Monday, 2009-04-27

../irclogs/#mantishelp.2009-04-27.log
--- scribe started ---00:00
[KK]Kirillhi07:10
mantisbotNew bug: Bug 10410 - paontis - open - new09:09
mantisbotNew bug: Upload and edit MS Office documents, they are corrupted - http://www.mantisbt.org/bugs/view.php?id=1041009:09
[KK]Kirillgiallu: you here?09:26
gialluya09:26
[KK]Kirillhi.. How I can get priority value of given issue?09:27
gialluusing soap?09:27
gialluor what?09:27
[KK]Kirillnot. I want override function09:28
[KK]Kirillwith write of priority status09:28
mantisbotNew bug: Bug 10411 - paontis - open - new09:40
mantisbotNew bug: Changes to project_view_state and view_state to create only private projects - http://www.mantisbt.org/bugs/view.php?id=1041109:40
[KK]Kirillgiallu: I try override function custom_function_override_roadmap_print_issue( $p_issue_id, $p_issue_level = 0 )09:51
[KK]Kirillbut I don't see any results.09:52
gialluah ok, now I get it09:52
gialluwell, to get the priority you use the bug_get_field function09:53
mantisbotNew bug: Bug 10412 - karthisudar - open - new09:55
mantisbotNew bug: defect - http://www.mantisbt.org/bugs/view.php?id=1041209:55
mantisbotNew bug: Bug 10413 - karthisudar - open - new09:55
mantisbotNew bug: testing for LR - http://www.mantisbt.org/bugs/view.php?id=1041309:55
mantisbotNew bug: Bug 10414 - karthisudar - open - new09:55
mantisbotNew bug: defect - http://www.mantisbt.org/bugs/view.php?id=1041409:55
giallugrrrr09:55
mantisbotNew bug: Bug 10415 - karthisudar - open - new10:00
mantisbotNew bug: testing defects - http://www.mantisbt.org/bugs/view.php?id=1041510:00
mantisbotNew bug: Bug 10416 - karthisudar - open - new10:00
mantisbotNew bug: testing defects - http://www.mantisbt.org/bugs/view.php?id=1041610:00
mantisbotNew bug: Bug 10417 - karthisudar - open - new10:15
mantisbotNew bug: testing defects - http://www.mantisbt.org/bugs/view.php?id=1041710:15
mantisbotNew bug: Bug 10418 - karthisudar - open - new10:45
mantisbotNew bug: testing defects - http://www.mantisbt.org/bugs/view.php?id=1041810:45
[KK]Kirillok. I did this10:47
[KK]Kirillgiallu: thanks10:47
mantisbotNew bug: Bug 10419 - jonathh - open - new11:05
mantisbotNew bug: Versions are not sorted - http://www.mantisbt.org/bugs/view.php?id=1041911:05
thraxispgiallu: can't we just disable this guy's account (karthisudar)11:07
gialluthraxisp, yeah11:09
gialluI think I'm going to hack the server and gain adminsitrator status...11:10
[KK]KirillWho know git well?11:41
[KK]KirillJohn, wake up11:55
[KK]Kirillgiallu: close this bug 1038412:08
mantisbotBug 10384 - ruslan - open - new12:08
mantisbotRoadmap can/should show priority  symbol for issues. - http://www.mantisbt.org/bugs/view.php?id=1038412:08
mantisbotNew bug: Bug 10420 - kesterhabermann - open - new12:10
mantisbotNew bug: Custom fields with newline break CSV export - http://www.mantisbt.org/bugs/view.php?id=1042012:10
mantisbotNew bug: Bug 10421 - kesterhabermann - open - new12:15
mantisbotNew bug: Default filename corrupted if project name contains special chars - http://www.mantisbt.org/bugs/view.php?id=1042112:15
nuclear_eclipsemorning all12:28
[KK]KirillHi12:28
[KK]KirillJohn, nice to see you12:29
nuclear_eclipsebtw, the green labels in gitweb mark local branches, pink marks remote branches, and yellow marks tags12:34
[KK]Kirill:) thanks12:40
[KK]Kirillnuclear_eclipse: close this bug 1038412:41
mantisbotBug 10384 - ruslan - open - new12:41
mantisbotRoadmap can/should show priority  symbol for issues. - http://www.mantisbt.org/bugs/view.php?id=1038412:41
nuclear_eclipse[KK]Kirill: I'm assuming you're requesting that 10384 be committed to master?13:07
[KK]Kirillnuclear_eclipse: Yes. How right request apply patch to master?13:07
nuclear_eclipsethe "proper" method is to ask here or on the dev list; I just need to review it and/or make sure other devs agree that it should be included13:08
[KK]Kirillmay be add resource where will be approved or disapproved patches?13:11
