Recent changes to engine-bugs

FireFox 4

Ricardo Cantu — Tue, 22 Mar 2011 19:45:30 -0000

Default font is not monospace.
Extension is not compatible.

Correct CON_SRCH config search routine

Dean Lyles — Thu, 17 Mar 2011 17:21:05 -0000

The following programs had to be recompiled due to a bug in the CON_SRCH routine that searches the engine.conf file, where a lone "[" character terminated a search in a group, such as [keys]. Programs that had to be recompiled were: PRE-SORT, TIPSORT, QX1007, QX1011, QX1012, QX1013, QX1017, and QX1039.

Locking

Ricardo Cantu — Thu, 10 Mar 2011 18:28:58 -0000

When canceling via tistat and browser is not up, there is a sh locking race from buzy thread. Kill buzz somehow or timeout on thread join

Buffering

Ricardo Cantu — Sat, 18 Dec 2010 23:23:12 -0000

When the screen is in the rendering phase (Facebook gif) keys are not buffered and display on the whited out screen. Also paste does not buffer at all.

Check Box gets Unchecked

Dean Lyles — Tue, 19 Oct 2010 18:57:30 -0000

I had a template with a Check Box on it, that was checked. I also had another field just preceding the check box, and I wanted to completely fill that field to the end. When I entered the last character in the preceding field, the cursor jumped to the Check Box, and THE CHECK MARK GET ERASED! But since I wasn't focusing on the Check Box, I didn't notice that the check mark got erased.

Deleting QXM32 TEMP using SYSM32.

Dean Lyles — Wed, 08 Sep 2010 19:39:44 -0000

QXM32 existed in both TEMP and T$MP. I went into SYSM32 and deleted QXM32 in TEMP. Afterwards, I could not do another action in the program. It complained that the LINE=03 data no longer existed.

$SEC - Fails to HOLD before update.

Dean Lyles — Fri, 03 Sep 2010 22:56:05 -0000

Engine fails to successfully HOLD a data line before update, when file $SEC is not in the database.

SIZEOF 2BYTE VS 4 BYTE

Tom Garrison — Fri, 03 Sep 2010 20:34:45 -0000

SIZEOF opcode only updates the last 2 bytes of the field that holds the length. That should always be big enough for any size field . BUT if the length field was a negative number to begin with (ie -30 hex (FFFFFFFC) when the SIZEOF is run it will only update the last 2 bytes. So if the field is 256 (hex 00000100 it will leave the length field as FFFF0100 which is -62380 or something like that. I have to zn the length before the SIZEOF if the length is a field used for other things like _BIN1.