Hi Magnus
I will file the bug, and I agree with you that it is OK to fail in that
case.
On 2008-04-03 Thu 15:51 +0200，Magnus Svensson wrote：
> He Zhenxing skrev:
> > Hi Magnus
> >
> > Sorry for the delay! I missed the mail, found it today.
> >
> > On 2008-03-28 Fri 13:47 +0100，Magnus Svensson wrote：
> >> He Zhenxing skrev:
> >>> Hi Magnus
> >>>
> >>> Thank you for pointing out the spelling errors, will be fixed.
> >>>
> >>> Unfortunately its not possible to use
> >>> skip requires '$_engine_type' storage engine;
> >>>
> >>> 'skip' do not expand variables :(
> >> ok, please file a bug against mysqltest for that one. Or do you like to
> >> fix it? I'll show how.
> >>
> >
> > Good, I'd like to do it, and thank you!
>
> Ok, just file a bug, set me as reviewer and go.
>
> >
> >>> When there is no information_schema, it seems $_engines_table will be
> >>> empty.
> >> ok
> >>
> >> Maybe we should fix mysqltest to set the variable to something like
> >> "Query failed with error: <mysql_errno>" ? Would that be better than
> >> empty string...
> >
> > Actually I was wrong on this, if there is no infomation_schema table,
> > the test would fail with error:
> > 1049 Unknown database 'information_schema'
> >
>
> If that is what ends up in $_engines_tables variable, it's ok. Don't you
> think?
>
> Magnus