nuclear_eclipsewould be a nice idea, but it's more of a concern of how to put something like that together without it just being a list of patch files...13:11
[KK]Kirillto view what thank each of all devs?13:12
[KK]Kirillok.. I try to see what I can found of this topic13:12
[KK]KirillJohn, I test your brunch strings13:13
[KK]KirillIt's work.13:13
nuclear_eclipsethis is giving me a good idea for a new plugin though :)13:13
[KK]Kirillnuclear_eclipse: :)13:14
nuclear_eclipse[KK]Kirill: well, it "works" because I haven't actually hooked anything up to the new API functions yet :P13:14
[KK]Kirillnuclear_eclipse: I in my local brunch change mbfunction to string_*13:15
nuclear_eclipseI'm going to try some `sed` magic to see if I can't automate much of the conversion process....13:16
[KK]Kirill:)13:16
[KK]KirillI talk yesterday with ruslan13:16
nuclear_eclipseI noticed13:17
[KK]KirillHe has one wish13:18
[KK]Kirillfor change version system13:18
[KK]KirillVersion [ 3] [b1]13:19
[KK]KirillVersion [ 3] [b2]13:19
[KK]KirillVersion [ 3] [b35]13:19
[KK]KirillVersion [ 3] []   << release (contain all change in beta version)13:19
[KK]Kirilland after release in changelog - delete all betas13:20
nuclear_eclipsewell, that's more of a management issue - it's always possible for an admin to update the fixed-in version after the fact13:21
[KK]KirillI don't think about it.13:22
[KK]KirillHow you set short numbers of mantis issue in mantisbt on your site?13:22
nuclear_eclipseor alternatively, a team could target issues for alpha/beta releases, but always set the fixed-in version to the final release13:22
nuclear_eclipsefor our own site, we prefer to track what changed in each pre-release, and keep that history13:22
[KK]KirillOk13:25
[KK]KirillHow you set short numbers of mantis issue in mantisbt on your site?13:26
nuclear_eclipsewhat do you mean?13:26
[KK]KirillIn your site Issue number length 3 chars13:27
[KK]KirillIn main mantis 713:27
[KK]KirillHow you did this13:27
[KK]Kirillfor example http://leetcode.net/mantis/view.php?id=3113:29
nuclear_eclipse$g_display_bug_padding = 3;13:29
nuclear_eclipse$g_display_bugnote_padding = 3;13:29
[KK]Kirillsorry, I found. Thanks13:30
[KK]KirillJohn, I right rollback last commit? see next13:36
[KK]Kirillgit tag wplast13:36
[KK]Kirill//edit files13:37
[KK]Kirillgit commit -a13:37
[KK]Kirillgit formate-patch ... wplast > ....13:37
[KK]Kirillgit reset --hard wplast13:38
[KK]Kirill?13:38
nuclear_eclipse[KK]Kirill: it might be better to use branches for something like that, such as:13:39
nuclear_eclipsegit checkout -b branchname13:39
nuclear_eclipsegit commit13:39
[KK]KirillI think you right13:39
nuclear_eclipsegit format-patch master > ....13:39
nuclear_eclipseat point, you can checkout master again to be back at the "official" place, but still keep the branch with your local changes13:40
paulr.13:41
[KK]Kirillpaulr: Hi13:42
paulrlo13:42
paulrmy interent broke last night13:42
* paulr couldn't push changes :(13:42
[KK]Kirillwhy?13:42
paulrdunno :)13:42
paulrisp broke :)13:42
nuclear_eclipsehi paulr13:42
[KK]KirillWhat now?13:42
* paulr @ work13:46
[KK]Kirillthis task can be closed http://www.mantisbt.org/bugs/view.php?id=826813:46
paulrnuclear_eclipse: can i break trunk a bit?13:46
paulr(plsssssssssssssssssssss)13:46
* [KK]Kirill *ROFL*13:46
* [KK]Kirill has 2 adsl internet @home :)13:47
nuclear_eclipsepaulr: how broken?13:48
mantisbotNew bug: Bug 10422 - ruslan - open - new13:56
mantisbotNew bug: [NEW] ability to mark some versions as ARCHIVE - http://www.mantisbt.org/bugs/view.php?id=1042213:56
giallunuclear_eclipse, ping14:12
nuclear_eclipsehi giallu14:12
gialluhi14:12
giallucan I kill the "this is a test" guy?14:12
Kirill_Krasnovgiallu: You MUST KILL HIM :)14:13
gialluare you able to check if he made _any_ real bug report?14:13
nuclear_eclipsegiallu: view issues page?14:14
Kirill_Krasnovnuclear_eclipse: too much test issues14:15
Kirill_KrasnovBAN HIM14:15
nuclear_eclipseaccording to view issues, he hasn't reported anything that hasn't yet been deleted :P14:15
