From developer@opengroupware.org Sat Jan 13 15:18:01 2007 From: developer@opengroupware.org (Adam Tauno Williams) Date: Sat, 13 Jan 2007 10:18:01 -0500 Subject: [OGo-Developer] person::qsearch In-Reply-To: References: <20061122112449.5fywxqksmc4o8kkk@www.mormail.com> Message-ID: <1168701481.4857.29.camel@aleph.whitemice.org> > > However, when I run - > > results = [[self getCTX] runCommand:@"person::qsearch" > > @"qualifier", _query, nil]; > > - in my code I get - > > lookup of person::qsearchqualifier failed, missing info > Was a bug, the "person::qsearch" was not properly registered in the > bundle-info.plist. > Fixed in r1848. I'm running r1850 (ZideStore) and still getting this same message when I try to run "person::qsearch". From developer@opengroupware.org Sat Jan 13 19:01:20 2007 From: developer@opengroupware.org (=?ISO-8859-1?Q?St=E9phane_Corth=E9sy?=) Date: Sat, 13 Jan 2007 20:01:20 +0100 Subject: [OGo-Developer] Bug in EOQualifierParser.m Message-ID: <9C68C004-0B47-4440-B6EC-038C4FC58A34@sente.ch> Hi, I found a bug in latest sope-core/EOControl/EOQualifierParser.m, line =20= 960: if (!(_buf[pos + toklen] =3D=3D '\0' || isspace(_buf[pos + = toklen]))) /* not at the string end or folloed by a space */ continue; The following qualifier string, which is valid, is not parsed =20 correctly: the 'nil' token is wrongly interpreted as a string: (not (attribute =3D nil) and attribute.className =3D =20 'com.webobjects.foundation.NSTimestamp') The fix is easy, I guess: simply check for the right parenthesis too: if (!(_buf[pos + toklen] =3D=3D '\0' || isspace(_buf[pos + = toklen]) || =20 _buf[pos + toklen] =3D=3D ')')) /* not at the string end or followed by a space or a right =20 parenthesis */ continue; St=E9phane From developer@opengroupware.org Mon Jan 15 19:05:12 2007 From: developer@opengroupware.org (Helge Hess) Date: Mon, 15 Jan 2007 20:05:12 +0100 Subject: [OGo-Developer] Bug in EOQualifierParser.m In-Reply-To: <9C68C004-0B47-4440-B6EC-038C4FC58A34@sente.ch> References: <9C68C004-0B47-4440-B6EC-038C4FC58A34@sente.ch> Message-ID: <437C85E8-E138-4B87-9677-237739A861F0@opengroupware.org> Hi, thanks a lot, patch is applied! Greets, Helge On Jan 13, 2007, at 20:01, St=E9phane Corth=E9sy wrote: > Hi, > > I found a bug in latest sope-core/EOControl/EOQualifierParser.m, =20 > line 960: > > if (!(_buf[pos + toklen] =3D=3D '\0' || isspace(_buf[pos + = toklen]))) > /* not at the string end or folloed by a space */ > continue; > > The following qualifier string, which is valid, is not parsed =20 > correctly: the 'nil' token is wrongly interpreted as a string: > > (not (attribute =3D nil) and attribute.className =3D =20 > 'com.webobjects.foundation.NSTimestamp') > > > The fix is easy, I guess: simply check for the right parenthesis too: > > if (!(_buf[pos + toklen] =3D=3D '\0' || isspace(_buf[pos + = toklen]) || =20 > _buf[pos + toklen] =3D=3D ')')) > /* not at the string end or followed by a space or a right =20 > parenthesis */ > continue; > > St=E9phane > > -- > OpenGroupware.org Developer > developer@opengroupware.org > http://mail.opengroupware.org/mailman/listinfo/developer --=20 Helge Hess http://www.helgehess.eu/ From developer@opengroupware.org Thu Jan 18 03:18:14 2007 From: developer@opengroupware.org (Adam Tauno Williams) Date: Wed, 17 Jan 2007 22:18:14 -0500 Subject: [OGo-Developer] Patches In Bugzilla Message-ID: <1169090295.7518.0.camel@aleph.whitemice.org> There are allot of patches hanging around in Bugzilla. Some of them relate to quite interesting features. Is there anything we can do to scoot the process along of getting patches out of Bugzilla and into trunk? Or even of getting patches rejected. It is rather discouraging to get a patch into Bugzilla and see it sit, as well as labourious to have to maintain patched local copies for extensive periods of time. I understand that those with commit access/responsibility have numerous other duties. [ (*) indicates ones I've tried applying to my trunk checkout, and see at least to produce no issues with building. ] Mine: http://bugzilla.opengroupware.org/bugzilla/show_bug.cgi?id=1766 * http://bugzilla.opengroupware.org/bugzilla/show_bug.cgi?id=1792 * http://bugzilla.opengroupware.org/bugzilla/show_bug.cgi?id=1793 * Oliver's, at least related to the fancy-dock theme: http://bugzilla.opengroupware.org/bugzilla/show_bug.cgi?id=1732 * http://bugzilla.opengroupware.org/bugzilla/show_bug.cgi?id=1734 * http://bugzilla.opengroupware.org/bugzilla/show_bug.cgi?id=1735 * Also Oliver's: http://bugzilla.opengroupware.org/bugzilla/show_bug.cgi?id=1312 * (applies but you have to be in the WebUI/Templates/EnterprisesUI directory) Other: http://bugzilla.opengroupware.org/bugzilla/show_bug.cgi?id=1807 http://bugzilla.opengroupware.org/bugzilla/show_bug.cgi?id=1157 (?) http://bugzilla.opengroupware.org/bugzilla/show_bug.cgi?id=1167 http://bugzilla.opengroupware.org/bugzilla/show_bug.cgi?id=1311 http://bugzilla.opengroupware.org/bugzilla/show_bug.cgi?id=1627 http://bugzilla.opengroupware.org/bugzilla/show_bug.cgi?id=1800 http://bugzilla.opengroupware.org/bugzilla/show_bug.cgi?id=1795 From developer@opengroupware.org Thu Jan 18 08:20:13 2007 From: developer@opengroupware.org (Sebastian Reitenbach) Date: Thu, 18 Jan 2007 08:20:13 -0000 Subject: [OGo-Developer] Patches In Bugzilla Message-ID: <20070118082013.0FCC178C007@ogo.rapideye.de> Hi all, developer@opengroupware.org wrote: > There are allot of patches hanging around in Bugzilla. Some of them > relate to quite interesting features. Is there anything we can do to > scoot the process along of getting patches out of Bugzilla and into > trunk? Or even of getting patches rejected. > > It is rather discouraging to get a patch into Bugzilla and see it sit, > as well as labourious to have to maintain patched local copies for > extensive periods of time. I understand that those with commit > access/responsibility have numerous other duties. > > [ (*) indicates ones I've tried applying to my trunk checkout, and see > at least to produce no issues with building. ] > A great idea, I would really like to have mine, and other patches included, or at least rejected with a reason, so that I can work on it to have them included some time later. Sebastian From developer@opengroupware.org Thu Jan 18 10:29:34 2007 From: developer@opengroupware.org (Helge Hess) Date: Thu, 18 Jan 2007 11:29:34 +0100 Subject: [OGo-Developer] Patches In Bugzilla In-Reply-To: <1169090295.7518.0.camel@aleph.whitemice.org> References: <1169090295.7518.0.camel@aleph.whitemice.org> Message-ID: <07A55C34-2E2C-4131-BC7E-962180E5C175@opengroupware.org> On Jan 18, 2007, at 04:18, Adam Tauno Williams wrote: > Is there anything we can do to > scoot the process along of getting patches out of Bugzilla and into > trunk? Or even of getting patches rejected. OK. I usually do a full review of Bugzilla about once every 6 months, which obviously is not a lot and its no fun work at all ;-) Plus I'm still quite busy with a specific project, so this is scheduled to free slots ... My suggestion: lets give Adam full commit permissions. Adam, you would need to check the patches one-by-one, test them and possibly affected areas. And then approve/commit the patches (adding proper revision numbers of patches in Svn). I would then do another quick review prior doing a release. Adam, would you volunteer for that? Thanks, Helge PS: please remember that properly reviewing an Objective-C code related patch from a beginner is often as much work as implementing a given feature from scratch :-/ Thats not meant as a discouragement to post patches, but please remember that applying them is a LOT of work. -- Helge Hess http://www.helgehess.eu/ From developer@opengroupware.org Thu Jan 18 11:12:57 2007 From: developer@opengroupware.org (Adam Tauno Williams) Date: Thu, 18 Jan 2007 06:12:57 -0500 Subject: [OGo-Developer] warning: =?UTF-8?Q?=E2=80=98=5FOBJC=5FINSTANCE=5F4=E2=80=99?= defined but not used Message-ID: <1169118777.7518.7.camel@aleph.whitemice.org> When I compile SOPE I see a lot of messages like: warning: ‘_OBJC_INSTANCE_4’ defined but not used Are these innocuous or am I missing a piece somewhere? Everything seems to work. And it seems that compiling anything Obj-C can produce these messages. From developer@opengroupware.org Thu Jan 18 11:14:36 2007 From: developer@opengroupware.org (Helge Hess) Date: Thu, 18 Jan 2007 12:14:36 +0100 Subject: [OGo-Developer] =?WINDOWS-1252?Q?Re:_[OGo-Developer]_warning:_=91=5FOBJC=5FINSTA?