From owner-freebsd-questions Wed Dec 18 06:45:43 1996 Return-Path: Received: (from root@localhost) by freefall.freebsd.org (8.8.4/8.8.4) id GAA09527 for questions-outgoing; Wed, 18 Dec 1996 06:45:43 -0800 (PST) Received: from nora.pcug.co.uk (Nora.PCUG.CO.UK [192.68.174.71]) by freefall.freebsd.org (8.8.4/8.8.4) with SMTP id GAA09522 for ; Wed, 18 Dec 1996 06:45:40 -0800 (PST) Received: from imdb.demon.co.uk by nora.pcug.co.uk id aa27695; 18 Dec 96 14:42 GMT Message-Id: <199612181442.OAA04492> Subject: Update DB To: freebsd-questions@freebsd.org Date: Wed, 18 Dec 1996 14:42:32 +0000 (GMT) From: Rob Hartill Organization: Internet Movie Database Ltd. X-pgp-public-key: http://us.imdb.com/pgp.html X-Mailer: ELM [version 2.4 PL24 ME8a] Content-Type: text Sender: owner-questions@freebsd.org X-Loop: FreeBSD.org Precedence: bulk Are there any plans to update the DB library ? perl 5.004 is due very soon and is likely to flag a "make test" failures because of the old version of DB that comes with 2.1.6.1 (and earlier). FreeBSD uses version 1.74, and the most "recent" is 1.85. I tried dropping 1.85 of db into the src tree and was able to compile it ok (after making mpool.h look at the new file in src instead of /usr/include) but a make install failed. It was 5am and I was too tired to think to save the error message before I reinstalled the old db and went to bed, however I think it failed to find "btree.0" The perl developers have already had a number of bug reports regarding this. What's the solution ? The perl people say: > bval is ignored. I've already had a report of this problem with > FreeBSD. > > So assuming you don't muck around with bval, you shouldn't have any > problems with 5.003_11. > > Personally, I would upgrade to Berkeley DB 1.85 -- Rob Hartill. Internet Movie Database Ltd. http://www.imdb.com/ -=-=-=-==-==-=-=-=-=-=-=-=-=-=-=-=-=-= 1.84 -> 1.85 recno: #ifdef out use of mmap, it's not portable enough. 1.83 -> 1.84 Thu Aug 18 15:46:07 EDT 1994 recno: Rework fixed-length records so that closing and reopening the file now works. Pad short records on input. Never do signed comparison in recno input reading functions. 1.82 -> 1.83 Tue Jul 26 15:33:44 EDT 1994 btree: Rework cursor deletion code yet again; bugs with deleting empty pages that only contained the cursor record. 1.81 -> 1.82 Sat Jul 16 11:01:50 EDT 1994 btree: Fix bugs introduced by new cursor/deletion code. Replace return kbuf/dbuf with real DBT's. 1.80 -> 1.81 btree: Fix bugs introduced by new cursor/deletion code. all: Add #defines for Purify. 1.79 -> 1.80 Wed Jul 13 22:41:54 EDT 1994 btree Change deletion to coalesce empty pages. This is a major change, cursors and duplicate pages all had to be reworked. Return to a fixed stack. recno: Affected by cursor changes. New cursor structures should permit multiple cursors in the future. 1.78 -> 1.79 Mon Jun 20 17:36:47 EDT 1994 all: Minor cleanups of 1.78 for porting reasons; only major change was inlining check of NULL pointer so that __fix_realloc goes away. 1.77 -> 1.78 Thu Jun 16 19:06:43 EDT 1994 all: Move "standard" size typedef's into db.h. 1.76 -> 1.77 Thu Jun 16 16:48:38 EDT 1994 hash: Delete __init_ routine, has special meaning to OSF 2.0. 1.74 -> 1.76 all: Finish up the port to the Alpha.