nuclear_eclipseKirill_Krasnov: I would if I had the access permissions to14:15
giallunuclear_eclipse, I just hasck my account to admin level...14:16
gialluhacked even...14:16
gialluthat's why I'm asking what to do14:16
nuclear_eclipseI was tempted to do the same14:16
nuclear_eclipseI suggest just disabling his account without actually deleting it14:16
Kirill_Krasnovnuclear_eclipse: may be add checked system for ip?14:16
Kirill_Krasnovban for some submask14:17
nuclear_eclipsebanning by IP is generally a very inefficient/poor method to handle that14:17
gialluok. disabling it is (tbh, I don't see the point in keeping the account disabled. it just means we will not be able to register again with the same name)14:17
nuclear_eclipsedisabling his account should be plenty14:17
nuclear_eclipsegiallu: that's the point ;)14:17
nuclear_eclipseby disabling it, we at least prevent him recreating the account...14:18
gialluok then...14:18
nuclear_eclipseI'm all for sending him a scathing email as well...14:18
* giallu thinks about a feature to show a custom, per user, message when the account is disabled14:18
nuclear_eclipseyes, that would be handy14:19
nuclear_eclipseeg, 'reason for being disabled'14:19
Kirill_Krasnovwho has right in IRC?14:19
nuclear_eclipseI do14:19
nuclear_eclipseand paulr does14:19
Kirill_Krasnovkick [KK]Kirill14:19
Kirill_Krasnov:)14:21
Kirill_KrasnovNickname is already in use.: [KK]Kirill14:21
Kirill_Krasnov:(14:21
nuclear_eclipseKirill_Krasnov: in the future, if you register your nick with freenode, you can "recover" your account14:21
Kirill_Krasnovnuclear_eclipse: How?14:22
nuclear_eclipse /msg nickserv help14:22
nuclear_eclipsefor now though, you'll either need to wait for your shadow to drop, or visit #help (iirc) and talk to a server op14:23
Kirill_Krasnov[17:25]Notice from NickServ (FreeNode): You may not ghost [KK]Kirill.14:23
[KK]Kirillnuclear_eclipse: Thanx14:27
nuclear_eclipseyou're welcome14:27
[KK]Kirill:)14:27
[KK]Kirillnuclear_eclipse: If I create patch, I must create issue in mantsi before or just write in mail why I wish apply patch in dev build?14:44
nuclear_eclipseattaching it to an issue is probably the best method, but not required14:44
[KK]Kirillok14:45
mantisbotNew bug: Bug 10423 - Kirill - open - new15:07
mantisbotNew bug: Get all issue from given user - http://www.mantisbt.org/bugs/view.php?id=1042315:07
giallu[KK]Kirill, I beliece that is covered by an older "monitor user acitivity" bug report15:12
[KK]Kirillgiallu: What issue number? I don't found issue15:12
[KK]Kirillgiallu: I create patch for this15:13
[KK]KirillI have local trouble15:13
[KK]Kirill#1146: Table 'kkrasnov.mantis_bug_revision_table' doesn't exist, запрос: SELECT COUNT(id) FROM mantis_bug_revision_table15:13
[KK]KirillWHERE bug_id=3 AND type=3 AND bugnote_id=0.15:13
[KK]KirillI create new mantis some mounth ago from git15:14
[KK]KirillAnd table not created?15:14
[KK]Kirillplease check new installation15:14
giallubug 815715:15
mantisbotBug 8157 - giallu - open - assigned15:15
mantisbotUser activity report - http://www.mantisbt.org/bugs/view.php?id=815715:15
[KK]KirillI forgot about note15:16
[KK]Kirillbut I don't think that more needly15:17
[KK]Kirillupgrade database from admin - don't help - table not created15:21
[KK]Kirillwhere I can see structure of this table?15:25
[KK]KirillI found in schema.php15:26
[KK]KirillHow I can execute create this table?15:27
* [KK]Kirill go home15:30
paul_.17:20
nuclear_eclipsehi paul_17:21
paul_lo17:21
paul_internet at home works :))17:21
nuclear_eclipseyou should register paulr as your nick, and boot out anyone who connects with it...17:21
paul_I have17:21
paul_nuclear_eclipse: I fixed some utf8 things over weekend18:28
paul_how can i link the commits to bugs18:28
nuclear_eclipse"fix #xxx"18:28
nuclear_eclipseor "issue #xxx"18:28
paul_after commit18:28
nuclear_eclipsehave you already pushed the commits to official repo?18:29
paul_nope18:29
