| ../irclogs/#mantishelp.2009-06-20.log | ||
| --- scribe started --- | 00:00 | |
| [KK]Kirill | siebrand: mo | 04:12 |
|---|---|---|
| mantisbot | New bug: Bug 10609 - Kirill - open - new | 07:20 |
| mantisbot | New bug: Move notice to admin workgroup - http://www.mantisbt.org/bugs/view.php?id=10609 | 07:20 |
| dhx_m | hey | 09:30 |
| dhx_m | lol @ japanese translation of my new s_stick_bug_button string being: Sticky設定 | 09:33 |
| dhx_m | hybrid english-japanese | 09:33 |
| paul_ | moo | 09:59 |
| paul_ | dhx does weekends? ;/ | 09:59 |
| dhx_m | all day every day :p | 10:08 |
| dhx_m | I don't do this for work ;) | 10:08 |
| dhx_m | are you on overtime at the moment? :p | 10:08 |
| paul_ | no | 10:09 |
| paul_ | I don't do this for work ;) | 10:09 |
| dhx_m | ah that sucks, I was hoping I could say "GET BACK TO WORK SLACKER!" :p | 10:09 |
| paul_ | you still testing mantis? | 10:09 |
| dhx_m | to which you could make a "PHP is compiling" comeback | 10:10 |
| dhx_m | well I have tested the basic parts | 10:10 |
| dhx_m | and [KK]Kirill has done a lot of testing too | 10:10 |
| dhx_m | 1.2 seems close? :) | 10:10 |
| paul_ | well | 10:10 |
| dhx_m | I'm just adding a few bits and pieces | 10:10 |
| dhx_m | but nothing breakable | 10:10 |
| paul_ | depends if someone breaks smothing | 10:10 |
| dhx_m | what did you have in mind? :p | 10:11 |
| dhx_m | actually one thing I couldn't work out is why we need an event in string_display_line | 10:11 |
| paul_ | mm? | 10:11 |
| paul_ | so you can add bbcode and stuff | 10:12 |
| dhx_m | ah, gotcha :) | 10:12 |
| dhx_m | does the event know the bug ID, etc? | 10:12 |
| dhx_m | or does it only get the string being printed? | 10:13 |
| dhx_m | ie. there must be some sort of context for the event? | 10:13 |
| dhx_m | I haven't done anything with events before | 10:13 |
| dhx_m | but that is what I'm used to in other programs (c) | 10:13 |
| paul_ | did you see 10607 ? | 10:22 |
| dhx_m | nope, let me check | 10:24 |
| dhx_m | you can't be serious... MS SQL doesn't support Unicode text? :o | 10:25 |
| dhx_m | not even UTF16 (the default on Windows) | 10:25 |
| dhx_m | ? | 10:25 |
| dhx_m | do I really need to drive on over to Microsoft and let them know it is 2009 already? :p | 10:25 |
| paul_ | it supports ucs-2 or something? | 10:26 |
| paul_ | (which i think might be utf16?) | 10:26 |
| dhx_m | yep | 10:26 |
| dhx_m | that's what it is | 10:26 |
| dhx_m | and it gets more fun, because UTF16 still has multibyte characters | 10:27 |
| dhx_m | so we'd technically need a UTF16 library to handle that | 10:27 |
| dhx_m | I assume MS SQL supports surrogate pairs for UTF16 (2x 16bit characters) | 10:28 |
| dhx_m | s/multibyte/surrogate pairs | 10:28 |
| dhx_m | ok I replied | 10:35 |
| dhx_m | bug 10607 | 10:38 |
| mantisbot | Bug 10607 - cbasset - open - new | 10:38 |
| mantisbot | Non-ASCII characters, collation and MS SQL Server - http://www.mantisbt.org/bugs/view.php?id=10607 | 10:38 |
| dhx_m | driver? | 11:19 |
| dhx_m | ah I see | 11:19 |
| dhx_m | http://blogs.msdn.com/sqlphp/ | 11:19 |
| paul_ | I need to write an irc plugin for mantis | 11:34 |
| dhx_m | that'd be nice :) | 11:35 |
| dhx_m | but how would it communicate with IRC? | 11:36 |
| dhx_m | you'd need a separate daemon running wouldn't you? | 11:36 |
| paul_ | i mean to replace the mibbit one we have atm | 11:36 |
| dhx_m | ah I see | 11:36 |
| dhx_m | I thought you meant IRC announcing (mantisbot) | 11:36 |
| nuclear_eclipse | paul_: other than the current breakdown between freenode and mibbit, why do you want to replace it? | 12:16 |
| nuclear_eclipse | dhx_m: how long will you be around? | 12:16 |
| dhx_m | nuclear_eclipse: 2hrs? | 12:16 |
| nuclear_eclipse | k | 12:17 |
| nuclear_eclipse | what time is it there, on the ass-end of the world? :P | 12:17 |
| dhx_m | lol | 12:19 |
| dhx_m | just after 10PM | 12:19 |
| nuclear_eclipse | oh, lol | 12:19 |
| nuclear_eclipse | I *just* woke up at 8AM :P | 12:19 |
| dhx_m | haha | 12:20 |
| nuclear_eclipse | you silly ausies | 12:20 |
| dhx_m | we're ahead of you :p | 12:21 |
| dhx_m | btw if we ever have an award for worst configuration variable name.... I'd like to nominate show_product_version and show_version | 12:21 |
| nuclear_eclipse | I talked with victor and siebrand yesterday about the succinct buttons, and they both agree with the idea, and siebrand will coordinate getting translatinos handled between the RC and final releases | 12:22 |
| dhx_m | nuclear_eclipse: ok sounds good by me... although I don't really see the need to translate this one | 12:22 |
| dhx_m | nuclear_eclipse: unless of course the translators are under direction to pick whatever single word in their language describes the button best | 12:23 |
| nuclear_eclipse | so if you could get 10274 rebased onto master, and push all three of your remaining proposals before you head out, that'd be great | 12:23 |
| dhx_m | nuclear_eclipse: yep, rewriting 10274 as we speak... almost done actually | 12:23 |
| nuclear_eclipse | ok | 12:23 |
| dhx_m | nuclear_eclipse: I'll fix the others too (can't remember what they are heh) in a moment | 12:24 |
| dhx_m | ah, not much else to do (yay for no work!) | 12:24 |
| nuclear_eclipse | I'm assuming that siebrand will probably just advise translators to remove any rendundant verbage if possible; but you're right, it's not really pressing that the translations happen, etc | 12:24 |
