[00:00] http://pastebin.ubuntu.com/398071 [00:01] hmm maybe i need a full copy of gtk_entry_completion also? [00:01] looks like that calls back into regular gtkentry [00:02] i had just added the parts of that class that were new [00:03] yes, feels like it [00:03] and the new parts didn't use GtkEntry at all from what i could tell [00:03] let me double check [00:03] ok [00:05] i think things would be easier if we would derive from gtkentry [00:05] then that would just work [00:05] i would hope [00:08] how do we go about doing that, have a derived class with all the bits newer than the one in 2.12? [00:08] ccheney`: so try to derive the GtkEntry [00:08] that might just work [00:09] you would drop all fields from EphyGtkEntry and EphyGtkEntryClass [00:09] use GTK_TYPE_ENTRY in G_DEFINE_TYPE rather than GTK_TYPE_WIDGET [00:09] also you would use GtkEntry and GtkEntryClass instead of GtkWidget and GtkWidgetClass in the EphyGtkEntry and EphyGtkEntryClass === Loki_ is now known as Loki [00:09] struct as the only member (the topmost that isnt a pointer) [00:09] then you would use override_property [00:10] rather than install_property for all the properties [00:10] that are from GtkEntry [00:10] and that should be it afaict [00:10] ok i will try that out and see how it goes [00:11] yeah [00:11] might be a bit tricky [00:11] but probably easier than copying all completion and so on [00:11] as then the accessor functions should just call us [00:11] yea [00:12] what did you mean by 19:08 < asac> struct as the only member (the topmost that isnt a pointer) [00:12] ccheney`: search for the EphyGtkEntry and EphyGtkEntryClass struct [00:12] empty them and have a struct inside them of type GtkEntry ? [00:12] the first elemtn is not a pointer [00:12] thats the parent class/object [00:13] ah yea GtkWidget widget; [00:13] thats currently GtkWidget and should be GtkEntry/Class [00:13] oh ok [00:13] and all the other funcs/fields that are already in trhe gtk 2.12 entry [00:13] should be dropped [00:13] ah ok, i thought that was what you meant but wasn't sure and wanted to make sure i didn't do something wrong :) [00:13] and then where we assign the funcs in _class_init [00:13] we have to assign them to GTK_ENTRY_CLASS (klass) [00:13] so then anything in my copied code that refers to old items need to point through to GtkEntry [00:14] rather than EPHY_GTK_ENTRY_CLASS (klass)->set_background (not sure if that function exists) [00:14] ccheney`: give me an example for the last [00:14] looking at it to see if i understand [00:14] e.g. for "refers to old items need to point through to GtkEntry" [00:14] eg widget_class->unmap = ephy_gtk_entry_unmap; [00:14] yeah right. widget_class is still widget_class [00:15] ah ok [00:15] but there probably also is entry_class or something [00:15] eg class->move_cursor = ephy_gtk_entry_move_cursor; [00:15] ccheney`: right so class-> should more or less be GTK_ENTRY_CLASS(class)-> [00:15] yeah [00:16] class is currently ephy_gtk_entry_class_init (EphyGtkEntryClass *class) [00:16] you can make a gtk_entry_class = ... on top [00:16] same as done with widget_class [00:16] so that should be ephy_gtk_entry_class_init (EphyGtkEntryClass *class) [00:16] er stupid return [00:16] ccheney`: that func should still be named the same [00:16] just the class-> ... assignments have to go against the casted entry_class (aka GTK_ENTRY_CLASS (class)) [00:16] ah ok [00:17] ccheney`: i think one function didnt exist in 2.12 ... so that function might need to be added to our ephy class then [00:17] yea [00:18] then code calling it in ephy should use the new class throughout? [00:18] and/or cast back from the new class to GtkEntry if needed, i guess? [00:19] ccheney`: well. we overload the functions [00:20] so things we have changed to ephy_gtk_entry_something [00:20] can usually be reverted to gtk_entry_something [00:20] except that thing that wasnt available in gtk_entry [00:21] ok [00:21] i think i can probably determine how to do it with what you have told me above :) [00:22] probably take a hour or two to get it fully building with everything reverted though [00:22] and its late where you are so if you want i