Re: Many regress tests failing due to latest psql changes

This seems to have been like that for a while. I put in a fflush() at the
appropriate spot, hope that helps. Man, I thought we had all of them
caught already... (It's time I start running the regress tests
religiously, ey?)

On 2000-01-13, Tom Lane mentioned:

> It looks like the latest psql has a buffer-flush-timing problem.
> Error messages are appearing out-of-sync with script echoing,
> for example:
> 
> *** expected/boolean.out    Sat Jan  8 19:31:26 2000
> --- results/boolean.out    Thu Jan 13 20:41:24 2000
> ***************
> *** 111,118 ****
>   -- This is now an invalid expression
>   -- For pre-v6.3 this evaluated to false - thomas 1997-10-23
>   INSERT INTO BOOLTBL2 (f1) 
> -    VALUES (bool 'XXX');  
>   ERROR:  Bad boolean external representation 'XXX'
>   -- BOOLTBL2 should be full of false's at this point 
>   SELECT '' AS f_4, BOOLTBL2.*;
>    f_4 | f1 
> --- 111,118 ----
>   -- This is now an invalid expression
>   -- For pre-v6.3 this evaluated to false - thomas 1997-10-23
>   INSERT INTO BOOLTBL2 (f1) 
>   ERROR:  Bad boolean external representation 'XXX'
> +    VALUES (bool 'XXX');  
>   -- BOOLTBL2 should be full of false's at this point 
>   SELECT '' AS f_4, BOOLTBL2.*;
>    f_4 | f1 
> 
> 
> This is making it difficult to look for actual backend bugs,
> so I respectfully request a fix ASAP.
> 
>             regards, tom lane
> 
> 

-- 
Peter Eisentraut                  Sernanders väg 10:115
peter_e@gmx.net                   75262 Uppsala
http://yi.org/peter-e/            Sweden