= =?WINDOWS-1252?Q?NCE=5F4=92_defined_but_not_used?= In-Reply-To: <1169118777.7518.7.camel@aleph.whitemice.org> References: <1169118777.7518.7.camel@aleph.whitemice.org> Message-ID: <454BDB5D-365E-4C97-9F1B-F3996FEBB382@opengroupware.org> On Jan 18, 2007, at 12:12, Adam Tauno Williams wrote: > When I compile SOPE I see a lot of messages like: > warning: =91_OBJC_INSTANCE_4=92 defined but not used > > Are these innocuous or am I missing a piece somewhere? Innocuous (complicated word ;-), I think its a non-critical bug in =20 the compiler. Greets, Helge --=20 Helge Hess http://www.helgehess.eu/ From developer@opengroupware.org Thu Jan 18 14:18:16 2007 From: developer@opengroupware.org (Adam Tauno Williams) Date: Thu, 18 Jan 2007 09:18:16 -0500 Subject: [OGo-Developer] Patches In Bugzilla In-Reply-To: <07A55C34-2E2C-4131-BC7E-962180E5C175@opengroupware.org> References: <1169090295.7518.0.camel@aleph.whitemice.org> <07A55C34-2E2C-4131-BC7E-962180E5C175@opengroupware.org> Message-ID: <1169129896.4914.13.camel@aleph.whitemice.org> > > Is there anything we can do to > > scoot the process along of getting patches out of Bugzilla and into > > trunk? Or even of getting patches rejected. > OK. I usually do a full review of Bugzilla about once every 6 months, > which obviously is not a lot and its no fun work at all ;-) Plus I'm > still quite busy with a specific project, so this is scheduled to > free slots ... Understood. > My suggestion: lets give Adam full commit permissions. Yikes. > Adam, you > would need to check the patches one-by-one, test them and possibly > affected areas. And then approve/commit the patches (adding proper > revision numbers of patches in Svn). I would then do another quick > review prior doing a release. > Adam, would you volunteer for that? Yes; although ultimately I might not be any faster than you. :) As for "Adding proper revision numbers of patches in Svn". Is there a guide for this? As in some kind of [at least rough] procedure as to the canonical steps? > PS: please remember that properly reviewing an Objective-C code > related patch from a beginner is often as much work as implementing a > given feature from scratch :-/ Thats not meant as a discouragement to > post patches, but please remember that applying them is a LOT of work. Yep. From developer@opengroupware.org Mon Jan 22 13:32:07 2007 From: developer@opengroupware.org (Olivier Hallot) Date: Mon, 22 Jan 2007 11:32:07 -0200 Subject: [OGo-Developer] Patches In Bugzilla In-Reply-To: <1169129896.4914.13.camel@aleph.whitemice.org> References: <1169090295.7518.0.camel@aleph.whitemice.org> <07A55C34-2E2C-4131-BC7E-962180E5C175@opengroupware.org> <1169129896.4914.13.camel@aleph.whitemice.org> Message-ID: <45B4BCD7.9060302@scinergy.com.br> Adam Tauno Williams escreveu: >>> Is there anything we can do to >>> scoot the process along of getting patches out of Bugzilla and into >>> trunk? Or even of getting patches rejected. >> OK. I usually do a full review of Bugzilla about once every 6 months, >> which obviously is not a lot and its no fun work at all ;-) Plus I'm >> still quite busy with a specific project, so this is scheduled to >> free slots ... > > Understood. > >> My suggestion: lets give Adam full commit permissions. > > Yikes. Vote: +1 for Adam. Whatever you need to help you please let me know, I'll be glad to support you with my (somehow limited) knowledge. -- Olivier Hallot Scinergy Consulting Tel (021) 8822-8812 Rio de Janeiro, Brasil http://www.scinergy.com.br From developer@opengroupware.org Mon Jan 22 17:47:32 2007 From: developer@opengroupware.org (Sebastian Reitenbach) Date: Mon, 22 Jan 2007 17:47:32 -0000 Subject: [OGo-Developer] Patches In Bugzilla Message-ID: <20070122174732.B0CCE78C003@ogo.rapideye.de> Hi all, developer@opengroupware.org wrote: > > > Adam Tauno Williams escreveu: > >>> Is there anything we can do to > >>> scoot the process along of getting patches out of Bugzilla and into > >>> trunk? Or even of getting patches