Hi everyone!
Sorry for posting in this group but there's no group for v6!
I need to do some actions on about 50 files.
But always I get error 88(INCOMPATIBLE_MODE_ERROR) on first few ones.
Then I changed the code to try as many times as needed to get a no
error result.
Now I always get 88 on the 1st file and operations on other files are
done with no error.
I know this error occures when a file is opened in 2 different modes
but in my code
each file is opened , processed , closed and then the next file.And
all in normal mode.
I'm working with Microkernel 6.15 on a single WinXP machine,not a
client server system.
And obviously I can't upgrade.

What should I do?
Can anyone help me.

Thanks.

<mohsen_fakhari (AT) yahoo (DOT) com> wrote

The only time I ever get this particular error is if I open a file in
read-only mode and attempt to modify the records (update/delete). Post a
snippet of your code, particularly where the error occurs.

If I had to guess, I'd say that you're running into one of the
following:
1) File is flagged ReadOnly. Flag the file ReadWrite.
2) A conflict with an Anti-Virus application is preventing opening the
file properly. Disable AV & see if the program works.
3) You are not initializing the KeyNumber parameter on the FileOpen
call. Be sure this is set to 0.


As for your last statement -- what is obvious about this? Pervasive
PSQL v9.5 will run most Btrieve 6.x applications with no modifications
whatsoever.
Goldstar Software Inc.
Pervasive-based Products, Training & Services
Bill Bach
BillBach (AT) goldstarsoftware (DOT) com
http://www.goldstarsoftware.com
*** Chicago: Pervasive Service & Support Class - 07/2007 ***


mohsen_fakhari (AT) yahoo (DOT) com wrote:

Hi
Thank you for your answers.In fact I was disappointed.
well,I disabled AV completely, nothing changed.
If the files were readOnly,I think I couldn't get any results but I
get NO_ERROR after some attempts.
However I checked them.
And this is the code.It's not readolny mode.

memset(dataBuf , 0 , sizeof(dataBuf) );
memset( keyBuf, 0, sizeof( keyBuf ) );

strcpy(dataBuf , "MA48");
strcpy( (BTI_CHAR *)keyBuf, fileName );

keyNum = 0;
dataLen = 5;
/* open the DAT file */
i = 0;
status = BTRV(B_OPEN, posBlock, dataBuf, &dataLen, keyBuf, keyNum );
while(status == 88){
fp = fopen("unLog.txt" , "a");
fprintf(fp , "%d," , i);
fclose(fp);
Sleep(SLEEP_TIME);//Some time this sleep reduces the wasted time in
the loop
status = BTRV(B_OPEN, posBlock, dataBuf, &dataLen, keyBuf, keyNum );
i += 1;
}

And if it's important this code is in a C function that is called from
a Java method
using JNI.

<mohsen_fakhari (AT) yahoo (DOT) com> wrote

I don't program in C, but I can follow the code (I think) but I'm curious as
to why you are only checking for status 88. If your open returns any other
non-zero status, you'll continue as if the file is opened properly, when it
is not. I suggest you place some code to handle a status other than 88 or
zero.

You don't have your outer loop shown there. Keep in mind that there
will be a short period of time when a database file is closed that the
engine (and subsequently, the OS) is still busy writing out the data.
If you are truely running in a small open/access/close loop, then you
may just be seeing the wait for the OS cache to get flushed.

Not sure why else you'd get this, though I suspect that there is some
OS limitation you are seeing. With the 6.15 workstation engine, you
can probably run SysInternals/Microsoft "FileMon" tool to gain a bit
more insight.
Goldstar Software Inc.
Pervasive-based Products, Training & Services
Bill Bach
BillBach (AT) goldstarsoftware (DOT) com
http://www.goldstarsoftware.com
*** Chicago: Pervasive Service & Support Class - 07/2007 ***


mohsen_fakhari (AT) yahoo (DOT) com wrote:

I'm checking just 88 because just this error is solved after some
tries.
Other errors should be handled basically.

I can't upgrade since I don't have access to other versions.
This is what I'm given.