Tuesday, 2009-04-14

../irclogs/#mantishelp.2009-04-14.log
--- scribe started ---00:00
CIA-11Mantisbt: thraxisp * r5019f259ff4d /bug_update.php: fix for #10035: Custom multi-selection list fields don't allow00:47
CIA-11Mantisbt: thraxisp master-1.1.x * re4ca98339277 /bug_update.php: fix for #10035: Custom multi-selection list fields don't allow00:50
CIA-11Mantisbt: thraxisp master-1.1.x * r45934909ad91 / (config_defaults_inc.php core/form_api.php): Merge branch 'master-1.1.x' of git@mantisbt.org:mantisbt into master-1.1.x00:50
husker_inhello, my mantis version is 1.1.6. I use LDAP for authentication.05:40
husker_inFor OU=foo, I want to limit access to only specific users.05:41
husker_inCan someone point me a way accomplish this?05:41
mantisbotNew bug: Bug 10327 - harpreet_singh - open - new06:42
mantisbotNew bug: Installation of Mantis - http://www.mantisbt.org/bugs/view.php?id=1032706:42
djpok, so I report an issue in mantis and if I dont fill out a required field I get an error07:18
djpwhen I click back I have lost all the information I had entered07:18
djpIs there a fix for this?07:18
djpthe version we are using is 1.1.207:22
djpok I found it is fixed in 1.1.307:47
paulr_lo10:17
paulr_hmm11:48
paulr_nuclear_eclipse: you done anything yet on utf?12:10
* paulr_ is playing12:10
CIA-11Mantisbt: paul * r3fd3eb8b2c6c / (5 files in 3 dirs): Fix: #0006833: Incorrect handling of some UTF-8 strings12:29
CIA-11Mantisbt: paul * rc18173d939a1 /manage_tags_page.php: HTML Validation12:30
CIA-11Mantisbt: paul * rc7569b1d1389 /lang/strings_catalan_8859-15.txt: Fix test_langs.php failures in catalan_8859.12:30
nuclear_eclipsepaulr_: I try to stay as far away from Unicode/UTF-8 as I possibly can14:06
paulr_hows your cw?14:07
nuclear_eclipsecw?14:08
paulr_work course14:08
nuclear_eclipseah14:08
nuclear_eclipsewell, I think I'm going to be pepetually overloaded for the next month or so until I graduate... :(14:09
nuclear_eclipseI just got back from class, and have yet another project to work on....14:09
paulr_hehe14:09
paulr_CREATE DATABASE dbname CHARACTER SET utf8 COLLATE utf8_general_ci; is how to create a mysql utf8 db ?14:10
nuclear_eclipseI have no clue14:10
nuclear_eclipseI always use 'CREATE DATABASE name' or phpMyAdmin... :P14:10
paulr_seems that doesn't support inserting utf8 data14:10
paulr_mantis live db strips utf 8 ;/14:13
paulr_http://www.mantisbt.org/bugs/view.php?id=921814:14
* paulr_ ponders14:14
paulr_should that be resolved ? ;/14:15
dhx_mhi14:15
paulr_hi14:15
dhx_mnuclear_eclipse: ah, UTF-8 is nice :p14:15
paulr_dhx_m: well volunteered14:15
dhx_mwhat is the problem?14:15
* paulr_ doesn't do utf8 :)14:16
dhx_mone thing I hate is software that doesn't play nicely with UTF814:16
paulr_and i'm trying to fix utf8 stuff14:16
* paulr_ makes note.... daryn dates... nuclear_eclipse bugobject feedback... dhx_m every utf8 bug14:16
dhx_mso basically I should put chinese characters all over my Mantis install14:16
dhx_mto see what breaks?14:17
paulr_heh, nah14:17
* paulr_ is currently doing that ;p14:17
dhx_mI was horrified to see: http://git.mantisbt.org/?p=mantisbt.git;a=commitdiff;h=f6d71ff36607383a7123e6c2f9a5b1e9f64349ab14:17
dhx_mhardcoded integer representation of projections, eta, resolution, etc! :)14:17
dhx_mI have a patch coming out to fix that14:17
paulr_it's always been like that14:18
dhx_mbug_submit_resolution, etc are needed14:18
dhx_mI know14:18
dhx_mseems very ugly to me14:18
nuclear_eclipsedhx_m: perhaps you would like to assist with bug 10303?14:18
mantisbotBug 10303 - jreese - open - assigned14:18
mantisbotCreate and use string_api wrappers for multibyte string functions - http://www.mantisbt.org/bugs/view.php?id=1030314:18
