Re: A couple of issues with psql variable substitution - Mailing list pgsql-hackers

I wrote:
> Andrew Dunstan <andrew@dunslane.net> writes:
>> We could just add -b unconditionally to the flex flags and then count 
>> the number of lines in lex.backup. If it's greater that 1 whine loudly, 
>> or even fail, otherwise remove lex.backup. Would that avoid locale 
>> dependencies?

> Hm, yeah, seems like that ought to work.

Done in HEAD, but only for the Makefile-based build mechanism.  Anybody
want to add the comparable logic to the MSVC scripts?
        regards, tom lane