All times shown according to UTC.

Time Nick Message
00:11 DGMurdockIII left #miro-hackers
00:54 willkg time to jet.  night all!
00:56 willkg left #miro-hackers
01:00 romus joined #miro-hackers
01:00 romus is now known as Roman_qa_
01:11 Roman_qa_ left #miro-hackers
01:14 maggie_s left #miro-hackers
01:25 glee joined #miro-hackers
01:27 Roman_qa_ joined #miro-hackers
01:37 Roman_qa_ left #miro-hackers
01:49 xS^away joined #miro-hackers
01:51 xS^- joined #miro-hackers
01:53 xS^ left #miro-hackers
01:53 xS^away left #miro-hackers
02:42 xS^- left #miro-hackers
04:02 glee left #miro-hackers
04:37 odenthaa joined #miro-hackers
05:07 DGMurdockIII joined #miro-hackers
05:33 glee joined #miro-hackers
06:01 glee left #miro-hackers
06:07 bendk left #miro-hackers
06:07 bendk joined #miro-hackers
06:17 glee joined #miro-hackers
06:36 bendk left #miro-hackers
06:38 glee left #miro-hackers
06:39 bendk joined #miro-hackers
06:59 bendk left #miro-hackers
06:59 xS^ joined #miro-hackers
06:59 bendk joined #miro-hackers
06:59 sedat joined #miro-hackers
07:12 xS^ left #miro-hackers
07:19 xS^ joined #miro-hackers
07:26 xS^ left #miro-hackers
07:49 bendk left #miro-hackers
07:50 bendk joined #miro-hackers
08:20 morganknutson left #miro-hackers
08:27 morganknutson joined #miro-hackers
09:19 glee joined #miro-hackers
09:24 janetPCF joined #miro-hackers
11:13 sedat left #miro-hackers
11:13 sedat joined #miro-hackers
12:19 xS^ joined #miro-hackers
12:53 willkg joined #miro-hackers
12:55 arpu left #miro-hackers
13:10 willkg janetPCF: i'm building a new osx nightly now.
13:10 janetPCF oh that's awesoem
13:10 willkg janetPCF: glee fixed the problem, but after the nightlies had already run.
13:11 janetPCF cool -
13:12 willkg janetPCF: new osx nightly posted.
13:12 sorry about that.
13:17 afk a bit....
13:18 arpu joined #miro-hackers
13:34 janetPCF1 joined #miro-hackers
13:35 janetPCF1 left #miro-hackers
13:37 janetPCF left #miro-hackers
13:55 paulproteus deardiary: Fixing more issues (hopefully the last batch of issues) with the new theme prepared by mattjohnson
14:01 z3p good morning
14:01 willkg oo...  glee broke miro on linux and windows.
14:01 z3p I was wondering about that when I read the checkouts in bed
14:02 glee willkg: Yeah I know, I broke it.  Just working out how to fix it.
14:02 willkg glee: Gradient is already in drawing--it's not in widgetset.
14:03 glee willkg: Yep, there's that one.  And then there's css_to_color which is needed but there's some circular import issue, so I'm just going to stick that import in the function.
14:03 willkg glee: fixing that brings up the css_to_color problem.  that's defined in style.py.
14:03 z3p miro
14:03 willkg glee: all the widgetset code should be self-contained.  it's really bad for widgetset code to import the portable widgets code.
14:03 z3p whoops, this isn't gmrun
14:04 glee willkg: Let's dup the function instead?
14:04 willkg glee: i'd just copy css_to_color and stick it in drawing.  it's a trivial function.  it's not going to change.
14:04 glee: yeah--what you said.
14:05 glee: alternatively, we could move css_to_color somewhere.  like to base.py or something in both the widgetset implementations.
14:06 glee You mean dup it in both the base.py of GTK and Cocoa?
14:06 willkg no...  so you could either copy it to gtk/drawing.py, or we could move css_to_color to base.py in both the widgetset implementations.