paulr_*NOT* when reporting bugs14:18
paulr_ Incorrect string value: '\xEF\xBB\xB0' for column 'description' at row 1 for the query: INSERT INTO mantis_bug_text_table14:18
paulr_( description, steps_to_reproduce, additional_information )14:18
paulr_VALUES14:18
paulr_( '?','','').14:18
* paulr_ sighs14:18
dhx_mI was actually thinking of starting a new "Work Log" plugin14:18
dhx_mexcept it won't be called "Work Log"... maybe more like "UserMetrics" or "UserStats"14:19
paulr_mysql> show variables like "character_set_database";14:19
paulr_+------------------------+-------+14:19
paulr_| Variable_name          | Value |14:19
paulr_+------------------------+-------+14:19
paulr_| character_set_database | utf8  |14:19
paulr_+------------------------+-------+14:19
paulr_1 row in set (0.01 sec)14:19
paulr_mysql>14:19
paulr_mysql> show variables like "collation_database";14:19
paulr_+--------------------+-----------------+14:19
paulr_| Variable_name      | Value           |14:19
paulr_+--------------------+-----------------+14:19
paulr_| collation_database | utf8_unicode_ci |14:19
paulr_+--------------------+-----------------+14:19
paulr_1 row in set (0.00 sec)14:19
paulr_that should be fine no? ;/14:19
dhx_mthe aim is to have other plugins hook into the UserMetrics plugin to show up information per user14:19
dhx_ma list of commits, number of commits, list of bugs and the time spent on them, etc14:20
dhx_mdoes that sound like the correct approach to use?14:20
nuclear_eclipsedhx_m: seems reasonable I guess14:21
dhx_mthe idea is that the UserMetrics plugin is extremely generic (it'd be quite simple)14:21
nuclear_eclipseI gotta run though, cheers all14:21
dhx_mand other plugins hook into it to provide the stats14:21
dhx_mUserMetrics could of course provide basic built-in stats from Mantis itself14:21
dhx_mok14:21
dhx_mcya nuclear_eclipse14:21
paulr_hmm14:22
paulr_utf814:22
dhx_mstrlen() is the same for UTF8 and ASCII14:22
paulr_nope14:22
dhx_mUTF8 reserves 0x00 for string termination14:22
dhx_mso it won't appear mid-string14:22
dhx_mat least this is how it works in C14:23
paulr_php fails then :)14:23
dhx_mstring_lower is actually very complex14:23
dhx_mbecause some languages have very special rules on case conversion14:23
dhx_mI assume PHP has this functionality built in... if not... don't even bother trying14:24
dhx_mhttp://au.php.net/manual/en/function.mb-strtolower.php14:24
dhx_mseems to handle that OK from what I am reading14:24
giallupaul_, live mantis DB has still lots of fields in latin mode last time I looked14:25
dhx_mthose mb_ functions are also very slow, because they are very high complexity14:25
paulr_dhx_m: some of projects like joomla/wikipedia etc have php implementations of the mb_ functions15:00
paulr_tending to use big arrays15:00
dhx_mhmmm so they 'roll their own'?15:01
* paulr_ nods15:01
paulr_for example, see http://dev.splitbrain.org/view/darcs/dokuwiki/inc/utf8.php15:01
paulr_lo15:10
darynhello15:11
paulr_daryn: i've made a new version of the date upgrade stuff15:11
dhx_mlooks quite ugly15:11
dhx_mloading all those arrays for each page load... ouch15:11
paulr_heh15:11
darynok.15:12
paulr_dhx_m: it's simpel really :)15:12
daryni'm quite busy today but i'll try to give it a test at some point15:12
paulr_php6 has utf8 support, php4/5 dont15:12
paulr_php4/5 give utf8 support via the mb_ extension15:12
paulr_however, mb_* isn't a default extension i.e. needs to be enabled15:12
paulr_so, the question is15:13
paulr_do we require mb_*15:13
paulr_do we add a wrapper around ~5 functions to use mb_* if available (NOTE: that can be done by using in php.nii via mb_'s function overload stuff15:13
paulr_do we borrow from somewhere a non_mb* version of stuff15:14
