*** Joins: dregad (~dregad@155.250.128.35) | 00:10 | |
GitHub40 | [mantisbt] dregad pushed 2 new commits to master-1.2.x: https://github.com/mantisbt/mantisbt/compare/218c5129e0b9...9db97b999a67 | 01:12 |
---|---|---|
GitHub40 | [mantisbt/master-1.2.x] Add event to allow plugins access to the bug change status form - Daryn Warriner | 01:12 |
GitHub40 | [mantisbt/master-1.2.x] Renaming EVENT_BUG_CHANGE_STATUS_FORM to EVENT_UPDATE_BUG_STATUS_FORM - Damien Regad | 01:12 |
GitHub10 | [mantisbt] dregad pushed 1 new commit to master: https://github.com/mantisbt/mantisbt/commit/6bbaa7b9b8282b2bd62d7772792f2763db0c329a | 01:12 |
GitHub10 | [mantisbt/master] Renaming EVENT_BUG_CHANGE_STATUS_FORM to EVENT_UPDATE_BUG_STATUS_FORM - Damien Regad | 01:12 |
*** Joins: kirillka (~Miranda@195.242.142.17) | 02:19 | |
*** Joins: Paul_46 (~IceChat09@cpc1-enfi15-2-0-cust580.hari.cable.virginmedia.com) | 02:57 | |
GitHub49 | [mantisbt] dregad pushed 2 new commits to master-1.2.x: https://github.com/mantisbt/mantisbt/compare/9db97b999a67...eb026965bf60 | 06:04 |
GitHub49 | [mantisbt/master-1.2.x] Cosmetic changes and comments in admin/check.php - Damien Regad | 06:04 |
GitHub49 | [mantisbt/master-1.2.x] Fix #14552: Update admin checks to make sure ADOdb extension is not used - Damien Regad | 06:04 |
*** Quits: kirillka (~Miranda@195.242.142.17) (Quit: kirillka) | 07:03 | |
*** Joins: renaud (5043b076@gateway/web/freenode/ip.80.67.176.118) | 07:47 | |
dregad | hi Paul_46 | 08:20 |
dregad | was trying to replace a call to db_query() by db_query_bound() in file_add() function | 08:21 |
Paul_46 | heh | 08:21 |
Paul_46 | don't | 08:21 |
Paul_46 | the reason that file_add and the other place use db_query was deliberate | 08:21 |
dregad | because of line ending encoding for text files | 08:22 |
Paul_46 | and kinda comes down to how adodb handles binary files | 08:22 |
Paul_46 | ? | 08:22 |
dregad | well while testing, i realized that uploaded text files with new functions are not dealt with properly | 08:22 |
dregad | i.e. with db_query_bound, it's a single line with \n chars | 08:22 |
dregad | wherease with db_query it's interpreted with line feeds as expected | 08:23 |
Paul_46 | we need to get off this db layer | 08:23 |
Paul_46 | it's stupid :P | 08:24 |
dregad | stuck with it for now :( | 08:24 |
Paul_46 | or well, it's stupid fixing stuff atm | 08:24 |
Paul_46 | what concern's me is whether it's possible to do a release of any branch | 08:24 |
Paul_46 | without breaking something else :) | 08:24 |
dregad | well using db_query is preventing oracle users from uploading files due to limitation of literal SQL to 2000 chars | 08:25 |
Paul_46 | really | 08:25 |
dregad | yep | 08:25 |
Paul_46 | we should be saying that mysql is only platform that really works | 08:25 |
dregad | ORA-01704 | 08:25 |
Paul_46 | as atm, our 'fix' for the fact that pgsql/oracle are broken in 1.2 is to replace adodb | 08:26 |
Paul_46 | but we are also trying to 'fix' adodb | 08:26 |
Paul_46 | so the question is | 08:26 |
Paul_46 | to ensure future proofing whether we support upgrades from 1.2 to 'next' on oracle etc [which we know are completely broken] | 08:26 |
Paul_46 | i'd personally be inclined not to allow automatic updates from mssql/pgsql/oracle etc from 1.2 to 2.0 | 08:27 |
Paul_46 | and ahve users contact us | 08:27 |
Paul_46 | so we can ensure with them that there database schema matches what we expect it to be in oracle | 08:27 |
Paul_46 | i.e. i'm 'concerned' that we may do a release that fixes the non-mysql db's that we've broken | 08:28 |
Paul_46 | users might see that in a set of release notes and start using them | 08:28 |