| dhx_m | yep | 12:25 |
| dhx_m | I'm surprised with how many translators we have | 12:25 |
| nuclear_eclipse | once that stuff gets committed, I'm going to work on rolling an official RC release sometime today in my free time | 12:25 |
| dhx_m | those new stick/unstick strings got translated in no time at all | 12:25 |
| nuclear_eclipse | translatewiki does wonders :) | 12:25 |
| dhx_m | exciting! :) | 12:25 |
| dhx_m | I was thinking about extending that email notification thing to also let users know when their access level to a project has changed | 12:26 |
| dhx_m | any thoughts on that... or leave it until later? | 12:26 |
| nuclear_eclipse | the key is that translatewiki is used for MediaWiki translations, so we're piggy-backing off their hard work :P | 12:26 |
| dhx_m | is that for Mantis only? | 12:26 |
| nuclear_eclipse | I say leave it the way it is for now | 12:26 |
| dhx_m | yep | 12:26 |
| nuclear_eclipse | I think they do some other projects too, but (iirc) siebrand is actually a member of the mediawiki translation team as well, which helps a lot :P | 12:27 |
| nuclear_eclipse | anywho, I gotta go grab some breakfast, bbiab | 12:27 |
| dhx_m | haha, was wondering why we had such an A grade translation team :) | 12:27 |
| dhx_m | ok | 12:28 |
| dhx_m | version_get_all_rows isn't very scalable (having to search through all versions in a project doesn't seem nice to me) | 12:32 |
| dhx_m | I might see about making a better query in version_api for that | 12:35 |
| dhx_m | or maybe not... seems like I'd have to bother with caching, etc | 12:36 |
| paul_ | right, what to do this weekend | 12:37 |
| dhx_m | break Mantis once 1.2 is branched? :p | 12:37 |
| paul_ | well | 12:37 |
| paul_ | I have 3 choices | 12:37 |
| dhx_m | we've had bugobjects... now what about ... htmlobjects | 12:37 |
| paul_ | no | 12:37 |
| dhx_m | or something else :p | 12:37 |
| paul_ | I want to play with utf8/sql stuff | 12:37 |
| dhx_m | that sounds like a good cause | 12:38 |
| paul_ | i.e. test against the various db's | 12:38 |
| dhx_m | yep | 12:38 |
| dhx_m | ahhh @ count( version_get_all_rows (project_id ) ) | 12:39 |
| dhx_m | inefficiency to the extreme :p | 12:39 |
| paul_ | ? | 12:40 |
| dhx_m | looping through every version for a project, and then placing those results from cache into an array | 12:40 |
| dhx_m | then returning it... then counting the number of elements in the array | 12:40 |
| dhx_m | as opposed to just having a function like count_versions_in_project that does count() on the global cache? | 12:41 |
| paul_ | where is that? | 12:41 |
| dhx_m | grep -Rn version_get_all_rows | 12:41 |
| dhx_m | a few places | 12:42 |
| paul_ | well first off | 12:42 |
| paul_ | it should be !empty ? | 12:42 |
| dhx_m | probably | 12:42 |
| dhx_m | at the moment it is count (...) > 0 | 12:42 |
| paul_ | right so we need to further review use of count() accross mantis | 12:43 |
| dhx_m | sounds like a good idea to me | 12:43 |
| dhx_m | it is generally a sign of inefficiency I guess | 12:43 |
| paul_ | I can imagine john cringing already | 12:43 |
| dhx_m | hah | 12:43 |
| dhx_m | he's having breakfast... if we type fast enough he won't have time to read all the scrollback, and might not notice :p | 12:44 |
| paul_ | ok! | 12:44 |
| paul_ | more importantly | 12:44 |
| paul_ | we should fix nusoap class | 12:44 |
| dhx_m | what's up with it? | 12:44 |
| paul_ | what i'd like to do | 12:44 |
| * dhx_m wiki's SOAP | 12:44 | |
| paul_ | on a seperate note | 12:45 |
| paul_ | is have | 12:45 |
| paul_ | mantis repo | 12:45 |
| paul_ | then seperate repo's for the libraries | 12:45 |
| paul_ | but then when you checkout the mantis repo | 12:45 |
| paul_ | it does what hg/svn do and include the 'library' repos | 12:45 |
| dhx_m | splitting into other repo's sounds like a good idea to me | 12:45 |
| dhx_m | it can do that? :) | 12:45 |
| dhx_m | how does it know where to store them locally? | 12:46 |
| dhx_m | when you do the initial pull, does it create it's own directory structure (1 directory per project) | 12:46 |
| dhx_m | as opposed to letting you choose what name you want to give the local repo directory? | 12:46 |
| nuclear_eclipse | dhx_m: you have to realize that i have an android phone with ssh access to my server, where i have my irc connection running in screen -- I can read irc from anywhere at any time ;) | 12:47 |
| dhx_m | nuclear_eclipse: geeky! :p | 12:47 |
| nuclear_eclipse | that's me := | 12:47 |
| paul_ | nuclear_eclipse: git can do what I describe yes? | 12:48 |
| nuclear_eclipse | and btw paul_, add submodule repos and i will kill you... | 12:48 |
| paul_ | why? :) | 12:48 |
| nuclear_eclipse | git submoduels is *not* something i want to use.... | 12:49 |
| nuclear_eclipse | afk | 12:49 |
| paul_ | :) | 12:49 |
| paul_ | dhx_m: so yea submodules | 12:51 |
| dhx_m | lol | 12:51 |
| paul_ | if it's anything like what i've used with hg/svn | 12:51 |
| paul_ | you'd have a repo for 'adodb' | 12:51 |
| paul_ | then when you clone the 'mantis' repo | 12:52 |
| paul_ | it clones mantis | 12:52 |
| paul_ | and automatically clones in 'adodb' | 12:52 |
| dhx_m | and that would appear as mantis-git/adodb ? | 12:52 |
