A BUGNOTE has been added to this bug. ====================================================================== http://www.dbmail.org/mantis/bug_view_advanced_page.php?bug_id=0000193 ====================================================================== Reported By: nik Assigned To: paul ====================================================================== Project: DBMail Bug ID: 193 Category: General Reproducibility: always Severity: major Priority: high Status: assigned ====================================================================== Date Submitted: 10-Apr-05 09:56 CEST Last Modified: 12-May-05 09:16 CEST ====================================================================== Summary: build fails with "imaputil.c:1792: error: 'timezone' redeclared as different kind of symbol" Description: When building the latest (7th April) snapshot of the 2.0 branch on FreeBSD 5.3 R, "gmake all" fails. (Currently the only method of installing a post 2.0.4 version on FreeBSD 5.3 appears to be an svn checkout from the 23rd of March.) ======================================================================
---------------------------------------------------------------------- paul - 10-Apr-05 11:20 CEST ---------------------------------------------------------------------- Ok. Another botched attempt at fixing http://www.dbmail.org/mantis/bug_view_advanced_page.php?bug_id=0000079. I'll change autoconf to check for tm.tm_gmtoff (provided in linux and bsd) and fallback to the current behaviour only if it's not there. ---------------------------------------------------------------------- paul - 10-Apr-05 21:52 CEST ---------------------------------------------------------------------- I assume this applies to *bsd, not just freebsd. The smell of ancient bsd/sysv controversy lingers. If anyone knows a cross platform method for getting at the local gmt-offset speak please. Until then, I'll start mucking with autoconf and see how that works out. I just downloaded some Xen stuff so I can bootstrap netbsd and freebsd virtual domains. This will help include more thorough independent cross platform prerelease testing. ---------------------------------------------------------------------- aaron - 11-May-05 22:55 CEST ---------------------------------------------------------------------- Ok, so apparently this might be a release blocker: Date: Wednesday, May 11 2005 01:21 pm From: Chris Reichow <[EMAIL PROTECTED]> add to contacts To: dbmail-dev@dbmail.org add to contacts, [EMAIL PROTECTED] add to contacts Reply-To: DBMAIL Developers Mailinglist <dbmail-dev@dbmail.org> add to contacts Subject: [Dbmail-dev] 2.0.5 Showstopper - FreeBSD does not compile cleanly Full Headers: Display Headers This is really in reply to Aaron's last email regarding 2.0.5, but I just joined the list 5 seconds ago, so I don't have a copy. =( Bug 193: http://www.dbmail.org/mantis/bug_view_advanced_page.php?bug_id=0000193 dbmail will not compile on FreeBSD without editing imaputil.c and reverting back to a tm_gmtoff type solution. I manually edited this to compile yesterday, but I am not well-versed in cross platform development and I have no idea if it would compile in Linux, etc. Here is my interim solution if anyone is interested. Not the prettiest, but it gets the job done. char *date_sql2imap(const char *sqldate) { time_t temp = time(0); struct tm *tms = localtime(&temp); long gmt_offset=0; struct tm tm_sql_date; char *last; /* defined by tzset */ //extern long timezone; //timezone=0; //tzset(); last = strptime(sqldate,"%Y-%m-%d %T", &tm_sql_date); if ( (last == NULL) || (*last != '\0') ) { trace(TRACE_DEBUG, "%s,%s, error parsing date [%s]", __FILE__, __func__, sqldate); strcpy(_imapdate, IMAP_STANDARD_DATE); return _imapdate; } gmt_offset = (-tms->tm_gmtoff)/3600; if (tm_sql_date.tm_isdst) gmt_offset++; snprintf(_imapdate,IMAP_INTERNALDATE_LEN,"%s %c%02ld00", sqldate, (gmt_offset >= 0 ? '+': '-'), gmt_offset); return _imapdate; } ---------------------------------------------------------------------- paul - 12-May-05 09:16 CEST ---------------------------------------------------------------------- Aaron, Using tm_gmtoff will effectively re-open bug http://www.dbmail.org/mantis/bug_view_advanced_page.php?bug_id=0000079. Other than that I don't have a problem with it. It must be possible to use autoconf to check for the presence of the tm_gmtoff struct member, and use that info. That would give us a longterm solution (hopefully). Bug History Date Modified Username Field Change ====================================================================== 10-Apr-05 09:56nik New Bug 10-Apr-05 11:16paul Assigned To => paul 10-Apr-05 11:16paul Status new => assigned 10-Apr-05 11:20paul Bugnote Added: 0000656 10-Apr-05 21:52paul Bugnote Added: 0000657 10-Apr-05 21:52paul duplicate_id 0 => 79 10-Apr-05 21:52paul ETA none => < 1 week 10-Apr-05 21:52paul Priority normal => high 10-Apr-05 21:52paul Projection none => minor fix 11-May-05 22:55aaron Bugnote Added: 0000700 12-May-05 09:16paul Bugnote Added: 0000701 ======================================================================
