Everything Else
TS1713: Server 6.0 Crash on Data Replace


The Problem

A crash has been observed at multiple locations using Helix Server 6.0 (058). The trigger for this crash is obscure, but basically it comes down to a failure to cope with a specific error that can occur when the implicit "find next (or previous)" that occurs when a record on a queried entry view is replaced and that replacement causes data to change in a way that the query re-evaluates and no record is next to be displayed.

Instead of trying to understand that, the best way to determine whether or not this bug is affecting you is to examine the crash log that is produced by macOS when an application crashes. You can find these logs in the user’s ~/Library/Logs/CrashReporter/ folder. The name of the Server's crash log is, quite naturally Helix Server macOS.crash.log. Open the crash log with any text editor (double clicking it should open the Console application, which also opens another window — don't let that confuse you) and you should see something very similar to the following lines:

Version: 6.0 (058) (6.0)

PID: #####
Thread: n

Exception: EXC_BAD_ACCESS (0x0001)
Codes: KERN_PROTECTION_FAILURE (0x0002) at 0x00000000

Double check the date of the crash (a few lines above that) and make sure you are looking at a current crash. (The crash log is appended to each time there is a crash: the earliest crash is at the top and the most recent crash is at the bottom of the file. For this reason you should delete this file — or move it out of the CrashReporter folder — after reviewing it to avoid re-reading the same data.)

Note the thread number and scroll down through the log until you find that thread. When you get there, look for the following call chain:

Thread n Crashed:
0 Helix Server macOS ... GetNext + 212
1 Helix Server macOS ... P_FVIFNext + 116
2 Helix Server macOS ... FVIFNext + 48
3 Helix Server macOS ... MulFNext + 64

If those are the calls you find there, your Server is crashing because of this specific bug. If your crashed thread has any other sequence of calls, then it is not this bug and applying this fix will not help you. But as always, send your crash logs (using the proper procedure!) in so we can find and fix your problem too!

The Solution

Updating to Helix Server 6.0.1 (059) addresses this issue. (It is not necessary to update the Clients; only Server must be updated.)

Alternatively: replacing the Enter command in sequences with Static Enter has proven to be the most effective workaround.


This information applies to Helix Server 6.0 (058) only.


This bug is resolved in 6.0.1 (059).


Special thanks go to Tim Annis for providing us with a reproducible example of this bug.