14:06 glee willkg: If it's going to be dup'ed anyhow then I'm happy to go either way
14:07 willkg the latter plan requires more work because you'd have to fix it everywhere.
14:07 i don't have a preference one way or the other.
14:07 i'm just thinking out loud.
14:07 z3p why does css_to_color need to be in widgetset?
14:07 willkg z3p: glee added some code in gtk/drawing.py that uses it.
14:07 z3p glee: why not just figure out what css_to_color returns for your color, and use that?
14:08 it's not like we're passing CSS colors into widgetset code
14:08 willkg z3p: i suspect it's because morganknutson sends us colors in rgb-hex.  having the color in the code the way morgan stated it makes it easier to maintain.
14:09 z3p willkg: that's true, but all of the widget code deals with colors in RGB triplets
14:09 er, widgetset
14:09 that's what the frontend passes in
14:09 glee Hm I guess either way it's going to be suboptimal, really hate duplicating code but let's go with that for now, then figure something out later
14:10 Maybe we could have some colorutil module that contained "stuff" like this
14:12 willkg well, if i have my way, at some point p, i'm going to move all the color and font information out of there and into a centralized place that we can theme off of.
14:12 but...  that's probably a 4.1 thing.
14:21 janetPCF joined #miro-hackers
14:37 willkg afk a bit to grab some coffee....
14:45 janetPCF left #miro-hackers
14:53 janetPCF joined #miro-hackers
14:54 CarlFK joined #miro-hackers
15:05 CarlFK left #miro-hackers
15:05 CarlFK joined #miro-hackers
15:15 z3p left #miro-hackers
15:17 z3p joined #miro-hackers
15:24 CarlFK1 joined #miro-hackers
15:24 CarlFK left #miro-hackers
15:33 bendk willkg: sorry about that os x bug, I can fix it if you want
15:34 willkg bendk: no worries--glee already fixed it.
15:34 ajonas joined #miro-hackers
15:34 z3p deardiary: I'm not sure cookies have ever worked on Windows
15:34 willkg z3p: very probably.
15:37 zanoi glee, i have pushed magnet links. it includes the protocol association code for OS X you gave me. Would be great if you could test if that works
15:38 paulproteus ajonas: I think I fixed all overlay problems on the POV theme. Can you check for me?
15:38 glee zanoi: sure I'll get back to you
15:38 ajonas paulproteus: exciting! i will in a min.
15:38 paulproteus P.S. Oh my God, that was a pain.
15:38 zanoi glee, thx
15:38 bendk zanoi: hey, i have a question about the magnet link code
15:38 got a sec?
15:38 zanoi sure
15:39 paulproteus ajonas: + http://bugzilla.pculture.org/s[…]_bug.cgi?id=16620
15:39 bendk so if you have a magnet link, you can calculate an info hash right?
15:40 then with an info hash you can create the torrent info?
15:40 I guess I'm checking if 1) that's possible, so that I can use it with some code I'm trying
15:40 and wondering 2) if it is possible, could we combine the codepaths a bit more?
15:41 zanoi i think the short answer is yes, the long answer is "it's more complicated"
15:42 bendk so I'm trying to get the torrent info so that I can call files(), do you think that will work?
15:42 zanoi yes
15:42 i do that in my magnet implementation
15:43 however, you have to start the torrent downloaded, then wait until you have received the metadata and only then you can call files
15:44 bendk ahh, okay
15:44 zanoi which gave me quite some problems because it meant the download has started before you had the metadata to figure out whether you're downloading a file or a directory :)
15:46 bendk yeah, I'm seeing
15:46 so... should we start the dev cal?
15:46 zanoi bendk, if you're interested i have written a magnet test script where i do stuff like calling files
15:46 yes, but npr is busy
15:46 so someone else will have to do it :)
15:47 willkg oh...  i can do it.  i'll start it now.
15:48 z3p woah, totally spaced on the call reading xpcom docs
15:49 is nobody talking, or is my skype totally fubared?
