Known Problem

Helix’s logging feature prevents data loss by recording all transactions in a logfile. If Helix crashes, it attempts to apply the logfile the next time the crashed collection is opened. This transaction logging is a major benefit of using Helix.

The format of the logfile is not the same when running Helix for PowerPC as it is when running Helix for Intel. Attempting to apply a logfile created under one architecture while running under the other will result in an error dialog informing you that the logfile is damaged.

The simple solution is to always reopen a crashed collection with the same Helix application (being sure to match the original architecture) as was running at the time of the crash. Then Save and Quit: Helix will automatically dispose of the logfile when the collection is closed correctly.

If you are unable to reopen the collection with the same Helix application, you can continue by responding to the dialog as appropriate for your situation. The dialog offers three choices: Locate, Cancel, & Proceed.

• Locate: In this situation, Locate is pointless. Helix knows where the logfile is, it simply can not read it correctly. (Unless the problem really is that you have moved the logfile and Helix needs help locating it.
• Cancel: Close the collection, leaving it and its logfile intact. You can then open the collection and logfile with the original Helix appication/architecture and apply the logfile.
• Proceed: Open the collection, wiping out the old logfile and creating a new one. All transactions that were preserved in the old logfile will be lost if you choose this option.

This problem affects version 6.1 or higher of Helix Server & Helix Engine.

Helix Client is not affected by this issue, as it does not maintain a logfile. (Logging is done by the Server.)

Helix RADE will also be affected by this issue when it is released as an OS X native application.

This issue will not be addressed. Eventually PowerPC machines will be obsolete and this will no longer be a problem.