paulr_so that utf stuff works for everyone15:14
paulr_obvious issues with each method :)15:14
paulr_how much of what I wrote did you get ;/15:17
dhx_mI got a link to utf8 (to which I had a one line reply)15:18
dhx_mthe next I heard was daryn: i've made a new version of the date upgrade stuff15:18
dhx_mand that is it15:18
paulr_[16:12:21] <paulr_> dhx_m: it's simpel really :)15:28
paulr_[16:12:30] <paulr_> php6 has utf8 support, php4/5 dont15:28
paulr_[16:12:30] <daryn> i'm quite busy today but i'll try to give it a test at some point15:28
paulr_[16:12:39] <paulr_> php4/5 give utf8 support via the mb_ extension15:28
paulr_[16:12:59] <paulr_> however, mb_* isn't a default extension i.e. needs to be enabled15:28
paulr_[16:13:13] <paulr_> so, the question is15:28
paulr_[16:13:18] <paulr_> do we require mb_*15:28
paulr_[16:13:41] <paulr_> do we add a wrapper around ~5 functions to use mb_* if available (NOTE: that can be done by using in php.nii via mb_'s function overload stuff15:28
paulr_[16:14:00] <paulr_> do we borrow from somewhere a non_mb* version of stuff15:28
paulr_[16:14:08] <paulr_> so that utf stuff works for everyone15:28
paulr_[16:14:22] <paulr_> obvious issues with each method :)15:28
dhx_mah, I see15:28
dhx_mwell PHP4 is deprecated so I don't see why it needs to be supported by Mantis anymore15:29
dhx_mno one else supports it (including dinosaur distros like Redhat)15:29
dhx_mPHP5 would be a problem however15:29
dhx_mpaulr_: maybe that wrapper idea is best?15:30
dhx_mpaulr_: if using PHP < 6, include the MB string functions that we've defined (is there a library for this)15:31
dhx_mwith the exception of using PHP5 and the mb_ functions are available use them15:31
dhx_menglish no good heh15:31
dhx_mit sounds like PHP 6 is turning into the next Perl 6 :p15:36
onetowerPLhi16:01
onetowerPLhow secure mantisbt after first instalation ?>16:02
onetowerPLver 1.1.616:02
onetowerPLi've delete admin directory16:02
CIA-11Mantisbt: paul * r31955f9282cd /lang/strings_english.txt: Fix: #0010306: "E-mail on note Added" => "E-mail on Note Added"16:33
CIA-11Mantisbt: paul * r5b26043b5f20 /bugnote_add.php: Fix: #0010294: Bugnote add reads bug_readonly_status_threshold from current project not the bug's one16:51
gialluonetowerPL, it depends on your definition of "secure". but if you just removed the admin directory you still have an "administrator" user with a default password18:08
onetowerPLso what kind of providence should i take ?18:08
gialluchange its password18:09
gialluor create a new one18:09
gialluand then disable/remove it18:09
gialluit == the default one18:09
onetowerPLok18:09
onetowerPLany additional hints ?18:09
onetowerPL:)18:09
gialluguess not. right now we have no nkown security issues18:10
onetowerPLgiallu thx for your help18:18
paulr_lo gaiallu18:29
giallupaul_, hi18:32
giallupaul_, so what are you doing about dates?18:33
giallupaul_, any public git branch I can look at?18:34
paulr_i've got two branches in mantisforge forks18:37
paulr_so yea18:37
paulr_one was bugobjects18:37
paulr_and one was dates18:37
giallubugobjects == class Mantis_Bug ?18:40
mantisbotNew bug: Bug 10328 - kec161 - open - new18:41
mantisbotNew bug: Added ability to append the handler to the history of specified custom fields - http://www.mantisbt.org/bugs/view.php?id=1032818:41
gialludinner time... will come back later18:44
aptituz_hm18:51
* aptituz_ wonders if my mail today reached the list18:51
paulr_hm19:16
paulr_aptituz: one mail did19:21
paulr_about phpmailer19:21
aptituzok19:22
gialluaptituz, you around?19:42
aptituzyea19:42
gialluhi19:42
gialluso about phpmailer, which version are you on debian?19:43