15:49 willkg ben's talking now.
15:49 zanoi you skype is fubared
15:49 willkg i was talking before, but i've got a cold.
15:50 z3p neat, skype broke all the sound on my phone
15:50 that's a new one
15:50 janetPCF Hey - did you forget me????
15:51 z3p willkg: gonna restart it and try again
15:51 zanoi heh, poor janet
15:51 so now i dropped out :(
15:52 willkg, call me plz!
16:16 CarlFK1 left #miro-hackers
16:23 CarlFK joined #miro-hackers
16:41 xS^ left #miro-hackers
16:52 willkg afk for food....
17:02 CarlFK1 joined #miro-hackers
17:02 CarlFK left #miro-hackers
17:15 CarlFK joined #miro-hackers
17:15 CarlFK1 left #miro-hackers
17:38 CarlFK left #miro-hackers
17:40 sedat left #miro-hackers
17:41 sedat joined #miro-hackers
17:43 janetPCF deardiary: someone broke the download tab: http://develop.participatorycu[…]se_to_a_text_file
17:43 oops wrong link
17:43 bug #16653
17:44 willkg deardiary: spending an hour or two seeing how miro 4 works on ubuntu natty.
17:49 janetPCF: i'll grab 16653.
17:52 janetPCF cool
17:53 looks like the bottom chrome got messed up today too - the volume and shuffle repeat buttons are pushed left
17:54 willkg glee: zanoi's magnet uri checkin breaks building miro on osx.  can you take a look at it?  i don't understand the bits involved.
17:55 glee willkg: righteo, but it'd probably in a couple of hours
17:55 willkg glee: really?  i think it's a plist issue.  i'm guessing it's the code zanoi checked in that he got from you.
17:56 glee willkg: I'll quickly build it and check
17:56 willkg glee: it's impossible to miss.  it happens in setup.py.
17:58 glee willkg: Should be okay now.
17:59 willkg glee: you rock.  thank you!
17:59 glee: yup...  i never would have figured that one out.
18:02 wow that connect display has small text.
18:02 at first i thought it was a weird gradient, but then i noticed those are tiny letters.
18:03 zanoi willkg, sorry was eating dinner
18:04 glee, oops, sorry for putting the osx stuff in the wrong place
18:04 glee zanoi: No, that's alright
18:06 willkg deardiary: miro works on natty--only a minor tweak to the helperscript.
18:13 janetPCF: looks like that downloading tab bug got fixed after i redid the osx nightly, so i'm doing another one now.
18:18 CarlFK joined #miro-hackers
18:19 willkg janetPCF: new osx nightly posted.  this time with more hot sauce.
18:21 sedat left #miro-hackers
18:22 willkg wow...  134 bugs created in the last week.
18:23 dev call notes: http://bluesock.org/~willg/blo[…]all_20110309.html
18:23 sedat joined #miro-hackers
18:23 willkg it's totally worth skimming the last few weeks of dev call notes to see the bugzilla numbers.  talk about serious bug churn.
18:30 Sergeiqa joined #miro-hackers
18:53 CarlFK left #miro-hackers
18:57 willkg bendk: you on?
18:57 bendk yup
18:58 willkg bendk: so i'm fixing problems with the New Watched Folder dialog.  on OSX it's a "sheet".  when the user opens the dialog, then clicks on the "Choose" button, it opens an ask_for_directory dialog that's transient to the New Watched Folder dialog.
18:59 bendk: but since the New Watched Folder is a sheet and not a dialog, apparently it doesn't have an nswindow and then everything goes to hell.
18:59 bendk hmm
18:59 willkg bendk: so...  i'm not sure what to do in this case.  should i make the ask_for_directory dialog not transient on the New Watched Folder dialog?
18:59 bendk I think there should be an nswindow somewhere
18:59 willkg that seems to be messy.
19:00 bendk aren't all our OS X "dialogs" really sheets?
19:00 like the bug reporter dialog?
19:00 willkg i have no idea.  i don't do much in osx land.
19:00 it's very possible.
19:00 i'm not sure there's a case where a sheet spins off another dialog, though.
19:00 bendk okay, nevermind I think I'm confusing the case
19:00 willkg the preferences window is a separate window.
19:00 bendk we should make it transient to the window that the sheet is over
19:01 willkg ok.  i'll go look at the sheet code and see what i can do.
19:01 thanks!
19:01 kcw the edit item dialog pops up a path chooser and seems to work
19:01 bendk in the Dialog class, self.sheet_parent.nswindow should do it
19:01 willkg kcw: is it a separate window or is it a sheet?
19:01 bendk where Dialog == our portable Dialog class, but we really display it as a sheet :)
19:02 willkg mmm....  i'll look into it.  maybe i implemented something wrong.
19:02 kcw it's a widgetset.Dialog
19:06 z3p deardiary: woo, cookies are saved on Windows now
19:08 janetPCF left #miro-hackers
19:09 willkg kcw: that's a regular dialog--it's not a sheet.
19:09 bendk: i figured out what's going on.  so New Watched Folder dialog is transient to the MainWindow.  then i go and make the ask_for_directory transient to New Watched Folder which is really a sheet.
19:10 bendk: so what happens is ask_for_directory goes to be transient to self.sheet_parent.nswindow and self.sheet_parent is New Watched Folder and it doesn't have an nswindow.
19:10 bendk agctually I think we should always make things transient to the MainWindow
19:10 willkg so the ask_for_directory should be transient to the MainWindow and not the dialog that spun it off?
19:11 janetPCF joined #miro-hackers
19:12 willkg that gives me mild icky feelings.
19:12 kind of like the feeling where i think i just stepped in gum.
19:13 CarlFK joined #miro-hackers
19:16 bendk willkg: hmm, yeah your right about that, never mind the suggestion
19:17 willkg bendk: i'm tossing around setting nswindow for dialog sheets on osx to the thing they're transient to (i.e. the main window).
19:17 bendk for Dialog that should definitely be true
19:18 willkg i think my New Watched Folder problem is the only place where we have something transient on something transient.
19:18 ok.  i'll test that out and see what happens.
19:18 wow...  i thought this dialog was going to be so simple.  i didn't expect the Spanish Inquisition.
19:31 bendk glee: you around?
19:33 janetPCF left #miro-hackers
19:36 willkg bendk: wow...  so that didn't work.
19:45 bendk: i ended up ditching the transience.  seems to work ok without it.
19:49 paulproteus z3p: I've been trying for a long time to get something working, and I wonder if you can give me some advice.
19:49 arpu left #miro-hackers
19:49 paulproteus er, actually, I'll try one thing first...
19:49 z3p paulproteus: my experience is generally in breaking working things, but I'll help if I can :)
19:49 DGMurdockIII left #miro-hackers
19:49 DGMurdockIII joined #miro-hackers
19:50 DGMurdockIII left #miro-hackers
19:50 DGMurdockIII joined #miro-hackers
19:50 DGMurdockIII left #miro-hackers
19:50 paulproteus z3p: Yeah, let me tty in 10ish
19:50 DGMurdockIII joined #miro-hackers
19:50 DGMurdockIII left #miro-hackers
19:51 DGMurdockIII joined #miro-hackers
19:51 DGMurdockIII left #miro-hackers
19:51 DGMurdockIII joined #miro-hackers
19:51 DGMurdockIII left #miro-hackers
19:51 DGMurdockIII joined #miro-hackers
19:52 DGMurdockIII left #miro-hackers
19:52 paulproteus z3p: Okay, hi.