nuclear_eclipsethen you can a) use `rebase -i` and/or `commit --amend` to edit the commit messages, or b) push to repo, and manually link the commits to the issues through the source integration interface18:30
paul_b sounds easier18:30
nuclear_eclipsea is rather easy to18:30
paul_can we talk about helper_ensure_post briefly18:31
nuclear_eclipsemaybe18:31
paul_# helper_ensure_post();18:31
paul_in query_delete.php18:31
paul_<form method="post" action="<?php print $t_delete_url; ?>">18:32
paul_<br /><br />18:32
paul_<input type="hidden" name="source_query_id" value="<?php print $f_query_id; ?>"/>18:32
paul_<input type="submit" class="button" value="<?php print lang_get( 'delete_query' ); ?>"/>18:32
paul_</form>18:32
paul_there's no form security stuff there right?18:32
nuclear_eclipseright18:32
paul_is it just the case of doing:18:32
paul_echo form_security_field("somefoo")18:33
paul_in the <form> tags18:33
nuclear_eclipsethat's half of it18:33
paul_and then just calling form_security_validate("somefoo") instead of helper_ensure_post18:33
nuclear_eclipseand form_security_purge("somefoo") after all the processing has been completede18:33
paul_any thing else to be aware of ?18:34
nuclear_eclipsejust make sure that "somefoo" is unique among all the other form security strings for that specific action18:35
paul_and now last thing18:35
paul_last night,  I went back to looking at my bugobject branch18:35
nuclear_eclipsein this case, "query_delete" maintains the current habit of naming the security string after the processing page18:35
paul_and had to do a merge18:35
paul_which conflicted18:35
paul_can i collapse the commits in this branch into one?18:37
nuclear_eclipseyeah, either using `rebase -i` and setting all commits after the first to 'squash', or by using `merge --squash`18:39
paul_so if I wanted to merge/squash bugobjects branch into bugobjects2 branch18:39
paul_how would I do this :)18:40
nuclear_eclipse`git co -b bugobjects2 master; git merge --squash bugobjects`18:40
paul_$ git merge --squash bugobjects18:41
paul_Auto-merging view_all_bug_page.php18:41
paul_CONFLICT (content): Merge conflict in view_all_bug_page.php18:41
paul_Squash commit -- not updating HEAD18:41
paul_Automatic merge failed; fix conflicts and then commit the result.18:41
nuclear_eclipsecongrats18:41
paul_...18:41
paul_stupid thing18:43
paul_nice18:44
paul_nuclear_eclipse: sorted :)18:45
SeabladeHey quick question, we are getting a couple of users reporting being unable to login or change passwords currently in an installation of Mantis, they are getting "Message: APPLICATION WARNING #2702: Your session has become invalidated."  What would be causing this, and is there a way to address it(Fix or workaround?)  I didn't find much in the BT.18:46
paul_Seablade: speak to nuclear_eclipse18:46
nuclear_eclipseSeablade: are you using a proxy server?18:47
Seabladenuclear_eclipse: For the hosting or the client?18:47
nuclear_eclipseeither18:47
paulrthe client user18:47
nuclear_eclipseSeablade: also, what verson of Mantis?18:47
Seabladenuclear_eclipse: I can't speak for the client, but can check.  The hosting is on a shared host, but I don't think there is a reverse proxy in effect there, I could be wrong18:47
Seabladenuclear_eclipse: 1.2.0a318:47
Seabladenuclear_eclipse: Dreamhost18:48
nuclear_eclipseSeablade: you can disable the session validation from the configuration file, but note that it opens a security hole, allowing potential session hijacking18:48
Seabladenuclear_eclipse: Is this a common issue if a proxy is involved?18:49
* paulr wants to know what proxy is in the loop btw18:49
nuclear_eclipseSeablade: it's a problem for any proxy server that doesn't obey the caching headers that Mantis sends18:49
nuclear_eclipseor wait18:49
nuclear_eclipsethat's a different issue18:50
nuclear_eclipsethe session issue happens with any client that connects to Mantis from a load-balanaced proxy server, where each connection can look like it's coming from a different IP address18:50
* Seablade is checking on the presense of a proxy with one of the users that reported the problem now, assuming he got the right user18:50