* giallu considering to split it from mantis on Fedora too19:43
gialluaptituz, see, someone replied you on list...19:43
aptituzpsc@lisa ~ % rmadison -sunstable  libphp-phpmailer19:43
aptituzlibphp-phpmailer |      2.1-1 |      unstable | source, all19:43
gialluit seems you're breaking ubuntu, please go fix it :P19:44
aptituzI don't see a reason to do the work of others :P19:44
gialluaptituz, is it a problem to push the newest phpmailer release there?19:44
aptituzno idea19:45
aptituzbut I don't see why this should be needed?19:45
aptituzwait, reading the mail once I've found it19:46
paulr_phpmailer is broken19:46
paulr_:)19:46
aptituzgiallu: problem isn't that libphp-phpmailer is too old, its that mantis misuses phpmailer19:46
gialluaptituz, yes19:47
gialluI know19:47
giallubut the point is19:47
aptituzits fixed in trunk, but last time I've checked it hasn't been fixed in 1.1.x19:47
giallupaul_ recently upgraded it to the latest releases in trunk19:47
gialluI assume we are not carrying patches now on that code19:47
gialluso we could use the same lib on 1.119:47
paulr_atm no19:47
gialluand fix this and some other issues19:48
paulr_but the newest release of phpmailer is still broken :)19:48
aptituzhmm.19:48
giallupaul_, please define "broken"19:48
paulr_doesn't work at the specification :)19:48
* paulr_ looks for bug report19:48
giallupaul_, you mena the /r/n stuff?19:48
paulr_http://www.mantisbt.org/bugs/view.php?id=1026919:49
aptituzgiallu: is this new upstream version _needed_ for the 1.1.x fix?19:49
aptituzor other said: would it be possible to patch it in lenny without a new upstream version and in sid with the new libphp phpmailer version?19:50
gialluaptituz, probably not19:50
giallubut upgrading 1.1 to an intermediate release looks not wise19:50
aptituzhm, otoh it doesn't matter19:51
paulr_tbh19:51
aptituzI just realized that the bug doesn't affect lenny ;)19:51
paulr_it's probably easier to fix mantis in this case or well19:51
paulr_heh19:51
paulr_problem you've got here is19:51
paulr_old mantis + old phpmailer19:51
paulr_= working combination19:51
aptituzyep19:51
paulr_dev mantis + new phpmailer = working combination19:51
paulr_old mantis + new phpmailer = broken combination19:52
aptituzwell I'm not the maintainer of libphp-phpmailer, so I'd have to ask him19:52
giallupaulr_, correct but incomplete19:52
aptituzwrt to ubuntu: they need to fix it themselves eventually19:52
paulr_giallu: oh?19:52
gialluaptituz, nah, they will wait for your fix...19:52
aptituzhowever I'd be able to push out a new 1.1.7 upload within a day once you release it19:52
giallupaul_, there is also the fact we want to fix some issue19:52
aptituzbut I don't see much sense in fixing 1.1.6 before19:53
gialluand we cant' with the old phpmailer19:53
aptituzunless you expect the release to take some weeks19:53
paulr_tbh19:53
paulr_the problem here is19:53
gialluaptituz, ah, I thought you were19:54
giallumaintainer19:54
paulr_we've not done a 'new' mantis release since Dec 200719:54
paulr_and we've done weekly commits to mantis since that date19:54
aptituzgiallu: for mantis, not libphp-phpmailer19:54
gialluaptituz, so, please see if you are going to hit problems if we decide to upgrade phpmailer above 2.119:55
aptituz(I wanted to ask the maintainer weither I can become the co-maintainer, but I haven't done yet)19:55
paulr_on top of that, I tend to look at whether an issue is resolved in 'git code'19:55
giallupaul_, which would be good _if_ we were doing regular releases...19:55
aptituzto summarize19:55
aptituz1) you plan to eventually move 1.1.7 to the new phpmailer? 2) Whats your intended timeline for 1.1.7?19:56
paulr_2-> john was talking this weekend iirc19:56
aptituzmeans eventually to late for ubuntu19:56
