Wednesday, 2009-12-16

../irclogs/#mantishelp.2009-12-16.log
--- scribe started ---00:00
CIA-27Mantisbt: vboctor * r77a0fb65a207 /core/user_pref_api.php: Fixes #11046: notification emails send to an unassigned user.09:29
CIA-27Mantisbt: vboctor master-1.2.x * r6ff6210c3e88 /core/user_pref_api.php: Fixes #11046: notification emails send to an unassigned user.09:29
killefizgiallu: what do you think about pushing a git-snapshot of 1.2.x to rawhide?10:36
giallukillefiz, I considered doing that about 6 month ago... I'm just glad I waited :)10:39
giallunow it seems we're _really_ near a release, so it could make sense10:40
killefizi'm using 1.2 in production for months now10:53
dhx_mI just came across the best thing I've seen in HTML forms ever11:00
dhx_mHTML5 has a text field that has a draggable corner, allowing users to resize the field to meet their needs as they type11:01
dhx_mand the new <datalist> tag also seems very useful11:05
dhx_messentially you can have autocomplete in text fields where values are taken from the datalist11:05
dhx_mthat might be handy for our project or user selector where we can have 100's of projects and 1000's of users to select from11:06
kirillkanuclear_eclipse: around?13:15
nuclear_eclipsehowdy13:31
paul__.13:59
paul__wuts rawhide13:59
nuclear_eclipsefedora's unstable/testing14:00
paul__have you seen the MS docs14:00
paul__on php stuff14:00
paul__erm14:00
nuclear_eclipseno14:00
paul__for their application deployment thingie14:00
paul__OMG14:02
* paul__ knows what he's doing tonight14:03
kirillkanuclear_eclipse: hi14:57
kirillkaJohn, did you know git web interface for closed project?14:57
kirillkasuch github with closed projects14:58
kirillkanuclear_eclipse: John?15:17
CIA-27Mantisbt: vboctor * r5955e8b9aa3f /bug_change_status_page.php: Fixes #10023: duplicate_id maxlength is hardcoded in bug_change_status_page.php.15:20
CIA-27Mantisbt: vboctor master-1.2.x * ref5688b47591 /bug_change_status_page.php: Fixes #10023: duplicate_id maxlength is hardcoded in bug_change_status_page.php.15:21
paul__dhx_m:15:23
paul__did you fix the url issuesxss?15:23
paul__a16:08
nuclear_eclipseomg, these splits are driving me insane16:15
nuclear_eclipseI've never seen freenode so split-happy before16:15
paul__nuclear_eclipse: :)18:13
nuclear_eclipsehi paul__18:31
* paul__ is having fun :)18:32
paul__or well18:32
paul__will be18:32
ezrawanyone know how I can get my custom fields into the csv export?18:39
kirillkanuclear_eclipse: did you know git web interface for closed project? Such github with closed projects.19:47
Dgtlrift_at_workAre there any plugins to render Gantt charts in mantis?20:16
gialluDgtlrift_at_work, good question... I never come across one20:17
Dgtlrift_at_workCan you make a guess at the time to implement such a plugin? Is there another plugin that does graphs that can be borrowed like one that dies a burn down?20:25
qwebirc611429This is much better... Dod you get my last question?20:27
* giallu still pondering20:28
gialluI'm not sure we wave any plugin with semi-complex graphs20:29
gialluwe _have_20:29
qwebirc611429unfortunate20:29
qwebirc611429It would seem for the gantt graph we /mosty/ everything - relationship tracking, issue complexity (effort in terms of time,) and project association...20:31
qwebirc611429.../mostly/ have evverything...20:31
qwebirc611429... what is lacking would be interproject resource tracking - ei joe is 100% on project "A" and has the expertiese associated with Ticket X, Y, and Z.20:33
qwebirc611429BTW: this is dgtlrift_at_work - I just couldn't handle the IRC client through the web page on my phone - so now I'm at a desktop.20:34
qwebirc611429We also have the graphing code for the roadmap that could be recycled.20:36
qwebirc611429giallu - do you agree?20:39
gialludgtlrift_at_work, sure it could20:39
dgtlrift_at_workWould I be better off documenting it in a ticket to get some feedback?20:42
dgtlrift_at_workgot to go - tnx for the advice20:45
paul__whos around?21:38
nuclear_eclipseme21:38
paul__hmm21:38
paul__you'll do21:38
paul__atm we can customise fields on bug report page right?21:39
nuclear_eclipseyep21:39
paul__but not order them?21:39
nuclear_eclipsedunno21:39
paul__or well, lay them out?21:39
nuclear_eclipsebut that reminds me, we wanted to change all those constants to normal strings21:39
paul__if we were to lay them out21:39
paul__nod21:40
