Meeting on irc.gnome.org:#gtk-devel Meeting started March 15 2004 17:09 EST (22:09 UTC) In attendance: Jonathan Blandford (jrb), Anders Carlson (andersca), Matthias Clasen (maclas), Tim Janik (rambokid), Tor Lillqvist (tml), Mark McLoughlin (markmc), Federico Mena Quintero (federico), Soeren Sandmann (ssp), Manish Singh (yosh), Owen Taylor (owen) OK, status is that GtkFileChooser is probably still a little rocky, but seems to be working a lot better let's all buy jrb a big present Everything else is, to my knowledge, basically ready to go federico: Some sleep? * jrb would settle for a nice mug of coffee * maclas goes for a mug of coffee fyi, I'm moving gtk.org to a new machine, probably tonight or tomorrow but you shouldn't notice anything beyond a changed ssh host key yosh: Hmm. OK. Timing collision :-) owen: so, one thing that maclas pointed out to me what I didn't think of before, is that I had to put in the no_pop_down hack in order to prevent selecting a row from immediately popping up the new options owen: is the release going to happen today or tomorrow? yosh: I'm guessing it's going to be tomorrow afternoon judging how long it will take for me to finish wirting up release documentaiton owen: that is pretty inconvenient for anyone using the custom entrycompletion stuff; can we change that behavior in a stable release? owen: the candidate fix would be to block the ::changed handler during the emission of ::match-selected owen: is anyone else but you going to be involved with mucking with files on the server? yosh: I can push it off a little bit if necessary to make sure that thenew machine has a bit of time to settle before we announce the release yosh: No, I'm the only guy who regularly does GTK+ web/ftp stuff Hey rambaby hey ok, then we should be able to coordinate it without much trouble I think yosh: My basic concern is that switching to a new machine usually reveals something unexpected... anyways just keep me posted on what you are doing when and how it goes * federico is away: I'm busy jrb: OK, I'm not sure I undersand this problem * federico is back (gone 00:00:01) owen: well, all the gtk stuff will be prepped tonight, except for possibly mail argh owen: but the dns won't change till tomorrow owen: if you modify the entry content's from your ::match-selected handler, you're triggering the ::changed signal from the entry, which causes the list to pop up again... so if you could check it out tonight/tomorrow morning that should work yosh: Are you going to sync over contents immediately when changing the DNS, or is there some "don't touch anything or it will be lost" period? owen: comment out the return on line 325: if (chooser_entry->no_pop_down) return FALSE owen: there will be a 5 min window of don't touch anything, for the sync to actually happen when dns changes jrb: Are you and maclas talking about the same issue, or different ones? but the tree will be prepped tonight, so all it will be is a data sync owen: same owen: same maclas: First should hte list be popping up at all if the only completion is an exact match? I guess we don't have an API to distinguish exact matches :-( yosh: you're switching wilber over to the new advogato machine? yep yosh: OK, Im not sure what you mean by "if you could check it out tonight/tomorrow morning that should work" * rambaby urg, hope my dns scripts don't break while i'm away from home yosh: But when I get close to getting ready to want to release, I'll get back into touch you and we'll sort things out owen: I'll setup beta.gtk.org with a snapshot of the tree. give it a once over yosh: is there any chance we can delay that by say a wekk or so? maclas: So, basically what you anted to do was to move the g_signal_handler_block (completion->priv->entry, completion->priv->changed_id); maclas: around the signal emission as well? rambaby: eh, it's basically blocking the new gimp website from going live owen: yes. although from looking at gtkentry.c, it blocks the changed handler inside gtk_entry_set_text(), so set_text() won't ever trigger the popup, only insert_text() yosh: sure, and i wanted to see this for a long time. but i'd rather not have potential problems collapse with the 2.4.0 gtk release maclas: Ugh! collide i mean ;) I doubt that will happen. and it's not like the old machine is going to be deconfigured, so it can be switched back in emergency owen: Are you ok with moving the handler_block(), or should we just document that ::match-selected handlers should use set_text(), not insert_text(), to modify the entry ? maclas: Let's move the handler_block ... the fact that set_text is