Re: No bison and NAMEDATALEN > 31: initdb failure?
От | Ian Barwick |
---|---|
Тема | Re: No bison and NAMEDATALEN > 31: initdb failure? |
Дата | |
Msg-id | 200207312334.40954.barwick@gmx.net обсуждение исходный текст |
Ответ на | Re: No bison and NAMEDATALEN > 31: initdb failure? (Tom Lane <tgl@sss.pgh.pa.us>) |
Список | pgsql-hackers |
On Tuesday 30 July 2002 00:23, Tom Lane wrote: > Ian Barwick <barwick@gmx.net> writes: > > - Does src/include/postgres_ext.h count as a parser definition file? > > No, it doesn't. Your experience sounds like you may have neglected to > do a full rebuild after altering NAMEDATALEN. (By default, we don't > compute object-file dependencies, so it's up to you to do "make clean" > after changing fundamental parameters.) Exactly what I thought. I actually deleted the source tree and unpacked at a different location, several times. It was only after installing initdb Bison that initdb suddenly worked... > It's unlikely that installing > Bison per se affected anything --- unless possibly you had corrupted > copies of gram.c etc. ... tried the same with a fresh source download on a Linux machine and got similar results. Clearly something odd is happening, but I assume it's a local problem. As it's not the most important issue right now, will see if I can make any sense (and report back should I find anything of interest). Thanks, Ian Barwick barwick@gmx.net PS The subject should have read "(...) NAMEDATELEN > 32 (...)", i.e. the default value.
В списке pgsql-hackers по дате отправления: