All times shown according to UTC.

Time Nick Message
00:05 maggie_s ajonas: if I downgrade, say, from Max to Plus, I can freely go back to Max in 15 minutes without even being taken back through Paypal - is it the same for you?
00:06 ajonas maggie_s: i haven't checked, that sounds buggy
00:06 maggie_s: i bet it has something to do with the fact that you were on max before
00:06 if you were?
00:06 maggie_s ajonas: I was
00:07 ajonas maggie_s: can you check your paypal recurring payments?
00:07 and see if there's one for the $75?
00:07 maggie_s ajonas: just a moment, I'll try
00:07 ajonas thanks
00:07 when you upgrade, we should make a new payment
00:08 and the old one should also still be there (until we cancel it by hand)
00:08 maggie_s ajonas: no, none for today
00:08 ajonas maggie_s: hmmm
00:09 maggie_s: that's odd - you're looking in the test account profile, yeah?
00:09 the same one you used?
00:10 maggie_s ajonas: it's -15$ for downgrading to Plus, then nothing for upgrading to Max, then another -15$ for getting down to Plus again
00:10 ajonas maggie_s: yeah that is wrong.
00:10 maggie_s: bug it up!
00:11 maggie_s ajonas: sure, I will
00:11 ajonas maggie_s: see if you can reproduce on another site if possible
00:11 willkg left #miro-hackers
00:11 maggie_s ajonas: hmm... I'll look if I can do that on dalmatia
00:12 ajonas maggie_s: thanks! night for now.
00:12 ajonas left #miro-hackers
01:34 CarlFK left #miro-hackers
01:40 maggie_s left #miro-hackers
01:46 glee left #miro-hackers
01:49 glee joined #miro-hackers
01:54 bendk left #miro-hackers
01:56 bendk joined #miro-hackers
02:09 Olena-QA left #miro-hackers
02:29 DGMurdockIII left #miro-hackers
03:01 maggie_s joined #miro-hackers
03:25 CarlFK joined #miro-hackers
03:29 morganknutson left #miro-hackers
04:29 DGMurdockIII joined #miro-hackers
04:45 CarlFK left #miro-hackers
05:42 DGMurdockIII left #miro-hackers
09:07 maggie_s left #miro-hackers
11:36 CarlFK joined #miro-hackers
12:02 Jarrhead joined #miro-hackers
13:18 willkg joined #miro-hackers
14:23 ajonas joined #miro-hackers
14:25 z3p good morning
14:25 deardiary: working on device bugs on OS X today!
14:25 CarlFK salut
14:30 willkg deardiary: spending some time fixing the SubtitlesSupport page so it reflects what Miro can actually do subtitles-wise like a proper design document so i stop wasting my time chasing down issues that it turns out we knew about.
14:44 ajonas left #miro-hackers
15:00 z3p left #miro-hackers
15:02 z3p joined #miro-hackers
15:24 Jarrhead left #miro-hackers
15:56 ajonas joined #miro-hackers
16:20 willkg afk a bit for food....
16:37 glee deardiary: transcoding bug fixing - multiple successive seeks flaky
16:48 zanoi deardiary: i keep forgetting to use deardiary lately..
16:49 deardiary: will be working on my list of magnet link issues today
16:59 janet_ joined #miro-hackers
17:00 janet_ deardiary: I appear to be back online for a bit - hopefully the whole evening.  I think they are still working on the lines, but have stopped for the day.
17:01 z3p: Paul, I did a bit more device testing, and found a few more crashes.
17:01 I was just about to go through that list I made and enter them all in as bugs - is there any I don't need to bother with?
17:01 z3p janet_: yeah, there were some issues with changes I made last night
17:01 janet_: if there were KeyErrors about 'settings' or 'enabled', you don't have to bother with them
17:04 janet_ ok thanks
17:09 z3p janet_: also, the bug with the duplicate selection, if you haven't filed it
17:28 janet_ z3p: I think I just filed them all
17:28 z3p janet_: I see; thanks!
17:28 janet_ but anyway - after the key error when I attached the droid today - I got a second crash
17:28 in _item_info_for
17:28 ValueError: dictionary update sequence element #0 has length 1; 2 is required
17:28 is that part of your key error stuff, or should it be filed?