paul__almost done that :P21:40
paul__if we were to lay them out21:40
paul__we'd need to store array( array(id, status), array( os,os_build,blank,eta) type structure?21:40
paul__serialise that into config ?21:40
nuclear_eclipsethat sounds insane :P21:44
paul__well table?21:55
paul__or what :P21:55
paul__you have 24 hours to come up with a good way to store rows :P21:56
nuclear_eclipsedo we really need to worry about that right now?21:56
paul__yep :P21:57
paul__you know it's xmas?21:57
nuclear_eclipsecan we worry about cleaning up 1.2 first?21:57
paul__I kinda am :P21:58
paul__i'm going through my list :)21:59
* paul__ wonders if dhx_m has a view22:10
paul__wow22:44
paul__wtf is helper_get_columsn_to_view22:44
nuclear_eclipseugh22:44
nuclear_eclipseI don't even feel like trying to explain it22:45
paul__? :)22:45
paul__well22:45
paul__it's a fairly random function22:45
nuclear_eclipseit's one of those things I have to re-explore every time I come across it because it's so weird and indirect22:45
paul__I've just coded a columns_get_default() function to use in config_Set22:46
paul__and now found helper_get_columns_to_view22:46
paul__which does something similar but not the same!22:46
paul__(columns_get_default is to replace the hardcoded columns in config_default's as the hard coded columns don't work properly as you can turn features on/off so need to be dynamic22:47
paul__now I need to work out if i've22:47
paul__a) just wasted my time22:47
paul__b) if I can just deleted the helper_* function22:47
nuclear_eclipseI'm pretty sure that helper_ takes care of all that22:47
paul__c) if we need both :/22:47
nuclear_eclipseeither way, don't break the way that plugin columns works22:48
paul__i'm not :)22:48
paul__I'm purely making default config_get take a default value of not ''22:48
paul__or well22:48
paul__removing $g_csv_columns from config defaults22:48
paul__and replace with a default value in config_get of column_get_default('csv')22:49
paul__such that the default values are appropriate for fields that are actually enabled22:49
paul__(such that we dont store columns that are default and unchanged in db)22:49
dhx_mhi22:51
paul__hi22:51
mourisjjoin #mirth22:52
nuclear_eclipsedinner time, might be on phone22:52
nuclear_eclipsecheers22:52
paul__mourisj: no22:52
mourisjsorry22:52
paul__:P22:53
mourisjbtw, they are currently mirthcorp.com22:54
paul__dhx_m: confirm: duplicate_id = dead?23:06
dhx_mpaul__: I'm not too familiar with "duplicate_id"...?23:07
paul__prior to relationships :P23:07
paul__can you review a patch? :P23:10
paul__also any thoughts on layout comment I had with john?23:11
dhx_mwell duplicate_id is still in the schema23:11
dhx_mwhatever it does23:11
dhx_mpaul__: I'm not 100% sure on the best way to approach ordering of fields23:12
dhx_mpaul__: it really should be something the user can customise in a UI (when they select what fields to show)23:13
dhx_mpaul__: however when we move to templating, it's going to get ugly when we have to change the order of templates23:13
* paul__ is more keen to allow users to customise fields in UI23:14
paul__kinda interested in ways to store data23:15
* paul__ goes to look at what deboutv did23:15
paul__dhx_m: so options23:24
paul__a) store sql tables23:24
paul__b) store a config entry (Array( '0_0'=>'reporter', 0_1=>foo etc23:25
paul__or multideminsional array23:25
dhx_mwell the problem I see is that some layout templates may have 2/3 columns23:29
dhx_mothers may have something completely different23:29
dhx_mso it becomes very hard to order stuff if we use that approach23:30
dhx_mthen again, we really need a generic way for printing fields23:30
dhx_mas the user may have 10 custom fields23:30
paul__I need a way to allow a user to modify the order of fields on http://www.mantisbt.org/bugs/view.php?id=435023:30
paul__atm, you can turn some of them off23:31
paul__but not hide all23:31
paul__and not fix layout if you hide some23:31
paul__what deboutv did with my view was23:31
dhx_mI'd say it should be a 2D array from most important to least important fields23:31
paul__set up some massive table23:32
paul__with fields 0_0,0_1,0_2 etc23:32
paul__then stored either blank, fieldname, or mergeleft/right23:32
paul__i'm not sure about merge left/right as that's a bit silly23:32
paul__but23:32
paul__fieldname, width23:32
paul__and row23:32
dhx_minteresting23:32
