7 messages in com.mysql.lists.bugsRe: type clash on default action: <NO...| From | Sent On | Attachments |
|---|---|---|
| Martin MOKREJŠ | 02 Dec 2002 02:36 | |
| Sinisa Milivojevic | 02 Dec 2002 04:36 | |
| Paul Eggert | 08 Dec 2002 00:17 | |
| Sinisa Milivojevic | 09 Dec 2002 03:30 | |
| Lenz Grimmer | 09 Dec 2002 04:22 | |
| Michael Widenius | 10 Dec 2002 05:42 | |
| Paul Eggert | 10 Dec 2002 10:51 |
| Subject: | Re: type clash on default action: <NONE> != <>![]() |
|---|---|
| From: | Michael Widenius (mon...@mysql.com) |
| Date: | 12/10/2002 05:42:04 AM |
| List: | com.mysql.lists.bugs |
Hi!
"Paul" == Paul Eggert <egg...@twinsun.com> writes:
From: Sinisa Milivojevic <sin...@mysql.com> Date: Mon, 2 Dec 2002 14:37:12 +0200
Actually we use 1.35 only and have not changed our code to support 1.75 or similar.
Paul> If I were you I wouldn't bother with 1.75; I'd wait for the next Paul> official release of Bison.
Paul> Is there an easy way that I can browse the development history of Paul> sql/sql_yacc.yy to see where the <NONE> stuff was first introduced and Paul> why? It might help us Bison folks improve backward compatibility if Paul> we know why the grammar is the way it is.
If I remember correctly we started to use NONE very early in MySQL development to go around some warnings that we got from bison and the none part has been growing ever sense.
Now we routinely add new stuff to the NONE part when we get an error about a type clash from bison.
Regards, Monty
-- For technical support contracts, goto https://order.mysql.com/ __ ___ ___ ____ __ / |/ /_ __/ __/ __ \/ / Mr. Michael Widenius <mon...@mysql.com> / /|_/ / // /\ \/ /_/ / /__ MySQL AB, CTO /_/ /_/\_, /___/\___\_\___/ Helsinki, Finland <___/ www.mysql.com