Paul_46 | then we might do an update that changes db engine, pick the data types that we want to use for engines, and then find we've got bugs later on due to users having done their own thing | 08:29 |
Paul_46 | a case in point, I think I currently run mantis at work with ODBC | 08:29 |
Paul_46 | and MSSQL | 08:29 |
Paul_46 | #mantisbt-help is a case in point atm of what i'm saying :) | 08:30 |
dregad | re: "and ahve users contact us" - like we have the resources to cope with this kind of support ::rolls eyes:: | 08:32 |
dregad | as broken as adodb may be, we're stuck with it for a while | 08:32 |
Paul_46 | given the fact that it's broken | 08:32 |
Paul_46 | pgsql/mssql etc dont work atm | 08:33 |
dregad | so i've been pushing all our fixes upstream | 08:33 |
dregad | (the ones I could find anyway) | 08:33 |
Paul_46 | so I dont think we are gonna have lots of people using it | 08:33 |
dregad | most of them have been included | 08:33 |
dregad | hence my pushing to upgrade from 5.10 to latest | 08:33 |
dregad | which should fix out of the box a lot of pgsql / mssql / oracle issues | 08:34 |
dregad | and let 1.2.x work +/- OK as we bundle it | 08:34 |
dregad | now if I can't use bind vars to upload files, that's kind of a show stopper | 08:34 |
Paul_46 | tbh, i'd rather you spent some time seeing if it's feasible to make a oracle engine for the db layer we had in next | 08:35 |
Paul_46 | atm, i'd probably propose dropping oracle/db2 support for mantis and just keeping mssql/pgsql/mysql and maybe adding a 'development' sqlite layer for non-production use | 08:36 |
Paul_46 | as I've never managed to get oracle/db2 setup for development | 08:36 |
dregad | no experience with db2, but oracle is almost good to go now | 08:36 |
Paul_46 | and especailly with db2, the core dev's dont have access to the db engines | 08:36 |
dregad | https://github.com/dregad/mantisbt/tree/oracle | 08:37 |
dregad | there are a few ppl using that | 08:37 |
dregad | so it'd be nice to pull it in at some point | 08:38 |
dregad | so, care to explain why uploading files works with db_query and not with db_query_bound ? | 08:39 |
Paul_46 | because db_Query has db_prepare_binary_file to encode the file | 08:40 |
Paul_46 | and in mssql we did something special iirc | 08:40 |
Paul_46 | $content = unpack( "H*hex", $p_string ); | 08:41 |
Paul_46 | return '0x' . $content['hex']; | 08:41 |
Paul_46 | is what we do for mssql for instance | 08:41 |
Paul_46 | however, there's no reason that the db insert stuff should change the line feeds | 08:41 |
dregad | see http://www.mantisbt.org/bugs/view.php?id=14563 the screenshot illustrates the problem | 08:51 |
dregad | btw this also screws up png files | 08:54 |
dregad | (that was on mysql) | 08:57 |
dregad | well if you have any ideas or thoughts, it would be appreciated | 09:16 |
dregad | going home now, may login again later - or tomorrow | 09:16 |
*** Quits: dregad (~dregad@155.250.128.35) (Quit: Ex-Chat) | 09:18 | |
*** Quits: giallu (~giallu@fedora/giallu) (Ping timeout: 246 seconds) | 09:35 | |
*** Quits: renaud (5043b076@gateway/web/freenode/ip.80.67.176.118) (Ping timeout: 245 seconds) | 12:20 | |
*** Quits: sdfjkljkdfsljkl (~sdfjkljkd@static.96.23.63.178.clients.your-server.de) (Remote host closed the connection) | 17:00 | |
*** Joins: sdfjkljkdfsljkl (~sdfjkljkd@static.96.23.63.178.clients.your-server.de) | 17:00 | |
*** Quits: Paul_46 (~IceChat09@cpc1-enfi15-2-0-cust580.hari.cable.virginmedia.com) (Quit: If at first you don't succeed, skydiving is not for you) | 17:20 | |
*** Joins: dregad (~dregad@155.250.128.35) | 23:52 |
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!