differnet than insert_text is a little odd, and I definitely would prefer not to recommend relying on that owen: ok, I'll do that after the meeting http://bugzilla.gnome.org/show_bug.cgi?id=69616 is probably relevant to how that should be done right OK, any other issues? not an issue really, but I'm not sure if the api docs are in good shape, or if we're missing some new symbols from gtk-sections.txt; I wanted to give them a quick look tonight maclas: For some reason gtk-sections.txt is complaining about msising all the class structures. I'm not sure what changed that caused that owen: well, they *are* missing maclas: I'd really wish I had some time to spend on updating the docs to describe 2.4 changes, fix the building section to be reasonably accurate, etc. But I don't think that's going to happen maclas: One thing I do want to do is put a symlink on widget so http://developer.gnome.org/doc/API/2.4/gtk/ goes to the right place the tutorial on gtk.org is looking pretty dated owen: yes, I don't want to do any time-consuming updates, just a quick check to see if there are missing symbols, markup typos, etc owen: do I still have time to do that tonight ? maclas: My guess is that I'm not going to start distchecking until tomorrow maclas: I don't have a ton of staying power tonight after working all weekend, and the release announcements are going to take a while understandable... * jrb would like to see updates to the docs of GtkEntryCompletion (http://people.redhat.com/~otaylor/tmp/glib-2.4-announce is the glib draft, but still working on the pango/gtk+ and the "all libraries" announcements. I'll mail out pointers when I'm doine) jrb: I've added a bit yesterday, do you still want more? Thanks to all the GLib 2.2 contributors, including: owen: don't forget to change that to a 2.4 :) andersca: Fixed maclas: oh. I didn't notice; lemme look. OK, let see I think what needs fixing on the web site will be: the Download section and putting up the release announcement in News. Anything else? I gues plan/ could use a quick update so that 2.4 is not still listed as in progress owen: if you give Tim-Philipp a proper ü, you might want to give ssp his umlaut as well... maclas: You mean his slash? owen: yes, instead of oe ø ides of march... * maclas didn't know how to type ø We really should start asking people to give their names both transliterated and in the native script, so that release announcements look cooler :-) owen: I think Søren Boll Overgaard would like a slash too, then maclas: alt-gr+o øøøøøøøøøøø !! ctrl-shift f8 aargh, create_folder doesn't use the info cache in gtkfilesystemunix ssp: He'd probably like Overgaard too. I wonder how that got dropped * kris gets all kinds of weird symbols * kris blames his non-utf8 terminal. owen: If you'd like, I can send you a patch for plan/ ssp: Do Søren Boll Overgaard,Henning Nielsen Lund sort under O/L or B/N ? owen: O/L owen: speaking about plan/, we may want to add bugzilla queries to plan/2.6 in the same way as we did for the willfix list. Thats not urgent, of course... * jrb laments the lack of cool characters in his name ssp: OK, that was my guess, but I thought I'd check. just call yourself Jønathan or Jönathan owen: are you sure he is "Henning Nielsen Lund"? "Henning Lund Nielsen" sounds a lot more likely? Ĵonathan ssp: http://lists.gnome.org/archives/gtk-devel-list/2004-March/msg00125.html OK (I guess that shows that the bar for getting listed as a contributor is fairly low :-) a one-character contributor... maclas: 2 :-) oh, I thought he only pointed out the missing trailing slash. ssp: Basiclaly, you can commit anything you want to plan/ and it will be good :-) Hmm, does anybody actually know muppet's name? It's generally been against my policy to credit anybody just by IRC nickname you can end the contributors list with like "and that muppet from irc" (: scott something or the other ask him www.asofyet.org yosh: if I see him, I will I googled for that piece of information once ah, Scott Arrington http://www.asofyet.org/muppet/ owen: ok, I'll do that tonight then. So for a short while we will see something as rare as outdated reality compared to a website OK, anything ele? owen: when is 2.4.1 due ? hi, sorry for being late maclas: I think we'll try for a week Hey tml whats the "topic" for 2.4.1 ? is it too late for a minor change to glib's configure.in? Otherwise HAVE_INT64_AND_I64 don't get defined with gcc, breaking 64-bit int g_printf... Only just noticed. tml: You didn't miss much, really. I'm being slow at writing the release announcements, will probably distcheck tomorrow, the website