19:52 DGMurdockIII joined #miro-hackers
19:52 DGMurdockIII left #miro-hackers
19:52 DGMurdockIII joined #miro-hackers
19:52 DGMurdockIII left #miro-hackers
19:53 DGMurdock joined #miro-hackers
19:53 z3p paulproteus: hi
19:53 DGMurdock left #miro-hackers
20:02 paroneay` joined #miro-hackers
20:02 paulproteus So I'm trying to programmatically pass data into get_video_list_for_user
20:02 This works: {% get_video_list_for_user "PBS" as pbs_list %}
20:02 What I want: {% get_video_list_for_user current_video.authors.all.0.username as pbs_list %}
20:02 That raises an exception.
20:05 paroneayea left #miro-hackers
20:06 paulproteus (Then my 'net connectivity broke, and the above messages were sitting in my send queue for 10 minutes.)
20:06 (In the meantime, I got myself a cookie and put some Nutella on it. That was good.)
20:06 I'll show you the exception (as reported by the Django test client) in a sec
20:08 ValueError: invalid literal for int() with base 10: 'PBS' http://pastebin.com/tDBusfut z3p
20:09 (I'm using an extra {% with %} around the username, but that shouldn't change anything.
20:10 I'm going to add some logging to find out what "author" is, inside the template tag, but let me know if you have any thoughts off the top of your head.
20:10 z3p paulproteus: if you do get_video_list_for_user with the ID, does that work?
20:10 paulproteus Empty list
20:11 z3p paulproteus: oh, I see what it is
20:12 paulproteus: you just want  {% get_video_list_for_user current_video.authors.all.0 as pbs_list %}
20:12 paulproteus: the node assumes that if you're passing in a variable, you're passing in a user object
20:12 paulproteus FWIW that doesn't do it either.
20:12 Empty list.
20:12 z3p paulproteus: really...
20:12 paulproteus er let me double check
20:13 DGMurdockIII joined #miro-hackers
20:13 DGMurdockIII left #miro-hackers
20:13 paulproteus It totally didn't work before.
20:13 But now it does.
20:13 So, thank you.
20:13 z3p sweet
20:13 DGMurdockIII joined #miro-hackers
20:13 DGMurdockIII left #miro-hackers
20:14 DGMurdock joined #miro-hackers
20:14 DGMurdock is now known as DGMurdockIII
20:28 zanoi bendk, did you say you had magnet links associated with miro on your machine?
20:34 bendk1 joined #miro-hackers
20:34 bendk1 kcw: hey, kaz you around?
20:35 paulproteus ajonas: http://coffee.mirocommunity.or[…]mansour-el-kikhia
20:35 w/r/t http://bugzilla.pculture.org/s[…]_bug.cgi?id=16631
20:36 Seem to fit the bill?
20:36 ajonas paulproteus: one min, call
20:36 paulproteus: new tab is funyn
20:37 paulproteus: and most watched
20:37 paulproteus lookin'
20:37 ermp agreed
20:39 ajonas also there's a slight spacing difference between category and author
20:39 paulproteus: but generally, looking good!
20:43 zanoi bendk1,  did you say you had magnet links associated with miro on your machine?
20:43 paroneay` left #miro-hackers
20:43 paroneayea joined #miro-hackers
20:43 Sergeiqa left #miro-hackers
20:44 bendk1 yup
20:44 by that I mean when I click on them in firefox, I get the option to open miro
20:45 paulproteus ajonas: retry, if you would
20:45 ajonas one min
20:46 zanoi bendk1, what os? and since when?
20:49 bendk1 linux
20:49 and I'm guessing since a while, since I don't think that running code from master changes my associations
20:49 that's what I was trying to say on the call, I think since firefox knows to run miro for torrents, it gives me the option for magnet links
20:50 zanoi well actually it shouldn't :-S
20:50 magnets aren't even necessarily torrents
20:51 ajonas paulproteus: author tab is semi missing
20:52 zanoi and everytime i tried firefox didn't know how to open magnet links
20:52 in any os, even if i installed programs that officially support magnet links
20:53 ajonas paulproteus: minor thing: can author come first?
20:56 zanoi bendk1, does it list other programs? you are running lucid right?