| dhx_m | or as mantis-git/* and separately as adodb/* | 12:52 |
| paul_ | well, for example | 12:53 |
| paul_ | when you cvs co http://cvs.php.net/viewvc.cgi/php-src/ | 12:53 |
| paul_ | you also get a Zend dir | 12:53 |
| paul_ | that contains http://cvs.php.net/viewvc.cgi/ZendEngine2/ | 12:53 |
| paul_ | lunch brb | 12:53 |
| dhx_m | hmm interesting | 12:55 |
| dhx_m | ok | 12:55 |
| dhx_m | paul_: the new calendar widget seems broken (firefox 3.5) | 13:10 |
| dhx_m | paul_: it works... but the popup has no background (it is just text-on-text) | 13:12 |
| dhx_m | paul_: being able to drag it around is cool though :) | 13:13 |
| paul_ | it's not new? | 13:13 |
| paul_ | are you on latest trunk? | 13:13 |
| dhx_m | paul_: oh... I thought it was sorry | 13:13 |
| dhx_m | yep | 13:13 |
| dhx_m | maybe I'm thinking of some other javascript thingy that was upgraded recently | 13:14 |
| dhx_m | I'll double check | 13:14 |
| paul_ | right | 13:19 |
| paul_ | what db's do we spport | 13:19 |
| paul_ | a) mysql | 13:19 |
| dhx_m | all of them? :p | 13:19 |
| paul_ | b) postgres | 13:19 |
| dhx_m | sqlite I heard mention of | 13:21 |
| dhx_m | as in... not working or currently supported... but a TODO item | 13:21 |
| paul_ | db2 | 13:21 |
| paul_ | mssql | 13:23 |
| paul_ | oracle | 13:23 |
| * paul_ ponders | 13:28 | |
| paul_ | I think first i might tidy my desk | 13:28 |
| dhx_m | haha | 13:32 |
| dhx_m | version_add doesn't add the new version to cache... is that OK? | 13:34 |
| nuclear_eclipse | yeah | 13:35 |
| paul_ | well the cache(s) are just there to reduce some DB calls | 13:35 |
| paul_ | php6 doesn't work :( | 13:35 |
| nuclear_eclipse | generally, versiod_add will be one of the last calls on a page, so it would be "useless" to cache that inof | 13:35 |
| dhx_m | but if you did a version_add followed by version_get_all_rows, I don't think you'd get the new version in the output array? | 13:35 |
| nuclear_eclipse | paul_: stop fretting about php6 | 13:35 |
| * paul_ isn't talking about mantis ;P | 13:36 | |
| dhx_m | yeah I think it is a bug that wouldn't really ever get noticed | 13:36 |
| paul_ | php5.3 next week :) | 13:36 |
| dhx_m | .. good? :) | 13:36 |
| paul_ | well, imo ew could drop php5.1 support once php5.2+5.3 are stable :P | 13:37 |
| * paul_ shrugs | 13:37 | |
| dhx_m | yeah it's not like any of the users will notice... they're still using mantis 0.6 on PHP 3 | 13:38 |
| paul_ | there wasn't a a mantis 0.6, first release of mantis was 0.9.0 in december 200 | 13:40 |
| nuclear_eclipse | mantisbt.org still runs php5.1 because that's what is in stable RHEL/CentOS distros | 13:40 |
| nuclear_eclipse | in these days of linux distros, the obvious php version number doesn't really mean much, because most distro apply patches from later versions to a known-stable older version | 13:41 |
| paul_ | which isn't necessarily helpful :) | 13:42 |
| nuclear_eclipse | I would agree with that | 13:42 |
| dhx_m | PHP package maintainers must have a lot of time | 13:42 |
| dhx_m | or alternatively, it might explain why it takes so long to see new releases :p | 13:43 |
| nuclear_eclipse | I'm just pointing out that we shouldn't drop "support" for php 5.1 until it's no longer used by a lot of people | 13:43 |
| paul_ | the 'stats' I used to look at | 13:43 |
| paul_ | seem to have stopped getting updated monthly ;/ | 13:43 |
| dhx_m | I think CentOS 5 is a good benchmark for what the oldest version of PHP in use is :p | 13:43 |
| paul_ | http://www.nexen.net/images/stories/phpversion/200810/evolution.milieu.png | 13:44 |
| paul_ | basically says no one uses 5.1 | 13:44 |
| paul_ | http://www.nexen.net/images/stories/phpversion/200810/mineure5.en.png | 13:45 |
| paul_ | as of 0ct2008, ~15% of php5 was 5.1 | 13:45 |
| paul_ | (wiht a 50-50 php4/5 split) | 13:46 |
| nuclear_eclipse | that's still a good percentage... | 13:46 |
| paul_ | as of 6 months ago :P | 13:46 |
| dhx_m | is it even supported by PHP? | 13:46 |
| nuclear_eclipse | yep | 13:46 |
| nuclear_eclipse | PHP4 only recently got EOL'i | 13:46 |
| dhx_m | but PHP 4 is a major version as opposed to 5.1 which is minor | 13:48 |
| paul_ | i'd say php dont support 5.1 | 13:48 |
| dhx_m | I can't find any info on 5.1 (where to download, etc) on the PHP website | 13:48 |
| paul_ | dhx_m: http://www.php.net/releases/ | 13:48 |
| nuclear_eclipse | anyways, I think the important part is that major distros still support it | 13:48 |
| dhx_m | paul_: those are unsupported though :) | 13:49 |
| dhx_m | nuclear_eclipse: fair enough, but which distros are we officially going to support? | 13:49 |
| paul_ | dhx_m: well | 13:50 |
| paul_ | I dont have 5.1 installed on any of my boxes :) | 13:50 |
| nuclear_eclipse | well, at least Debian, Fedora, and RHEL/CentOS | 13:50 |
| dhx_m | also the 1.3/trunk would be for a few months time | 13:50 |
| * paul_ has php5.2.8, php5.3 and phpsnapshots of dev stuff | 13:50 | |
| paul_ | as my versions | 13:50 |
| dhx_m | afaik it is easy to get 5.2/5.3 packages for CentOS | 13:51 |
| dhx_m | not sure if they're supported by the CentOS people though | 13:51 |
| dhx_m | I know RHEL wouldn't support it | 13:51 |
| nuclear_eclipse | I personally have 5.2.x on all of my machines/servers, but considering even mantisbt.org is running CentOS 5 with php5.1, I think it makes sense to keep upport for it | 13:51 |