gialluaptituz, yeah19:57
giallutoo late19:57
giallubut I hardly see that as our problem19:57
aptituzneither mine19:57
gialluso let's go on.19:58
paulr_our problem is ubuntu should be running a 1.2.0 :)19:58
aptituzbut I try to find a solution for them, too.. for whatever its worth19:58
giallumy summary is:19:58
giallu1. John is going to make a release, eventually as soon as this w.e.19:58
giallu2. we have some bugs wrt to email handling, fixed in trunk with a newer phpmailer _and_ mantis code changes19:59
giallu3. we can choose if/when backprot this stuff to 1.119:59
gialluso basically I am investigating if we want to do it (large changes like this are not usual in the stable release )20:00
aptituzwell I'm perfectly fine with upgrading to a newer phpmailer version20:00
aptituzmight take some time till it reaches unstable in Debian, but well20:00
aptituzafter all its a development branch of Debian.20:00
aptituzthe thing is that I can't push newer upstream versions in lenny anymore20:01
aptituzso after all I'd needed to backport fixes for security issues anyway20:01
gialluaptituz, yeah. the time it takes to propagate in distros is mostly out of our control20:01
gialluaptituz, and that has 1.1.6?20:02
aptituzyes20:02
aptituzwhat I mean is: I don't care if 1.1.7 in unstable diverges from 1.1.6 in lenny20:02
giallubut you can push only security fixes?20:02
aptituzwell and fixes for important bugs20:02
aptituzreally important bugs20:03
gialluok, I hope _you_ decide what's important...20:03
giallubut I understand the issue20:03
aptituzno, after all the release team needs to ACK each change, because they are responsible for the release20:03
gialluanyway, the current mantis policy is mostly compatible with your policy,i.e. you should not see big changes20:04
gialluin this case, if phpmailer is at 2.120:04
gialluwe could probably try that first20:05
aptituzno giallu, you get me wrong :)20:05
* giallu feels a sense of dejavu20:05
aptituzin my eyes there _is no_ issue20:05
paulr_DUE DATES20:06
paulr_ARGHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHHH20:06
aptituzlenny most likely won't get 1.1.7 and sid can be broken in every imaginable way20:06
paulr_giallu: did you look at my branches?20:07
aptituzit currently has 2.1, yes, but that doesn't matter because it should not be a too big problem to update20:07
gialluaptituz, uhm. but if lenny has phpmailer 2.1 and mantis 1.1.6, isn't that a broken pair (accessing private property?)20:08
giallupaul_, not yet20:08
aptituzgiallu: lenny doesn't have 2.120:08
gialluah20:08
aptituzpsc@lisa ~ % rmadison -sstable libphp-phpmailer20:09
aptituzlibphp-phpmailer |     1.73-6 |        stable | source, all20:09
gialluok, I see you have _no issue_20:09
gialluapart if you want to fix the double message-id bug...20:09
aptituzwell its annoying but not important enough that the release team would ack that20:10
aptituzso even if I wanted20:10
gialluok great20:11
MikeWHmmmmmmmmmmmmmmmmmmmmmmm20:33
MikeWSo I fail.20:33
MikeWI never had time to seperate out all my custom mantis code to patches I could make available, the codebase has changed a lot anyway, and when I left the team well over a year ago they moved to some other bugtracking system so the code is unmaintained. Yeah, I fail at tail'ing the project and being useful anymore.20:36
MikeWp.s. even though the docs (used to suck - I dunno if anything changed in the last 16 months) - mantis was still one of the most awesome totally hackable and extendable bugtrackers I've used - thanks for the ride guys. I think I can /part happy now.20:37
giallusad...20:45
* paulr_ can't recall mikey speaking much though ;/21:41

Generated by irclog2html.py