Thread: failed regression tests

I'm trying to (finally) get my rather extensive patches (mostly
addressing int8 versions of date/time storage) applied but am now having
trouble with the regression tests.

I'm sure this has been on the list, but I'm not recalling the
explanation or workaround. My guess is that it is related to collation
troubles with the new locale-always-enabled feature. I've tended to
never enable this stuff in the past.

The first symptom is failures in the char regression test. An example
diff is

*** expected/char.out    Tue Jun  5 07:20:01 2001
--- results/char.out    Sun Apr 21 10:04:08 2002
***************
*** 63,74 ****    WHERE c.f1 < 'a';  five | f1  ------+----
-       | A       | 1       | 2       | 3       |  
! (5 rows)

So the 'A' row is left out of the result on my machine.

All other failures (there are 7 tests total which fail) are likely
similar in nature. I've tried a "make clean", a "make distclean", and
need a hint on what to try next. I'd *really* like to get these patches
applied, and am almost certain that they are not related to these
regression failures, but...

Early help would be appreciated; I've got time in the next couple of
hours to get this stuff finished!! :)
                   - Thomas

Thomas Lockhart writes:

> I'm sure this has been on the list, but I'm not recalling the
> explanation or workaround. My guess is that it is related to collation
> troubles with the new locale-always-enabled feature. I've tended to
> never enable this stuff in the past.
>
> The first symptom is failures in the char regression test. An example
> diff is

initdb --no-locale

I'm pondering ways to make the regression tests locale-aware, but it
hasn't happened yet.

-- 
Peter Eisentraut   peter_e@gmx.net

> > The first symptom is failures in the char regression test...
> initdb --no-locale

Ooh. Thanks, that fixes it. Where would I have found this in the docs? I
was looking in the wrong place (in configure/build) rather than at
initdb. Should we have something in the release notes? I see (now that I
look) that there is a one-liner, but istm that this may deserve a
paragraph in the "significant changes" category.
                - Thomas

> > > The first symptom is failures in the char regression test...
> > initdb --no-locale
> 
> Ooh. Thanks, that fixes it. Where would I have found this in the docs? I
> was looking in the wrong place (in configure/build) rather than at
> initdb. Should we have something in the release notes? I see (now that I
> look) that there is a one-liner, but istm that this may deserve a
> paragraph in the "significant changes" category.

Since once a user do initdb without knowing he is enabling the locale
support, the only way to recover from it is doing initdb again.  I
suggest something like showing a message in the initdb time
emphasizing he is enabling the local support.
--
Tatsuo Ishii