can let you know how the patching goes and you can head off to bed or whatever (being really late in germany) [00:24] ccheney`: maybe keep what you have for now [00:25] so in case this road isnt good we can try someting else ;) [00:27] yea i am going to keep a backup of it [00:29] maybe also push somewhere online for desaster scenario ;) [00:29] ok off ... will check tomorrow [00:29] thanks [00:34] np :) have a good night [05:47] asac. you there? i have a question. [07:15] asac: haven't gotten to a point to upload yet and about to head to bed [07:15] asac: i'm currently looking at what to do for update_cursors in gtkentry for GTK_ENTRY_GET_PRIVATE [07:16] asac: i think probably copying that verbatim from gtkentry.c would work and copying the private struct into the derived class so it can see what the original gtkentry has [07:16] asac: at least the extra functions definitely need access to the private data from gtkentry to work from what i can see [07:16] asac: ttyl :) [11:53] ccheney`: hey [11:54] LLStarks: yews [15:05] hmm. did micah's sdk update ever land? [15:05] or is that staged in mozilla-devscripts? [18:37] asac: hi [18:40] hey [18:45] so i realized last night that i was way too tired to be sanely working on the code [18:45] i think the private members aren't needed in the new code, checking that now [18:45] yea it doesn't appear to use any of them [18:46] ccheney`: what private members are you talking about? [18:46] so i just need to create a ephyentry private struct [18:46] ccheney`: what did you do up to now [18:46] and copy the private ones it uses from the new version of gtk into it [18:46] how does ephyentry class look like? [18:46] asac: i reverted the ephy gtk entry calls in the code and deleted the old gtk entry copy and started copying just the bits over that are needed to derive from it [18:47] i'll pastebinit just a sec [18:47] http://pastebin.com/cww1GymP http://pastebin.com/7WYXZHQ6 [18:49] ccheney`: you dont have any TYPE_DEFINE [18:49] and class_init [18:49] etc. [18:49] all that is gone ;) [18:49] yea need to add that back [18:50] i added things as the compile complained about them [18:50] ccheney`: i dont think you can drop all the other funcs [18:50] my idea was to copy everything [18:50] just change the class_init function [18:50] we already copied everything before though? [18:50] and make the structs and the G_DEFINE_TYPE [18:50] * ccheney` is confused [18:50] ccheney`: you basically need to overload all funcs of GtkEntry [18:51] thats why you need all the impls [18:51] we had yesterday [18:51] and just set them to the parent class in our new class_init [18:51] so just take the code i had before and then change up the class_init, etc ? [18:51] ccheney`: well. if you have luck it works as you did it now [18:52] ccheney`: the original idea was to change G_DEFINE_TYPE ... class_init and the structs [18:52] e.g. from the structs you drop all the funcs that are in GtkEntry [18:52] and derive from GtkEntry [18:52] ah i think i misunderstood that to drop all funcs that are in GtkEntry period :) [18:52] in class_init you just bind all the funcs [18:52] so i redid the files [18:52] that will ensure that gtk_entry::something will call our new ::something [18:53] and instead of install_property in class_init [18:53] we use override_property [18:53] so if you set/get the properties on a GTK_ENTRY casted object it will also call our get/set_property funcs [18:54] so yeah. only drop the declaration from the structs because they are already there if you have GtkEntryClass parent_class; [18:54] in there [18:54] anyway. have to run now [18:54] the instructions should be in there [18:55] will be back in a couple of hours [18:55] ok [22:02] guys, can somebody check https://bugs.launchpad.net/ubuntu/+source/thunderbird/+bug/526290 ? :) [22:02] Ubuntu bug 526290 in thunderbird "Thunderbird 3 hyperlinks broken" [Undecided,Confirmed] [22:23] dupondje: commented [22:25] thunderbird-gnome-support was indeed not installed, tho it doesn't fix the issue it seems [22:25] after install it still asks me what application to open when I click on a link [22:30] bbin 5 [22:31] i'll grab some food also :P [22:39] dupondje: works for me :/ [22:39] dupondje: sure you restarted tbird properly after installing the pkg? [22:39] jl@laptopjl:~$ ps aux |grep thunder [22:39] jl 22662 0.0 0.0 7556 936 pts/1 S+ 23:39 0:00 grep thunder [22:39] its not running :x [22:40] starting it, and it asks for program :( [22:41] dupondje: in $HOME/.thunderbird (somewhere below) there is a compreg.dat [22:41] please paste that [22:42] |grep http or ? :) [22:42] no [22:42] full [22:42] there probably is no http in [22:43] http://paste.ubuntu.com/398497/ [22:51] dupondje: try running sudo touch /usr/lib/thunderbird-3.0.3/.autoreg [22:51] then start tbird [22:51] stop it [22:51] and paste again [22:53] http://paste.ubuntu.com/398502/ [22:53] oh and it works btw :P [23:00] hehe [23:00] right [23:01] dupondje: fix committed ;) [23:01] will be in tomorrows dailies [23:01] or next upload [23:01] :p [23:01] e.g. fix for componetns not getting registered if you install -gnome-support later [23:02] but still you should 'smell' to install the gnome package ? :) [23:03] dupondje: i commented that the other part should be tracked in a separate (wishlist) but [23:03] bug [23:03] e.g. in #4 [23:04] anyway thx for the fix [23:04] so yes. we should do what we do in firefox since lucid [23:04] surely for new users it was quite annoying :) [23:04] problem is we cannot recommend gnome-support [23:04] because of kubuntu [23:05] they complained all the time [23:05] dupondje: there is a nother issue ... in old tbird package [23:05] we had a safety net that called x-www-browser [23:05] if no gnome component was there [23:05] we should consider to resurrect that [23:06] chrisccoulson: ^^ maybe open those two bugs ;) ... 1. gnome integration should work even without -gnome-support ... 2. ensure that x-www-browser is used if no http handler is found through gnome integration :) [23:06] it quite improves usability when you don't need to select your browser :) but we are getting close :P [23:07] both are regression bugs over our tbird2 package [23:07] i think for 2. we need a patch to code iirc [23:07] as it will otherwise always prefer to use x-www-browser [23:07] but lets first see and get those bugs ;) [23:07] dupondje: or you can file those :) [23:07] :) [23:08] let me do that [23:08] 591 Open bugs [23:08] cool :P [23:10] filed bug 543060 [23:10] Launchpad bug 543060 in thunderbird "thunderbird - gnome integration should work even without -gnome-support" [Undecided,New] https://launchpad.net/bugs/543060 [23:12] :) [23:12] :( slow launchpad [23:12] to much thunderbird bugs ;) [23:15] filed bug 543064 about us not using x-www-browser as a safety net. [23:15] Launchpad bug 543064 in thunderbird "ensure that x-www-browser is used if no http handler is found through gnome integration" [High,Triaged] https://launchpad.net/bugs/543064 [23:15] dupondje: well. lots of noise ;) [23:15] those bugs are important enough, so they will get attention now :-P [23:15] feel free to help out on bug triage so we can filter the other important ones and ensure they get fixed ;) [23:16] starting with NEW: would be good ;) ... https://bugs.edge.launchpad.net/ubuntu/+source/thunderbird/+bugs?search=Search&field.status=New [23:16] dupondje: ^^ [23:16] enjoy ;) [23:16] i'll look :) btw, there is another bug, that thunderbird doesn't move mails when marking it as junk [23:17] but its set in the setting it should move them :s [23:19] that doesnt feel like an ubuntu specific bug. probably best filed upstream [23:19] dupondje: do you see any errors in error console? [23:20] I checked, nope :( [23:26] thx for the bugs btw :) [23:28] 5 bugs as incomplete :p [23:28] 1 as invalid [23:28] lets kill some more [23:31] good evening everyone [23:32] hi :) [23:32] asac left you some work ;) [23:32] yeah, i noticed ;) [23:33] i've got to disappear again. it seems my laptop drive has corrupted itself at some point today, so i have to go and recover [23:39] I use Mint-Gloria (Mozilla Thunderbird). Which ever email I try to open, I always (hapend from5 days) get it: [23:39] rofl @ stupid bugreports [23:42] I user a Thunderbird mail server ... [23:42] Thunderbird mail server [23:42] thats new :)