20:58 bendk1 lucid yes
20:58 and no other programs
20:58 willkg bendk: have you ever installed miro from our ppa?
20:59 zanoi ok, let me revise that statement. Installing Azureus in Windows makes magnets work in IE
20:59 but not in firefox :(
21:00 bendk1 wilIkg: I'm not sure on that one
21:00 zanoi could it be that you at some point manually selected miro as a program to open magnet links? I can't really think of anything else that could make it happen
21:00 :-S
21:00 sedat left #miro-hackers
21:01 bendk1 hmm, that could be true
21:01 if you're not seeing the same things, then i would just forget about me
21:01 I'm definitely a weird sample
21:02 sedat joined #miro-hackers
21:02 zanoi i wouldn't mind you as sample, i could mark the bug as fixed then :-)
21:02 or mark it "WORKSFORBEN"
21:07 paulproteus ajonas: re: author tab: try now
21:07 ajonas: Author can come first. Working on it.
21:07 ajonas paulproteus: looks good
21:08 paulproteus: there's still a slight shift
21:08 paulproteus Yeah, I don't know what's up with that.
21:08 ajonas the tabs in the second tab + are all the way at the left edge
21:08 paulproteus: maybe a margin?
21:08 paulproteus I kind of want to leave that to mattjohnson since he'll do it about 100x faster than I will.
21:08 http://coffee.mirocommunity.or[…]granbys-speech-ny -- one more thing there
21:09 Sometimes authors have no first_name and no last_name
21:09 I'll look into that case now
21:11 ajonas paulproteus: interesting - i wonder why that's different than pbsnewshour
21:14 paulproteus ajonas: regardless, fixed
21:14 (I use just "username" now, not first_name and last_name)
21:16 zanoi glee, should the  plist code make magnet links work in every browser on osx?
21:17 because there doesn't seem to be any "browser independent" way on any other platform
21:58 DGMurdockIII left #miro-hackers
21:58 DGMurdockIII joined #miro-hackers
21:58 DGMurdockIII left #miro-hackers
21:58 DGMurdockIII joined #miro-hackers
21:58 DGMurdockIII left #miro-hackers
21:59 DGMurdockIII joined #miro-hackers
21:59 DGMurdockIII left #miro-hackers
21:59 DGMurdockIII joined #miro-hackers
21:59 DGMurdockIII left #miro-hackers
22:00 DGMurdockIII joined #miro-hackers
22:00 DGMurdockIII left #miro-hackers
22:00 maggie_s joined #miro-hackers
22:00 DGMurdockIII joined #miro-hackers
22:00 DGMurdockIII left #miro-hackers
22:00 DGMurdockIII joined #miro-hackers
22:00 DGMurdockIII left #miro-hackers
22:01 DGMurdockIII joined #miro-hackers
22:01 DGMurdockIII left #miro-hackers
22:01 DGMurdockIII joined #miro-hackers
22:01 DGMurdockIII left #miro-hackers
22:02 JasonWoof left #miro-hackers
22:03 maggie_s ajonas: Hi ajonas!
22:03 ajonas hi maggie_s!
22:04 maggie_s ajonas: I sent you an email yesterday - just worried it might get lost among the messages from the robot
22:04 ajonas maggie_s: hey, yes, i just realized i had a group of unread emails
22:04 give me one min and i'll get back to you
22:04 maggie_s ajonas: OK
22:05 ajonas maggie_s: so, for bulk edit, my preference (and i'm pretty sure paulproteus's) would be for you to make a new ticket for each issue
22:06 JasonWoof joined #miro-hackers
22:06 JasonWoof left #miro-hackers
22:06 JasonWoof joined #miro-hackers
22:06 maggie_s ajonas: OK, will do
22:06 ajonas maggie_s: i'll cancel that payment now
22:06 maggie_s: re: downgrade
22:07 maggie_s: the cancellation messages are usually for upgrading
22:07 maggie_s: but also for downgrading during a free trial
22:08 maggie_s: if not downgrading during a trial, you should get a form to adjust your payment, but see http://bugzilla.pculture.org/s[…]_bug.cgi?id=16614
22:08 (basically, that case is not currently working as it should)
22:08 maggie_s ajonas: I see... my site should be well beyond the trial stage. So, there was no e-mail for the transaction I mentioned?
