=0. This could reduce performance or cause server instability on some platforms.


• String operations such as ReplaceNoCase and FindNoCase on strings that contain the German ß character (scharfes S or double S) caused errors.

Een mens vraagt bijna zich af wat er wél werkte…