| dhx_m | (officially) | 13:51 |
| paul_ | I'm just wondering who's supporting php5.1 :) | 13:52 |
| paul_ | given we've just concluded that all core dev's seem to be running 5.2, or 5.3 | 13:52 |
| dhx_m | btw RHEL is officially supported for 7 years after release | 13:53 |
| paul_ | dhx_m: they must have fun :) | 13:53 |
| dhx_m | paul_: yeah I think they probably stick with php 5.1 for 7 years from the release of RHEL 5... and do their own patching? | 13:55 |
| dhx_m | my theory is that anyone wanting that sort of "stability" probably wouldn't want to be using a new version of Mantis either? | 13:56 |
| nuclear_eclipse | lies :P | 13:57 |
| paul_ | dhx_m: well | 13:57 |
| nuclear_eclipse | my pov is: what do we actually *gain* by dropping support for 5.1? | 13:57 |
| paul_ | nothing | 13:57 |
| dhx_m | nuclear_eclipse: let's say that PHP X.Y comes out that supports Unicode properly... do we wait 7 years to use that in Mantis? :p | 13:57 |
| paul_ | until we find out we've commited some code tht doesn't work in 5.1 :) | 13:58 |
| paul_ | dhx_m: nope! | 13:58 |
| paul_ | dhx_m: I tend to test/fix mantis for new php releases before the new release happens | 13:58 |
| dhx_m | ifdefs! | 13:58 |
| dhx_m | yep | 13:58 |
| dhx_m | that is a good point about how many developers are using 5.1 | 13:58 |
| nuclear_eclipse | well, when php6 comes out, that' a completely different story -- we would actually *gain* something by supporting only php6, that being native utf8 support | 13:59 |
| dhx_m | yep | 13:59 |
| dhx_m | is PHP 6 vapourware or is it real? | 13:59 |
| nuclear_eclipse | but between 5.1 and 5.2, there's really nothing that we would actually gain by dropping support for 5.1 afaik | 13:59 |
| dhx_m | yeah I think the few minor features we could use like: | 14:00 |
| nuclear_eclipse | but like paul said, if we commit something and later find out it doesn't work in php5p1 for some weird reason, then we can reconsider what we want to support | 14:00 |
| dhx_m | "Hooks for tracking file upload progress were introduced." | 14:00 |
| paul_ | dhx_m: http://wiki.pooteeweet.org/PhP60 http://cvs.php.net/viewvc.cgi/php-src/unicode-progress.txt?view=markup http://www.php.net/~scoates/unicode/render_func_data.php | 14:00 |
| dhx_m | "Introduced DateTime and DateTimeZone objects with methods to manipulate date/time information." | 14:01 |
| dhx_m | etc | 14:01 |
| dhx_m | could be worked around fairly easily? | 14:01 |
| nuclear_eclipse | anywho, dhx_m, commit your stuffs! | 14:02 |
| dhx_m | oh yeah, forgot :p | 14:02 |
| paul_ | dhx_m: I believe the *main* changes in 51->52 were memory/stability/performance improvements | 14:02 |
| dhx_m | paul_: looks like PHP 6 is a way off then | 14:03 |
| paul_ | dhx_m: well, will find out in a couple of weeks I reckon | 14:03 |
| paul_ | i.e. after 5.3 | 14:03 |
| nuclear_eclipse | afterwards, 1) backup and update live site to latest master and test, 2) prepare final release builds, 3) release! | 14:03 |
| dhx_m | yep | 14:03 |
| paul_ | they'll need to decide whether next release is 5.4 or 6.0 | 14:03 |
| paul_ | nuclear_eclipse: does all the date stuff work correctly now? | 14:04 |
| dhx_m | ah HTML 5 is going to be a real pain in the arse | 14:04 |
| paul_ | mm? | 14:04 |
| dhx_m | "web worker threads" are the worst thing I've ever tried | 14:05 |
| paul_ | wtf? | 14:05 |
| dhx_m | multithreaded javascript | 14:05 |
| paul_ | if(( null === $p_truncated_length ) || ( utf8_strlen( $c_field_name ) != $p_truncated_length ) ) { | 14:05 |
| paul_ | $query = "SELECT id FROM $t_custom_field_table WHERE name = '$c_field_name'"; | 14:05 |
| paul_ | } else { | 14:05 |
| paul_ | /** @todo This is to handle the case where we only have a truncated part of the name. This happens in the case where | 14:05 |
| paul_ | * we are getting the custom field name from the history logs, since history is 32 and custom field name is 64. | 14:05 |
| paul_ | * This fix will handle entries already in the database, future entries should be handled by making the field name max lengths match. | 14:05 |
| paul_ | */ | 14:05 |
| paul_ | btw | 14:05 |
| paul_ | can we fix that? ;/ | 14:05 |
| nuclear_eclipse | I dunno | 14:05 |
| nuclear_eclipse | but dates seem to be working completely now | 14:06 |
| paul_ | /** @todo This is to handle the case where we only have a truncated part of the name. This happens in the case where | 14:06 |
| paul_ | * we are getting the custom field name from the history logs, since history is 32 and custom field name is 64. | 14:06 |
| dhx_m | "web worker threads" = "consume 100% CPU and make the browser completely unusable" in Firefox 3.5 | 14:06 |
| paul_ | describe mantis_bug_history_table; | 14:06 |
| paul_ | | field_name | varchar(64) | NO | | NULL | | | 14:06 |
| paul_ | | old_value | varchar(255) | NO | | NULL | | | 14:06 |
| paul_ | | new_value | varchar(255) | NO | | NULL | | | 14:06 |
| nuclear_eclipse | yeah, I think that may no longer be a problem :P | 14:07 |
| paul_ | are people likely to have fieldlengths>32 ? | 14:08 |
| paul_ | it's tempting to just drop that code | 14:09 |
| dhx_m | I'm confused as to why changing $g_roadmap_update_threshold = 999; | 14:09 |
| dhx_m | would prevent the target version showing in the advanced view page | 14:09 |