22:09 ajonas maggie_s: there were many emails, i haven't looked through them all yet
22:10 maggie_s: one sec, i'll cancel the one you mentioned and look up the other
22:11 maggie_s: thankfully you can search transactions by id number!
22:13 maggie_s ajonas: how do you do that? search by transaction ID does not return anything for me
22:13 ajonas maggie_s: ah, i'm in the biz account
22:13 but it didn't actually work yet :)
22:13 maggie_s: though it did last time i tried!
22:14 maggie_s ajonas: oh, I see
22:14 ajonas maggie_s: ok, canceling
22:15 maggie_s: you should get an email in you test account
22:17 maggie_s: it shouldn't otherwise affect your site state
22:21 sedat left #miro-hackers
22:22 maggie_s ajonas: right, it is now marked as canceled
22:22 ajonas maggie_s: cool
22:22 maggie_s: we added the info to the basic page, but the styling is off, so i need to fiddle with it more
22:23 maggie_s ajonas: OK, I'll have a look when it is ready
22:24 ajonas: now I'll go back to bulk edit page - I think it'll need more work
22:24 ajonas maggie_s: it's going to be hard to get more into this release - more edge cases might have to wait for 1.2.1 or 1.5
22:25 maggie_s ajonas: edited video info must at least save correctly
22:25 ajonas maggie_s: reasonable
22:25 :)
22:25 maggie_s ajonas: performance improvement can wait, of course
22:26 kcw bendk1: what's up?
22:32 Roman_qa joined #miro-hackers
23:01 bendk1 hey kcw still there?
23:01 I took a shower break
23:02 kcw hey
23:03 bendk1 so I think there's an issue with the TableView scroll position change you made
23:03 62e029764fc122dff946c330e10acd8148a3187a
23:03 can you help me understand the code?
23:03 maggie_s left #miro-hackers
23:03 bendk1 I get this:
23:04 if scroll_to: # widgetstate restoring a saved position
23:04 self.scroll_position = scroll_to
23:04 but not this:
23:04 else: # fixing position if it has changed
23:04 scroll_to = self.scroll_position
23:05 I'm not sure if what I'm saying is making sense :)
23:05 I get that we try to set the scroll position at the begining, but then why do we keep setting after each model_change() call?
23:06 kcw bendk1: model_change can happen after the scroll position is reset, particularly by bulk_change and possibly by other things. I don't know the situation on OS X as definitely as GTK, but I think it is the same
23:07 bendk1 so what's happening is when I go to a tableview whenever there's an update, the list scrolls (I think maybe to the initial position)
23:07 Also, I don't think on_scroll_changed is being called
23:10 kcw hm. I'll double-check that it doesn't work without calling set position all the time, but I think the problem is probably that it isn't being saved when it should be. What it's supposed to be doing in saving it whenever there's a change that's supposed to happen, and then restoring it whenever any change happens - because, unless there's information available that isn't in gtk, that seems to be the only way to handle the positio
23:10 bendk1 yeah, I think if we saved the scroll position on every scroll that would work
23:11 I also think we might be able to get rid of the cals to it all the time.  I made some changes it portable to only call start_bulk_change if it was going to add or delete a bunch of items
23:11 I gotta run though, I think I understand what's happening now better, so if you want to just pass it to me that's fine
23:11 see you later
23:12 DGMurdockIII joined #miro-hackers
23:12 DGMurdockIII left #miro-hackers
23:15 arpu joined #miro-hackers
23:47 ajonas left #miro-hackers

← Previous day | Index | Server Index | Channel Index | Today | Next day → | Atom Feed | Search | Google Search | Plain-Text | plain, newest first