paulr$t_columns = get_class_vars ('BugData');18:52
paulrreturns the default values too???18:52
nuclear_eclipseReturn Values18:52
nuclear_eclipseReturns an associative array of default public properties of the class. The resulting array elements are in the form of varname => value.18:52
SeabladeHmm I don't think Dreamhost is set up like that, in as far as they do have individual servers that they are sharing, not really a cluster setup in the way MediaTemple offers to my knowledge.  However the user I just checked with is having issues going through an iPhone getting that error, so it is a good bet that they are indeed running through a proxy18:52
nuclear_eclipsephp.net/manual ftw...18:52
nuclear_eclipseSeablade: the problem is most likely caused by a) EDGE/3G disconnecting/reconnecting with a new IP address, and/or b) hopping between wifi and EDGE/3G and back, etc18:54
nuclear_eclipseit's an eventual goal to implement the user option to disable IP-based session validation at login time18:54
paulrnuclear_eclipse: that just leaves bug api18:55
nuclear_eclipseor implement some other basis for session validation to help prevent hijacking  - see bug 9744 on the official tracker18:55
mantisbotBug 9744 - seiji - open - assigned18:55
mantisbotUsers behind proxy will get 'ERROR_SESSION_NOT_VALID' - http://www.mantisbt.org/bugs/view.php?id=974418:55
paulrsee john this is the problem with never doing a release :)18:56
paulrusers run beta's :)18:56
Seabladenuclear_eclipse: Ok, thanks.  I will check with the other user[s] that report this issue and see if we can get it sorted out without disabling the session validation preferably.  Ill probably end up being back in the future with other possible problems:)  Thanks for the help18:56
paulrfeel free to help out with suggestsions of fixes btw :)18:57
Seabladepaulr: Heh not sure why this project switched to the alpha, I am sure there was reasoning, but yea we upgraded since then based off the issues with web browsers, etc.18:57
Seabladepaulr: Heh anytime I can I certainly will, it is just a matter of whether I can or not, not the webmaster for this, just someone that helps out with the bugs:)18:57
nuclear_eclipseSeablade: you're welcome18:58
Wildmanafternoon everyone18:58
WildmanGiallu are you listening?18:58
paulrhi18:59
paulr-       $t_status = config_get( 'bug_submit_status' );19:06
paulr+       $t_starting_status = config_get( 'bug_submit_status' );19:06
paulr+       $t_original_status = $c_status;19:06
paulrnuclear_eclipse: what's this crap19:06
nuclear_eclipsecontext?19:07
paulrsome commit19:07
nuclear_eclipsethat helps19:08
Wildmangiallu this is Bill, we talked Friday regarding the import issues plugin if you have time to speak with me I have a couple of questions19:09
paulrnuclear_eclipse: what time are you leaving tonight?19:10
nuclear_eclipsepaulr: in about 10 minutes, for class :P19:10
nuclear_eclipseWildman: giallu is in Italian timezone, so he may not be around until a bit later, or much later :P19:11
paulrnuclear_eclipse: are you/will you beback after?19:12
nuclear_eclipseperhaps19:13
nuclear_eclipseyou can always just ping me in a couple hours and see if I respond :P19:14
Wildmanthanks Nuclear19:14
WildmanI talked to him last Friday around this time I think he was trying to get his kids to sleep LOL19:14
paulrI might just do bigpush tomorrow :)19:14
nuclear_eclipsepaulr: can you double check bug 10369?19:16
mantisbotBug 10369 - Kirill - open - assigned19:16
mantisbotVerification link not work - http://www.mantisbt.org/bugs/view.php?id=1036919:16
paulri think your right19:16
paulror well19:16
paulrI was doing a "it's html it should be &amp;19:16
paulrbut you mgiht be right: )19:16
nuclear_eclipseI think we need to revert your "fix" commit, as you were fixing a broken mail client and breaking everything else :P19:16
paulrI was wondering if we could just come up with a sequence and split it19:16
paulre.g.19:16
paulratm, we do something like19:17