and docs contain lots of out of date stuff, not much we can do about that. That's about all :-) tml: If you could run it by me, I'd appreciate it, but it shoudl be fine to make a change owen: OK One final thing ... I'll send the final release announcements around for people to check, but I wanted to run the "Team" and "Special Thanks" listings by people here For glib, I have Team:Matthias Clasen, Tim Janik, Noah Levitt, Tor Lillqvist, Manish Singh, Owen Taylor, and Sebastian Wilhelmi Special Thanks: Jonathan Blandford, Hans Breuer and J Ali Harlow, George Lebl For GTK+, I was thinking: [ sorry for the suspense here, badly prepared ] [ drum roll ] Team: Jonathan Blandford, Matthias Clasen, Tim Janik, Tor Lillqvist, Federico \Mena Quintero, Kristian Rietveld, Soeren Sandmann, Owen Taylor Special thanks to: Morten Welinder, Hans Breuer, J Ali Harlow, Dov Grobgeld This stuff is very abitrary, of course. but I don't really know how to make it not so. Anybody obvious I'm misisng for either? Anybody who should be categorized differently? owen: I'd say jamesh and andersca should both be in the special thanks for gtk+ other contributions go in plain "Thanks", right? maybe I should be categorized differently. but I dunno. kris: No, you don't get to get out of being inthe Team for 2.4 :-) also markmc, maybe. he did one of our new widgets... ssp: Yeah, I should add Jamesh for Glib as well, the autogen stuff was pretty massive ok if you think I did enough useful stuff ah, I did hack on the toolbar and the uimanager * andersca remembers now maclas: Ah, yes. Maybe alexl too, since the icon theme stuff is just his code stolen owen: i think yosh definitely belongs on the gtk team list owen: jamesh did the original uimanager implementation michael/jody for pounding on the GtkAction/UIManager stuff? (there're lots of small fixes in gtk+/ChangeLog from him if you want justification) federico: Hmm. Not sure there ,but definely Marco federico: if pounding counts, then marco also federico: yes, jody should be in "special thanks", imo. I had a couple of useful discussions regarding uimanager with him, and he provided patches yes, jody in special thanks for G_DEFINE:TYPE() ;) owen: I think maybe both Hans and I should be in the Special Thanks section for GTK, Hans did lots during the winter while I got active quite late the "win twins"... owen: gtk special thanks also go to michael meeks tml: I could move Hans up to the team section for both if you want, he certainly dives in and fixes a lot of stuff owen: fine with me seth, for the file chooser design? federico: Good idea. This is getting like the oscars... rambaby: Are you thinking of of anything in particular for Meeks or just general bugfix submission? owen: i was thinking of define_type originally and looked, and there're a couple of bug patches and commits from him in the changelog Oh, hmmm, probably Semion Chichelnitsky for all the RTL widget patches Of course, the problem is that this shades off very gradually ... Counting straight appearances of names in NEWS, the top counts are: ok folks, badly need sleep and have to get up early. night. --- You are now known as | <|> 72 Matthias Clasen <|> 43 Owen Taylor <|> 37 Soeren Sandmann <|> 27 Federico Mena Quintero <|> 16 Jonathan Blandford <|> 12 Morten Welinder <|> 12 Kristian Rietveld <|> 8 Marco Pesenti Gritti <|> 8 Hans Breuer <|> 8 Alexander Larsson <|> 7 Mark McLoughlin <|> 7 Christian Persch <|> 6 Tor Lillqvist <|> 6 Murray Cumming <|> 6 Manish Singh <|> 6 Damon Chaplin <|> 5 Padraig O'Briain <|> 5 Michael Natterer <|> 5 Jody Goldberg <|> 5 Anders Carlsson --- You are now known as owen But that doesn't really say a lot - Tor gets few mentions because I always summarize * Win32 [Tor Lillqvist] - asdf - asdf -asdf, etc.. owen: after the special thanls section, you can make normal thanls section and relist everyone from the 2.3.x NEWS ;) rambaby: Yes, that's what I've done before and was planning to do now we used to summarize * GtkTreeView [Kris] too i think. owen: well, you have to draw the arbitrary line *somewhere*. so add a forgotten-about-someone disclaimer and be done with it. kris: Yeah. In those counts GtkTreeModelFilter gives you the same count as Henning Nielsen Lund's gtk.org => gnu.org :-) rambaby: Indeed. heh OK, thanks for the feedback folks. I think we're all done here. When I send out the drafts, if people could find as many of the 2.2's that should be 2.4's as possible, that woudl be appreciated. --- markmc is now known as mcsleep ok * kris goes for this 6 hours of sleep night kris night. Meeting ended March 15 18:34 EST (23:34 UTC)