| paul_ | because admin = 100? | 14:09 |
| dhx_m | but update_threshold isn't used in bug_view_advanced_page | 14:10 |
| dhx_m | I have: | 14:10 |
| dhx_m | $t_can_view_roadmap = access_has_bug_level( config_get( 'roadmap_view_threshold' ), $f_bug_id ); | 14:10 |
| dhx_m | as the check | 14:10 |
| paul_ | nod | 14:10 |
| * paul_ thinks | 14:11 | |
| paul_ | did I want john to fix something | 14:11 |
| dhx_m | paul_: with bugobjects, what do you think of adding checks inside __get | 14:13 |
| paul_ | to check what? | 14:13 |
| dhx_m | paul_: for instance, roadmap_view_threshold and show_product_version ? | 14:14 |
| paul_ | I'm not sure | 14:14 |
| dhx_m | roadmap_update_threshold has me confused | 14:15 |
| paul_ | (we'd have to trigger an error if someone tried to request it that shouldn't - (which could be a good thing), but we'd still need to check for it in the code | 14:15 |
| dhx_m | yep | 14:16 |
| dhx_m | well what I was thinking towards was try/catch errors? | 14:16 |
| dhx_m | so we have all our access checks in one place | 14:17 |
| dhx_m | instead of scattered all over the codebase | 14:17 |
| dhx_m | it'd be 1000000 times easier to maintain that way | 14:17 |
| dhx_m | (I think) | 14:17 |
| dhx_m | or better yet, maybe it could just follow the lead of the custom fields | 14:18 |
| nuclear_eclipse | me is in favor of waiting until after 1.2 for considering that | 14:18 |
| dhx_m | and have a threshold for viewing and updating | 14:18 |
| dhx_m | oh yeah, not for 1.2 | 14:18 |
| paul_ | didn't we release 1.2 already? | 14:18 |
| * dhx_m gets back to working out why roadmap_update_threshold is messing my code up (when it isn't referenced at all in my checks) | 14:18 | |
| nuclear_eclipse | lol, no | 14:19 |
| paul_ | dhx_m: what yare your commits? | 14:19 |
| dhx_m | paul_: nuclear_eclipse was going to RC it today I gather :) | 14:19 |
| dhx_m | paul_: bug 10274 | 14:19 |
| mantisbot | Bug 10274 - dhx - open - new | 14:19 |
| mantisbot | Display target release date next to target version - http://www.mantisbt.org/bugs/view.php?id=10274 | 14:19 |
| paul_ | dhx_m: I'd hate to try co-ordinate an RC :P | 14:19 |
| dhx_m | bug 10598 | 14:19 |
| mantisbot | Bug 10598 - dhx - open - assigned | 14:19 |
| mantisbot | Shorten HTML button strings shown when viewing/updating a bug - http://www.mantisbt.org/bugs/view.php?id=10598 | 14:19 |
| dhx_m | bug 4772 | 14:19 |
| mantisbot | Bug 4772 - hugopedersen - open - assigned | 14:19 |
| mantisbot | Changing username doesn't advice user - http://www.mantisbt.org/bugs/view.php?id=4772 | 14:19 |
| dhx_m | those 3 | 14:19 |
| dhx_m | and 1 more (:p) which is my missing part from 10330 to handle the effectiveness ratings on the summary page | 14:20 |
| dhx_m | I needed to give it a bit of a cleanup | 14:20 |
| nuclear_eclipse | dhx_m: eta on all that? | 14:20 |
| dhx_m | nuclear_eclipse: very soon... just stumped on this roadmap_update thingy | 14:20 |
| dhx_m | I'm testing these before I commit :) | 14:21 |
| * paul_ starts to code | 14:21 | |
| paul_ | testing? | 14:21 |
| nuclear_eclipse | if you need me to wait for tomorrow to RC, that's fine; just want to know how much I should get into my WoW session :P | 14:21 |
| paul_ | heh | 14:21 |
| paul_ | well | 14:21 |
| dhx_m | actually tomorrow would be good :) | 14:21 |
| paul_ | yay | 14:21 |
| paul_ | I can break stuff today! | 14:21 |
| dhx_m | because I can give it more testing | 14:21 |
| dhx_m | I was planning on going "live" with the current RC | 14:22 |
| dhx_m | after a round of testing | 14:22 |
| nuclear_eclipse | paul_: no breaking anything! | 14:22 |
| dhx_m | so that should find bugs fairly quickly (if there are any) | 14:22 |
| paul_ | why do we have @require_once in instal.php ? | 14:22 |
| nuclear_eclipse | dhx_m: perhaps you'll be waking up before I'll be going to bed? :P | 14:22 |
| dhx_m | nuclear_eclipse: there seems to be a silent failure somewhere when you start setting config values to NOBODY? | 14:23 |
| nuclear_eclipse | dunno | 14:24 |
| nuclear_eclipse | so stop doing that! :P | 14:24 |
| nuclear_eclipse | "doctor, my knee hurts when I bend it backwards..." | 14:24 |
| dhx_m | nuclear_eclipse: I can set $g_roadmap_update_threshold to random numbers like 77 and my administrator can still view the roadmap | 14:24 |
| dhx_m | haha | 14:24 |
| paul_ | nuclear_eclipse: i've concluded that phputf8 classes fail btw :P | 14:25 |
| dhx_m | yet values over my admin threshold will prevent things from showing | 14:25 |
| nuclear_eclipse | fail how? | 14:25 |
| paul_ | try comparing php-compat functions to mb ones for all unicode chars | 14:25 |
| dhx_m | GRRRRRRRRRRRRRRRRRRRRRR | 14:26 |
| dhx_m | ohhh | 14:29 |
| dhx_m | silent error I missed | 14:29 |
| dhx_m | bug_update_advanced_page.php37--gpc_get_int( <string>'bug_id' ) | 14:33 |
| dhx_m | core/gpc_api.php108--call_user_func_array( <string>'gpc_get', <Array> { [0] => 'bug_id' } ) | 14:33 |
| dhx_m | --gpc_get( <string>'bug_id' ) | 14:33 |
| dhx_m | 50--trigger_error( <string>'200', <integer>256 ) | 14:33 |
| dhx_m | no one would have noticed this before, as it only shows up embedded within the raw html output | 14:36 |
| dhx_m | not on a new page | 14:36 |
| dhx_m | A required parameter to this page (bug_id) was not found. | 14:38 |
| paul_ | nuclear_eclipse: .? | 14:39 |