paulrusername=foo&key=dfjklsdjkl or something right?19:17
paulrso if we did say19:17
paulrkey=foo___***___dfjklsdjkl19:17
paulror something19:17
paulrwe could fix both19:17
paulri.e. take some char that isn't allowed in a username19:17
paulrand use it as a split19:18
nuclear_eclipsehmm, that sounds like a dirty hack; I'd much rather not...19:18
paulrwell, stupid question19:18
paulrsurely it should be &amp; though?19:18
nuclear_eclipsewe can just tell Entourage users to complain about getting the bug fixed in Entourage, rather than pandering to them...19:18
nuclear_eclipseno, the emails aren't HTML, so &amp; is invalid19:18
nuclear_eclipsethe only reason you need &amp; in HTML is because HTML is annoying...19:19
paulrhttp://www.tech-archive.net/Archive/Mac/microsoft.public.mac.office.entourage/2004-04/0558.html19:19
nuclear_eclipseanywho, I gotta run19:20
nuclear_eclipsecheers19:20
mantisbotNew bug: Bug 10424 - cbradney - open - new19:43
mantisbotNew bug: Requesting option to disable the use of some versions from the list - http://www.mantisbt.org/bugs/view.php?id=1042419:43
gialluWildman, I'm around now20:07
giallutoo late...20:07
paulrlo giallu20:16
gialluhi paulr20:17
gialluwhat kind of brakeage are you planning?20:17
paulrthe normal20:18
paulri've edited a regex to do with url's for utf820:18
paulrand started merging one of my branches in properly20:18
gialluwhich one?20:22
nuclear_eclipsepaulr: that sounds like it should be reviewud before pushing...20:22
paulrbugobjectone20:22
giallunuclear_eclipse, since when we review _before_ pushing?20:22
giallu:P20:22
paulrnuclear_eclipse: it's already had some input :)20:22
paulrtbh, i might make a bundle tpye thing20:23
nuclear_eclipseutf8 just seems like one of those things that's liable to make things even worse... :P20:23
nuclear_eclipsepaulr: please don't make bundles, just make formatted patches, they're *so* much easier to work with20:23
paulryour just lazy20:24
paulrI thought the point of bundles is they just apply to your repo20:27
paulr:)20:27
nuclear_eclipsethe point of bundles is to let one repo transfer data to another, not for human use20:31
nuclear_eclipseformatted patches and remote references are what humans should use...20:31
nuclear_eclipseas the manpage for git-bundle says: "This command provides support for git-fetch and git-pull to operate..."20:32
CIA-11Mantisbt: s.mazeland * re9cbf60f8c1b / (10 files in 5 dirs): Fix: #0007484: implement RTL css20:33
nuclear_eclipseie, it's a plumbing command, not porcelain20:33
paulrdo we really need a config variable for RTL css ;/20:36
gialluno, please, not another one....20:36
nuclear_eclipseapparently we do........ =\20:37
giallucan we state a goal of, let's say, max 5 config variables for mantis 2.0 ?20:37
paulrlol20:37
* nuclear_eclipse beats giallu with a big stick for mentioning 2.020:37
gialluhey nuclear_eclipse, if you like IO can propose that for 1.220:38
gialluso we release in 201220:38
* nuclear_eclipse hits head on desk...20:38
paulrnuclear_eclipse: review20:38
paulrpls20:38
paulrk20:38
paulrthx20:38
paulrwheres daryn20:38
gialludaryn hitted desk too strongly20:39
paulrdo we need to call string_attribute before we put it into a dropdown list?20:41
nuclear_eclipsepaulr: I need a URL before I can review anything....20:41
nuclear_eclipsepaulr: no, string_line should suffice20:41
paulrno i mean for editing20:41
paulri.e.20:42
paulr<?php print_category_option_list( $t_bug->category_id, $t_bug->project_id ) ?>20:42
paulrthat would normally be:20:42
paulrprint_category_option_list( string_attribute( $t_bug->category_id ), $t_bug->project_id );20:42
nuclear_eclipseI don't think you need it in that case...20:43
paulrgimme10min20:43
paulr                       <?php print_version_option_list( $t_bug->version, $t_bug->project_id, $t_product_version_released_mask ); ?>20:46
paulrwhat about that case?20:46
paulryes?20:46
nuclear_eclipseI don't know off hand, but I would think not20:53
nuclear_eclipsethe print_*_list functions sholud apply any stirng_* functions as necessary20:54
