>There is never any reason to do a lock table before a check, so I
>don't see this is a problem; We can just defined that for OS/2 you
I found 'lock table t1 read; check table t1' in the show_check.test, so
I was looking for a way to pass the test without changing it.
>have to do LOCK TABLE WRITE in this case.
But this a good workaround.
All test are passed, except for the different status message (check
table t1) in alter_table and the 'show index' different values.
Where can I start debugging?
* member of TeamOS/2 - Italy