| nuclear_eclipse | hi paul_ | 14:39 |
| paul_ | forms api | 14:39 |
| paul_ | we need to finish | 14:39 |
| nuclear_eclipse | what's not finished? | 14:39 |
| paul_ | it's not used everywhere right? | 14:40 |
| nuclear_eclipse | I prenny sure it is | 14:40 |
| paul_ | we dont use it in account_prefs_reset.php | 14:40 |
| nuclear_eclipse | ok | 14:40 |
| dhx_m | if someone is running a test version of Mantis, can you check the HTML output of bug_update_advanced_page.php | 14:40 |
| paul_ | nuclear_eclipse: nor 35 other locations | 14:41 |
| dhx_m | and see if there is an error showing up in there? | 14:41 |
| paul_ | do we need to add it for these locations? | 14:41 |
| * paul_ thought it was a security thing? | 14:41 | |
| nuclear_eclipse | dhx_m: check leetcode.net/mantis | 14:41 |
| nuclear_eclipse | paul_: it's only needed for any page that actually performs an *action* | 14:41 |
| dhx_m | nuclear_eclipse: thanks | 14:42 |
| paul_ | doesn't bugnote_add preform an action? | 14:42 |
| nuclear_eclipse | if it's just a post/get form to gather information or view something, you don't need to have form security involved | 14:42 |
| nuclear_eclipse | yes, bugnote_add would be an action | 14:42 |
| dhx_m | nuclear_eclipse: hmm can't update any :p | 14:42 |
| paul_ | nuclear_eclipse: could you have a look and see if i'm right | 14:42 |
| nuclear_eclipse | liar | 14:42 |
| paul_ | that we've missed 35 locations | 14:42 |
| nuclear_eclipse | dhx_m: missed the fact that you said *update*_page :P | 14:43 |
| dhx_m | ah :) | 14:43 |
| paul_ | nuclear_eclipse: also | 14:43 |
| paul_ | whos writing release note documentation? | 14:43 |
| nuclear_eclipse | I was planning to | 14:43 |
| paul_ | show me the draft | 14:44 |
| nuclear_eclipse | pushy much? :P | 14:44 |
| paul_ | yep :P | 14:44 |
| paul_ | if we're gonna do an RC not a beta | 14:44 |
| paul_ | :P | 14:44 |
| nuclear_eclipse | paul_: email a list of topics you think should be covered in the notes, just so I'm not missing anything | 14:45 |
| paul_ | to do that i'd need to review the changelog | 14:45 |
| dhx_m | nuclear_eclipse: never mind, I think it is a Firefox 3.5 bug | 14:46 |
| dhx_m | nuclear_eclipse: except the part where the error occurs after HTML output has started? | 14:47 |
| dhx_m | can't reproduce when using links from the command line | 14:47 |
| paul_ | HMM | 14:48 |
| * paul_ edits install.php | 14:49 | |
| * paul_ wants to remove INSERT admin account db upgrade ;/ | 14:51 | |
| nuclear_eclipse | why? | 14:51 |
| paul_ | have users specify password during initial install | 14:51 |
| dhx_m | that's a good idea | 14:51 |
| nuclear_eclipse | yes, but not for 1.2 please | 14:52 |
| paul_ | 1.2rc notes cover 1.1->1.2 or 1.2a3->1.2? | 14:53 |
| paul_ | nuclear_eclipse: well can you branch git then :P | 14:53 |
| nuclear_eclipse | paul_: I'm waiting for dhx_m's fixes before branching 1.2 | 14:54 |
| dhx_m | working on it :) | 14:54 |
| paul_ | branch then merge | 14:54 |
| paul_ | we'll have a bunch of stuff to merge anyway | 14:54 |
| nuclear_eclipse | and I think the RC notes should cover 1.1->1.2 | 14:54 |
| dhx_m | ugly | 14:54 |
| dhx_m | what is left to merge into RC? | 14:54 |
| nuclear_eclipse | paul_: merging wouldn't work if you started committing stuff | 14:55 |
| nuclear_eclipse | would have to start cherry-picking, which I'd rather not do | 14:55 |
| paul_ | we'll end up needing to do that anyway i'm sure | 14:55 |
| nuclear_eclipse | just keep your install thing in a local branch or on mantisfogre until we branch 1.2 | 14:55 |
| paul_ | nuclear_eclipse: I did that last week | 14:55 |
| paul_ | erm | 14:55 |
| paul_ | weekend | 14:55 |
| paul_ | and you saw what happened :P | 14:56 |
| nuclear_eclipse | just make a new branch for every new thing you want to do, and it will make everything *far* easier to handle.... | 14:56 |
| paul_ | nuclear_eclipse: where is irc plugin? | 14:56 |
| paul_ | can you rename it | 14:57 |
| paul_ | :P | 14:57 |
| paul_ | Mibbit->IRC | 14:57 |
| dhx_m | ok big question... | 14:57 |
| paul_ | dhx_m: YES | 14:57 |
| dhx_m | why can users without roadmap_update_threshold permission changed the "Fixed in" field | 14:57 |
| dhx_m | current ehmm... "documentation": threshold for updating roadmap, target_version, etc | 14:58 |
| dhx_m | maybe we'll fix that later | 14:58 |
| nuclear_eclipse | paul_: why not create a different irc plugin if you don't want to use mibbit? | 14:58 |
| dhx_m | seems any fix we do now breaks the traditional meaning of that variable | 14:58 |
| paul_ | as i'm going to make mibbit one of the options | 14:59 |
| paul_ | mibbit|qwebirc | 14:59 |
| paul_ | so it's going to a copy/paste of mibbit one | 14:59 |
| paul_ | and renamed :) | 14:59 |
| paul_ | there's a good chance freenode won't let mibbit back on | 14:59 |
| paul_ | imo | 14:59 |
| nuclear_eclipse | fork it and do the renaming stuffs, and I'll pull it in at some point | 15:00 |
| paul_ | well, we need to rename the repo | 15:00 |
| nuclear_eclipse | ah | 15:00 |
| nuclear_eclipse | right | 15:00 |
| paul_ | as current one is 'mibbit' | 15:00 |
| nuclear_eclipse | I'll put it on my todo list | 15:00 |
| paul_ | I know a reasonable amount of thep olitics behind mibbit | 15:00 |
| nuclear_eclipse | it seems to me the problem is more on freenode's side | 15:01 |