paulrhow do i push a branch to a fork again20:57
paulr(on mforge20:57
siebrandnuclear_eclipse: john, any planning for http://www.mantisbt.org/bugs/view.php?id=10271 (mb_string warning)21:01
* paulr doesn't think we should require mb_string21:02
siebrandpaulr: for proper support for certain methods in many languages, it is almost unavoidable. We need to devise some way to warn people if they choose a language that needs it.21:04
siebrandpaulr: only 15% of the world population understands English.21:04
paulrsure21:05
paulrat the same time, in some cases (over weekend) things I notice that we break on21:05
paulrdont require mb_string to fix21:05
paulrfor example as 2 examples:21:06
paulrin our email api we were using wordwrap to wrap the email width to 75 characters21:06
paulrthen using phpmailer's wordwrap functionality (which supports utf8 supposedly) to wrap to 80 characters21:06
paulr--21:06
paulr2nd example: the url parsing stuff nukes utf8 which again I think i've fixed over weekend21:07
siebrandyeah, it's full of bugs. Let's tackle them one at a time, and not forget that making MantisBT available in languages other than Lantin scripted is a Good Thing (tm)21:07
giallupaulr, your examples are fine, but sibrand is right. we can't support all languages without mb_*21:08
paulrgiallu: depending on what we are doing21:08
gialluso we definitely need to prevent users from using mantis with non-english languages if mb_String is not there21:08
giallunote:21:08
siebrandbtw, why is it that if I choose to sort on issue id in http://www.mantisbt.org/bugs/view_all_bug_page.php?filter=3255091 that I go from 28 results to 5?21:08
giallunot even that guarantees we have correct utf handling21:09
paulras a 3rd example21:10
paulrstrlen()21:10
paulr(which doesn't support utf8)21:10
paulrmy understanding is there's 3 possible ways to do a utf8 strlen:21:10
paulr1) call mb_strlen(&$foo, 'UTF-8');21:11
paulr2) call iconv_strlen(&$foo, 'UTF-8');21:11
paulr3) call strlen(utf8_decode(&$string));21:11
paulrif we chose option 3, it would be slower, but would work for everyone even without mb_*21:13
giallutbh, 2 and 3 sound like workarounds I don't really want to see21:13
siebrandstrtoupper (which I have on 257 lines on our code) completely messes up everything to do with Cyrillic. Compare it to having random X-es in your UI.21:14
giallusiebrand, right. but there are more21:14
* siebrand nods knowingly.21:14
paulrgiallu: iconv isn't... :)21:14
giallupaulr, so are we planning to outsmart php guys here?21:14
siebrandmb_string is a solution for a gamut of unicode issues.21:15
siebrandiconv is the work around.21:15
siebrandwhat exactly is the con for mb_string?21:15
gialluI have no problem with it... expecially because is for non-english locales21:16
siebrandIMO we just require it, and be done with it. It will enable to improve out product tremendously with regards to language support. Where will its quality decrease?21:16
paulrthere's not that much that breaks without it atm21:18
paulrsome of the things that break21:18
paulrare our regex's21:18
* giallu sighs... 21:18
paulrmb_string does't help with that21:19
siebrandpaulr, you are being Anglo-centric to the bone.21:19
paulrno21:19
siebrandwhy not?21:19
paulri've spent a fair amount of time recently starting to look at the issues that are coming up21:19
gialluif nothing is broken, can you explain http://tinyurl.com/cegwwy ?21:21
gialluand this is the tip..21:21
giallubtw, I'm tired...21:22
paulrgiallu: can I exclude the closed issues when responding to this? :)21:23
gialludon't bother21:23
paulrone of the main problems21:24
paulrthat I commented on over weekend21:24
paulrpaul_http://www.mantisbt.org/bugs/view.php?id=926912:4821:24
paulrpaul_^^ that seems to be 'biggest' problem12:4921:24
paulrpaul_we should force utf8 no?12:4921:24
paulrdhx_mI think so13:0621:24
paulrdhx_mI can't see any reason why you wouldn't be using UTF821:24
paulrbug 926921:24
mantisbotBug 9269 - 0xHH - open - acknowledged21:24