dhx_mnow you mention it23:33
dhx_mit should probably be a multi-dimensional array23:33
dhx_mwhere you can group fields together23:33
dhx_mit should be possible to design quick algorithms to work out spacing as appropriate23:33
paul__http://www.mantisbt.org/bugs/view.php?id=435023:34
paul__why does top row show23:34
paul__ID23:34
paul__000023:34
paul__whereas everything else show23:34
paul__Reporter: foo23:34
dhx_mI agree, that sucks :p23:35
paul__you could represent that table atm by store an array of rows23:35
dhx_mIMO it should be one or the other23:35
paul__where a row has a type (aka top/bottom, left/right)23:36
paul__and an array of fields23:36
paul__where a field has a variable name and a width23:36
dhx_mI was actually thinking it should be configurable on a per-project basis (and probably per-user or per-group basis too)23:36
paul__in all honesty, I think mantis is something that needs to be skinnable, have a bit of a facelift into ajax world23:37
paul__and after that templates comes down to something to make our lives easier, not endusers23:38
dhx_mthe main reason for templates IMO is to allow for different layouts (mobile, wide screen, simplified, etc)23:38
paul__right23:39
paul__23:34 < dhx_m> it should probably be a multi-dimensional array23:40
paul__23:34 < dhx_m> where you can group fields together23:40
paul__23:35 < dhx_m> it should be possible to design quick algorithms to work out spacing as appropriate23:40
paul__so erm23:40
paul__any thoughts of structure?23:40
dhx_mwe should be able to get spacing information from the fields themselves (maximum length, etc)23:41
dhx_mwhat we're really interested in is grouping of similar fields23:42
dhx_mas well as the order to display them23:42
dhx_mso the array structure would probably look like:23:42
paul__well grouping of similar user could do23:42
paul__if you allowed them to place them23:42
dhx_marray[] = array('os', 'os_version', 'php_version', 'whatever');23:43
dhx_marray[] = 'custom_field_something';23:43
dhx_marray[] = array('target_version', 'due_date', 'etc');23:44
dhx_mso with the current three column layout23:44
dhx_mIt'd show the os, os_version, etc as 2 rows of 2 columns23:44
dhx_mthe custom field thingy as it's own line (perhaps)?23:45
dhx_mand the target_version, etc as 1 row of 3 columns23:45
dhx_mof course, depending on field lengths... the 'etc' field may be a long text field that requires the entire page width23:45
paul__I partly wonder if one just does23:47
paul__(array( array('os',1') array('target_version',2))23:47
dhx_maren't those array orders implicit (keys by default are addressed numerically)?23:48
paul__i mean width23:48
dhx_mwidth doesn't make sense for every layout23:48
dhx_mie. single column view of fields23:49
paul__well, with a single column view, you'd have to put each field on it's own row23:50
dhx_myep, but you can still group them vertically using HTML fieldsets23:50
dhx_mfor example23:50
paul__ahhhmm23:51
paul__I see what you mean23:51
dhx_mI just think the "layout engine" should be responsible for working out widths, spacing, etc23:52
dhx_mbecause they're not the same on every layout23:52
dhx_mand it is something that could be determined quite easily by some quick PHP code23:52
paul__so your thinking an array of sets of fields23:53
dhx_myep23:53
paul__which could be rows or otherwise23:53
dhx_myeah23:53
* paul__ shall have a play23:53
dhx_messentially, it's a group of fields that are similar in some way23:53
dhx_mand because we don't know what fields exist at design time (users may add their own)23:54
dhx_mwe really need to let the users have control over how they're grouped together and how they're ordered23:54
paul__well, that's essentially what defining the table layout would do23:55
dhx_mbut it assumes there is a table :)23:56
dhx_mgrouping also could have other uses23:56
dhx_munrelated to generating views23:56
paul__tbh once you go down the route of it might be a table, it might be css it might be div23:57
paul__gl with plugins ;)23:57
dhx_mmy point is that plugins would specify a new type of field23:58
paul__who mentioned field :)23:58
dhx_maka 'column' in our current implementation23:58
* nuclear_eclipse shudders at the content of the backlog...23:59
dhx_mnuclear_eclipse: how else do we handle templating *and* custom fields at the same time?23:59
nuclear_eclipseumm, we don't? :P23:59

Generated by irclog2html.py