| nuclear_eclipse | they are blaming mibbit for not doing things that he's been trying to get them to let him do for the past few months... | 15:01 |
| dhx_m | also show_product_version option doesn't cover the Build field | 15:01 |
| dhx_m | so you can disable all versions from showing... but the build is still shown | 15:01 |
| nuclear_eclipse | dhx_m: write up a bug on the tracker about those inconsistencies, and we'll see what people say about it? | 15:02 |
| dhx_m | and the roadmap/etc links in the toolbar still show | 15:02 |
| dhx_m | yep will do | 15:02 |
| paul_ | grr | 15:07 |
| paul_ | who broke tests ;/ | 15:07 |
| paul_ | dhx_m: warn me before you push pls | 15:21 |
| dhx_m | paul_: ok | 15:21 |
| dhx_m | paul_: when showing dates, do I have to use a special function now? | 15:37 |
| dhx_m | or is date() ok? | 15:37 |
| dhx_m | I would assume it isn't anymore :) | 15:37 |
| paul_ | date is fine | 15:38 |
| paul_ | iirc | 15:38 |
| dhx_m | so the time zone of the user isn't factored in? | 15:38 |
| dhx_m | or is there a global set_timezone() thingy instead? | 15:38 |
| paul_ | yep | 15:39 |
| paul_ | the later | 15:39 |
| dhx_m | ok thanks :) | 15:39 |
| dhx_m | can I commit then? :D | 15:39 |
| paul_ | go for it | 15:42 |
| dhx_m | ok | 15:42 |
| dhx_m | I hate working in the wrong branch ;/ | 15:50 |
| CIA-17 | Mantisbt: hickseydr * r20c59001f25e / (13 files in 3 dirs): Option to show scheduled version release dates | 15:51 |
| CIA-17 | Mantisbt: hickseydr * rd62c08e3b682 / (5 files in 3 dirs): Notify user if an admin updates their account | 15:57 |
| CIA-17 | Mantisbt: hickseydr * r63e4dbcb6a77 / (5 files in 3 dirs): Merge branch 'manage-user-email-notify' | 15:57 |
| dhx_m | Warning: Illegal offset type in isset or empty in /var/www/localhost/htdocs/mantis-git/core/user_api.php on line 51 | 16:02 |
| dhx_m | Warning: Illegal offset type in /var/www/localhost/htdocs/mantis-git/core/user_api.php on line 75 | 16:02 |
| dhx_m | on manage_user_edit_page | 16:02 |
| dhx_m | ah I see | 16:05 |
| paul_ | you broke trunk? | 16:06 |
| dhx_m | my ADMINISTRATOR thing the other day had an error | 16:06 |
| dhx_m | fixing now | 16:06 |
| paul_ | wow | 16:06 |
| paul_ | dont you test? | 16:06 |
| paul_ | :P | 16:06 |
| dhx_m | (and reviewing my patch from yesterday again) | 16:06 |
| dhx_m | lol | 16:07 |
| dhx_m | I was sending a user object instead of a user id :o | 16:07 |
| CIA-17 | Mantisbt: hickseydr * r70a687a08c69 /manage_user_edit_page.php: Fix invalid argument to user_is_administrator | 16:11 |
| paul_ | dhx_m: your running 3.4? | 16:24 |
| paul_ | 3.5 | 16:24 |
| CIA-17 | Mantisbt: paul * r226eb0fe7c0b /core/classes/MantisEnum.class.php: actually go back to not using is_blank -> this change broke my phpunit test setup. | 16:24 |
| CIA-17 | Mantisbt: paul * rdb4c72a09ef1 / (core.php core/html_api.php core/wiki_api.php): Only load wiki support if wiki is enabled | 16:25 |
| dhx_m | 3.4 of what? :) | 16:25 |
| nuclear_eclipse | paul_: http://git.mantisforge.org/w/irc.git | 16:34 |
| CIA-17 | Mantisbt: hickseydr * re877da171c03 / (3 files in 3 dirs): Custom reporter effectiveness multipliers | 17:48 |
| dhx_m | nuclear_eclipse: everything except the string shortening patch is now applied | 17:49 |
| dhx_m | nuclear_eclipse: so from me, I have nothing left :) | 17:49 |
| nuclear_eclipse | ok | 17:51 |
| nuclear_eclipse | thanks | 17:51 |
| dhx_m | wait | 17:52 |
| dhx_m | I might apply Victor's suggestions | 17:52 |
| dhx_m | does that sound ok? | 17:53 |
| nuclear_eclipse | what suggestions? | 17:56 |
| dhx_m | 1. How about if we use an access level threshold rather than a bool. | 17:57 |
| dhx_m | The default of NOBODY will give same functionality as now. | 17:57 |
| dhx_m | 2. $g_show_scheduled_release_dates -> $g_show_version_dates_threshold. | 17:57 |
| dhx_m | The word "scheduled" only applies for future releases, but you use | 17:57 |
| dhx_m | the same configs for released ones. From other configs, I think we | 17:57 |
| dhx_m | use "version" as the entity and "release" as the action of releasing a | 17:57 |
| dhx_m | version. However, feel free to double check. I just had a quick | 17:57 |
| dhx_m | look. | 17:57 |
| nuclear_eclipse | dhx_m: whatever you think is good | 18:16 |
| dhx_m | ok | 18:16 |
| paul_ | build successful | 18:21 |
| dhx_m | paul_: you need to include wiki_api.php into bug_view_advanced_page now :) | 18:22 |
| dhx_m | possibly others... I haven't checked | 18:23 |
| paul_ | shouldn't need to if it's disabled... | 18:23 |
| dhx_m | well it makes a wiki_enabled() call | 18:23 |
| dhx_m | which is in wiki_api | 18:23 |
| paul_ | huh | 18:23 |
| paul_ | I grepped | 18:23 |
| paul_ | we need an is_enabled('wiki' type function imo | 18:32 |
| paul_ | I probably grepped for wiki_is_enabled not wiki_enabled as we normally use *IS* | 18:32 |
| dhx_m | yep | 18:32 |
| dhx_m | bug_view_page also affected | 18:33 |
| paul_ | so yea | 18:40 |
| paul_ | constant naming of api functions ftw | 18:40 |
| paul_ | feature_is_enabled('news') | 18:41 |
| paul_ | feature_is_enabled('wiki') | 18:41 |
| paul_ | or something :P | 18:41 |
| paul_ | imo | 18:41 |
| paul_ | thoughts? | 18:43 |
| dhx_m | guess so... I take it plugins are included? | 18:43 |
| dhx_m | my only question is why check strings like 'news' when you can check constants instead? | 18:44 |