mantisbotOn installation mantis create database with default collation settings. - http://www.mantisbt.org/bugs/view.php?id=926921:24
paulri.e. atm21:25
paulrif you upgrade/create a db21:25
paulrit doesn't necessarily use utf8 as type21:25
gialluthat's _another_ problem21:28
gialluthat we should try to fix21:28
paulrnod21:28
paulri've got some ideas on that21:28
paulri'm just waiting confirmation on if we should *always* be telling mysql utf8?21:28
gialluyes we should21:29
paulr(for new installs)21:29
siebrandpaulr: from what/whom? It's YES!21:29
gialluascii is a subset of utf21:29
gialluso going utf you don't break anyone21:29
giallubut fix those non ascii21:29
gialluso it's a must21:30
paulrsiebrand: one of my weekend work was to fix(or at least try to) the string_insert_href function21:30
paulrthat parses urls21:30
siebrandgiallu: <unknown> will have to export db, convert, import before being able to update.21:30
paulrand what not to work with utf821:30
siebrandpaulr: would be resolved if we *require* a db convert to UTF 8 for 1.2 branch and up.21:30
paulrnod21:31
giallupaulr, I had a working conversion script21:31
siebrandpaulr: and we could drop all the non-utf locales that are there for nothing anyway at the moment...21:31
giallubut never had time to convert to something usable by mantis21:31
siebrandbtw, any reason why I should not drop those in master anyway?21:31
giallusiebrand, if we have the corresponding utf one, I'd say kill'em21:32
siebrandgiallu: we do.21:32
siebrandok, easy enough :)21:32
siebrandgiallu: poof. They're gone.21:59
CIA-11Mantisbt: s.mazeland * re593fed9cecc /lang/ (32 files): Remove obsolete and outdated non-UTF8 language files. All have a better maintained UFT8 equivalent.22:00
giallu:D22:00
siebrandfinally it's easy to count that we have 44 locales.22:01
gialluwe should add it to the website...22:01
nuclear_eclipsejust to add my two cents, I think we should check/warn for installations without mb_*, but still support them by using standard string functions for cases where the hosting environment doesn't have the multibyte PHP extensions, and the site is never planning to use anything other than English anyways; ie, we should be as compatilble as possible while still warning/recommending having the multibyte extension22:19
nuclear_eclipsewhich is exactly the way I'm headed with my work for bug 1030322:20
mantisbotBug 10303 - jreese - open - assigned22:20
mantisbotCreate and use string_api wrappers for multibyte string functions - http://www.mantisbt.org/bugs/view.php?id=1030322:20
paulrnuclear_eclipse: http://git.mantisforge.org/w/mantisbt/paul.git?a=commitdiff;h=83784b10182a64013c0cdeb6ed0f4ddf737fe68d22:23
paulrnuclear_eclipse: http://git.mantisforge.org/w/mantisbt/paul.git?a=commitdiff;h=5226fbc4b721dd47627c987e57d2c7d5c5a1431a22:24
* nuclear_eclipse looks22:29
paulrfeel free to look at rest of that branch22:29
nuclear_eclipsethe wordwrap diff looks fine to me22:32
nuclear_eclipsewhat's with the $t_url_hex?22:32
paulrme trying to work out what the regex does ;p22:33
nuclear_eclipseI'm just not sure why you need the hex stuff since it's already covered by url_valid_chars....22:34
paulryou had it :)22:34
paulrthat's a null change22:34
nuclear_eclipse...  I don't remember that...22:34
paulrall i've done is22:34
paulra) added /pL22:35
paulrb) add /u22:35
paulrerm22:35
paulrmake that \pL22:35
paulrso /se becomes /sue22:35
paulrand [:alnum:] is now followed with a \pL which apparently is unicode letters22:35
nuclear_eclipsewhat's \pL ?22:35
paulrapparently utf8 leters of any case22:36
nuclear_eclipsewell, I gotta go eat dinner, I'll continue looking at it later22:37
CIA-11Mantisbt: s.mazeland master-1.1.x * r857ebb166f78 / (config_defaults_inc.php lang/strings_arabic.txt): Fix #0010405: Added Arabic localisation. Contributed by kawzaki.22:42
CIA-11Mantisbt: s.mazeland master-1.1.x * r9f5f8c95b4fa /lang/strings_polish.txt: Fix syntax error in strings_polish.txt. Where did that come from?22:45

Generated by irclog2html.py