rejected. > >> OK. I usually do a full review of Bugzilla about once every 6 months, > >> which obviously is not a lot and its no fun work at all ;-) Plus I'm > >> still quite busy with a specific project, so this is scheduled to > >> free slots ... > > > > Understood. > > > >> My suggestion: lets give Adam full commit permissions. > > > > Yikes. > > Vote: +1 for Adam. > > Whatever you need to help you please let me know, I'll be glad to > support you with my (somehow limited) knowledge. > add me too here, with my, also (somehow limited ;) knowledge, I'd be glad to test patches, give comments on others patches, or try to change mine to make them fit to OGo. kind regards Sebastian From developer@opengroupware.org Tue Jan 23 11:45:49 2007 From: developer@opengroupware.org (Rafel Amer) Date: Tue, 23 Jan 2007 12:45:49 +0100 Subject: [OGo-Developer] userDefaults Message-ID: <45B5F56D.6040303@ruth.upc.edu> Hi! We are running OGo and different persons log in with the same username. Some times OGo crashes with the following message: class: LSDBObjectCommandException name: LSDBObjectCommandException reason: userdefaults::delete failed: Could not write User-Defaults I think the crash occurs when different OGo sessions tries to write to disk the same userDefaults file. The line 254 of the file SOPE-4.4/libFoundation/Foundation/common.m is filename = atomically ? [path stringByAppendingString:@"~"] : path; i.e., if the write is atomic a temporary file is used. The name of the temporary file is the name of the original file with "~" appended. Do you thing that replacing this line with filename = path; if(atomically) { NSProcessInfo *proc = [NSProcessInfo processInfo]; filename = [NSString stringWithFormat:@"%@%@.txt", [path stringByDeletingLastPathComponent], [proc globallyUniqueString]]; } will solve the problem? Thanks. R. Amer Thecnical University of Catalonia } -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From developer@opengroupware.org Tue Jan 23 12:03:25 2007 From: developer@opengroupware.org (Helge Hess) Date: Tue, 23 Jan 2007 13:03:25 +0100 Subject: [OGo-Developer] userDefaults In-Reply-To: <45B5F56D.6040303@ruth.upc.edu> References: <45B5F56D.6040303@ruth.upc.edu> Message-ID: <35D68FAF-FFCB-427E-8BF4-E203BBC0E0B0@opengroupware.org> On Jan 23, 2007, at 12:45, Rafel Amer wrote: > We are running OGo and different persons log in with the > same username. Some times OGo crashes with the following > message: > class: LSDBObjectCommandException > name: LSDBObjectCommandException > reason: userdefaults::delete failed: Could not write User- > Defaults Check your Unix permissions. OGo probably can't write the file. 'strace' is your friend .... Greets, Helge -- Helge Hess http://www.helgehess.eu/ From developer@opengroupware.org Tue Jan 23 13:50:48 2007 From: developer@opengroupware.org (Adam Tauno Williams) Date: Tue, 23 Jan 2007 08:50:48 -0500 Subject: [OGo-Developer] userDefaults In-Reply-To: <45B5F56D.6040303@ruth.upc.edu> References: <45B5F56D.6040303@ruth.upc.edu> Message-ID: <1169560248.4580.2.camel@aleph.whitemice.org> > We are running OGo and different persons log in with the > same username. Some times OGo crashes with the following > message: > class: LSDBObjectCommandException > name: LSDBObjectCommandException > reason: userdefaults::delete failed: Could not write User-Defaults > I think the crash occurs when different OGo sessions tries to write to disk > the same userDefaults file. > The line 254 of the file SOPE-4.4/libFoundation/Foundation/common.m is > filename = atomically ? [path stringByAppendingString:@"~"] : path; > i.e., if the write is atomic a temporary file is used. The name of the > temporary file is the name of the original file with "~" appended. I log into OGo's WebUI multiple time simultaneously all day / every day. It has never given me an error about saving defaults. > Do you thing that replacing this line with > filename = path; > if(atomically) { > NSProcessInfo *proc = [NSProcessInfo processInfo]; > filename = [NSString stringWithFormat:@"%@%@.txt", > [path stringByDeletingLastPathComponent], > [proc globallyUniqueString]]; > } > will solve the problem?