| dhx_m | ok I'm done for today | 18:48 |
| CIA-17 | Mantisbt: hickseydr * r2c22ce06979c / (5 files in 3 dirs): Change show_scheduled_release_dates to threshold | 18:48 |
| dhx_m | nuclear_eclipse: all good to go, victor's changes are done | 18:48 |
| dhx_m | cya :) | 18:48 |
| paul_ | dhx_m: oi | 18:48 |
| paul_ | dhx_m: did you fix what I broke then? ;p | 18:49 |
| paul_ | or shall I quickly revert that patch | 18:49 |
| paul_ | and rename it | 18:49 |
| dhx_m | paul_: nope... want me to include it? | 18:49 |
| paul_ | well, could be useful | 18:49 |
| dhx_m | ohh | 18:49 |
| dhx_m | I see | 18:49 |
| paul_ | before john does a release ;p | 18:49 |
| dhx_m | I like your patch better | 18:50 |
| paul_ | mm? | 18:50 |
| dhx_m | just do a config_get | 18:50 |
| dhx_m | instead of using the wiki_enabled thing | 18:50 |
| paul_ | k | 18:50 |
| dhx_m | unless | 18:50 |
| dhx_m | wiki_enabled does other checks? | 18:50 |
| paul_ | nope | 18:50 |
| dhx_m | maybe in the future... but I think it needs the overhaul you mentioned | 18:51 |
| dhx_m | so that a new event hook can be created that lets plugins/etc disable certain features | 18:51 |
| nuclear_eclipse | dhx_m: one last thing that I just remembered -- config option adimn_site_threshold needs to be documented still | 18:51 |
| dhx_m | nuclear_eclipse: oh yep, will do that quickly | 18:51 |
| paul_ | I thought we dropped that threshold? | 18:51 |
| nuclear_eclipse | nope, we expanded it | 18:52 |
| paul_ | I thought we concluded that | 18:52 |
| dhx_m | paul_: in case you have multiple admin access levels :p | 18:52 |
| paul_ | ADMINISTRATOR was the only value you could set | 18:52 |
| dhx_m | nope | 18:52 |
| dhx_m | you could have multiple administrators still | 18:53 |
| dhx_m | for cosmetic reasons | 18:53 |
| dhx_m | or potentially other reasons such as controlling enum levels | 18:53 |
| paul_ | well | 18:53 |
| dhx_m | rare cases | 18:53 |
| paul_ | so admin_site_threshold is level you can do an install at ? | 18:53 |
| dhx_m | yes | 18:54 |
| dhx_m | it is essentially what the ADMINISTRATOR constant used to be | 18:54 |
| paul_ | we should add a check to make sure that admin_site_threshold is >= ADMINISTRATOR | 18:54 |
| dhx_m | also it'd be better to use existing functions like user_is_administrator or current_user_is_administrator | 18:54 |
| dhx_m | good point | 18:54 |
| paul_ | or more | 18:55 |
| dhx_m | wait... no | 18:55 |
| dhx_m | someone might unset the ADMINISTRATOR constant completely | 18:55 |
| dhx_m | and define their own | 18:55 |
| dhx_m | so we can't really | 18:55 |
| paul_ | well | 18:56 |
| paul_ | this is why I thought we agreed not to add the threshold | 18:56 |
| paul_ | imo, we *need* to add a check to ensure that only administrators can run installer | 18:56 |
| paul_ | and that users dont set it to anybody by accident (security risk potentially) | 18:56 |
| paul_ | which is why I thought we agreed 'ADMINISTRATOR' was really a defined constant we couldn't change ;/ | 18:57 |
| dhx_m | I don't know why they would...? | 18:57 |
| nuclear_eclipse | say if they configure it wrong, we can't hand-hold them | 18:57 |
| nuclear_eclipse | it's more of a security problem to have attachments in the webroot, and we can't possibly do any proper checks for that... | 18:57 |
| dhx_m | * Threshold at which a user is considered to be a site administrator. | 18:59 |
| dhx_m | * These users have "superuser" access to all aspects of Mantis including | 18:59 |
| dhx_m | * the admin/ directory. WARNING: DO NOT CHANGE THIS VALUE UNLESS YOU | 18:59 |
| dhx_m | * ABSOLUTELY KNOW WHAT YOU'RE DOING! Users at this access level have the | 18:59 |
| dhx_m | * ability to damage your Mantis installation and data within the database. | 18:59 |
| dhx_m | * It is strongly advised you leave this option alone. | 18:59 |
| dhx_m | is that ok for a warning then? | 18:59 |
| nuclear_eclipse | +1 | 18:59 |
| paul_ | dhx_m: your doing the wiki stuff? | 19:09 |
| paul_ | right? | 19:09 |
| dhx_m | uhmm nope :p | 19:12 |
| dhx_m | documenting | 19:12 |
| dhx_m | maybe just revert it until we fix it properly for 1.3? | 19:16 |
| paul_ | go for it | 19:16 |
| CIA-17 | Mantisbt: hickseydr * r37ebe80af11f / (2 files in 2 dirs): Document g_admin_site_command config option | 19:17 |
| CIA-17 | Mantisbt: hickseydr * rff31b8e7f61b / (3 files in 2 dirs): Replace all instances of wiki_enabled | 19:23 |
| dhx_m | ok all done | 19:23 |
| dhx_m | I'm REALLY out of here now :) | 19:23 |
| dhx_m | cya | 19:23 |
| paul_ | good night | 19:23 |
| dhx_m | :) | 19:23 |
| nuclear_eclipse | gn dhx_m | 19:23 |
| CIA-17 | Mantisbt: jreese * rce10860b00c9 /core/constant_inc.php: Version bump. | 21:27 |
| CIA-17 | Mantisbt: hickseydr * r0d1db44e853b /lang/strings_english.txt: Make HTML button strings shorter | 21:31 |
| paul_ | nuclear_eclipse: you've branched? | 21:53 |
| paul_ | The ManTweet plugin isn't currently working because it relies | 22:15 |
| paul_ | on the old-style date handling routines; Victor will need to either | 22:15 |
| paul_ | follow the existing Mantis schema upgrades and such, or do whhat I did | 22:15 |
| paul_ | for the source integration plugins | 22:15 |
| paul_ | surely | 22:15 |
| paul_ | *surely* | 22:15 |
| paul_ | that's a bad idea | 22:15 |
Generated by irclog2html.py