17:28 z3p janet_: I think that's related to some metadata stuff, but I'll take a look
17:29 janet_ ok - I'll put the whole crash log in the bug.
17:31 bug #16413 is interesting - I was thinking last night that a cool unintended feature of the device stuff is that someone could attach a digital camera, then convert the output files directly to their library, so like they could go from camera to ogg in 1 step and upload to a website
17:31 but it crashed when I attached my camera
17:51 willkg left #miro-hackers
17:52 zanoi willkg, bendk: so I was wondering whether I should add saving/loading the DHT state between restarts
17:52 bendk how much work is it?
17:52 zanoi i thought i'd add it but now i saw we don't save/load the session state
17:52 so i guess i'll just leave it
17:53 bendk does it take a while to recreate the state?
17:53 zanoi bendk: not that much
17:53 you mean performance wise? no idea
17:53 bendk seems okay either way to me, you make the call
17:53 zanoi ok
17:55 i'd say either we don't add it or add both DHT state saving/restoring and torrent session saving/restoring
17:55 doesn't make much sense to me to have DHT state restoring but not torrent session
17:56 bendk That makes sense
17:56 I think we weren't saving the session just because no one thought about it
17:56 so if you want to implement both, go for it
17:56 zanoi ok
17:57 willkg joined #miro-hackers
17:57 z3p deardiary: having different names for the same thing is really starting to annoy me
17:58 willkg z3p: me, too.  we should do a pass on fixing that for 4.1.
18:15 willkg left #miro-hackers
18:16 bendk kcw: what's the best way currently to get the album art for an item?
18:17 kcw bendk: to get the album art that's already extracted?
18:17 bendk yup
18:17 kcw it's in cover_art
18:18 bendk ahh okay
18:19 and for ItemInfo, I can use the thumbnail attribute to get either the cover art or the thumbnail?
18:20 kcw bendk: yeah
18:22 willkg joined #miro-hackers
18:38 willkg kcw: 2b216047 caused the failure in 16416.
18:39 kcw: looks like you changed get_title to pull from the store before it checks the user-changed title for the item.
18:39 kcw: i figured i'd check with you before switching the order.  i don't know why you did what you did--there's no documentation that i can see.
18:40 kcw: do you want me to assign this bug to you?
18:41 kcw willkg: yeah
18:41 willkg kcw: out of curiosity, did you run the unit tests after making those changes?
18:42 kcw willkg: I did. I was planning on looking at that this morning
18:43 willkg kcw: ok.
18:56 bendk heh, I was just about to comment on that very same changeset
18:57 it causes an issue with the UI as well, we used to be calling item.get_description(), now we're just using description directly
18:57 I just checked in a changeset that hacks together a fix, but I'm guessing there's a better way
19:00 zanoi: I'm getting errors when I start up miro now
19:00 File "/home/ben/miro-git/tv/linux/dist/usr/lib/pytho​n2.6/site-packages/miro/dl_daemon/download.py", line 262, in set_dht
19:00 self.session.start_dht()
19:00 ArgumentError: Python argument types in
19:00 session.start_dht(session)
19:00 did not match C++ signature:
19:00 start_dht(libtorrent::session {lvalue}, libtorrent::entry)
19:05 willkg bendk: what operating system are you using?
19:05 bendk ubuntu lucid
19:05 willkg bendk: what version of libtorrent is that using?
19:05 bendk 0.14.10
19:06 willkg that's probably the problem.  it works fine for me on ubuntu maverick with libtorrent 0.15.5.
19:07 bendk: probably best to throw it in a bug.
19:07 zanoi: i can look into the dht problem ben is having if that helps.
19:11 z3p: re: 3ae60f41, there are errant print statements in there.
19:11 z3p willkg: ack!
19:11 willkg z3p: it's not a big deal...  i was just skimming the changes and noticed it, so i mentioned it.
19:12 bendk http://bugzilla.pculture.org/s[…]_bug.cgi?id=16419
19:13 willkg zanoi: let me know if you want me to grab 16419 freeing you to work on other things.
19:41 bendk left #miro-hackers
19:41 epitron left #miro-hackers
19:41 epitron joined #miro-hackers
19:41 epitron left #miro-hackers
19:41 epitron joined #miro-hackers
19:48 zanoi willkg: sorry, was travelling back home from cafe
19:52 willkg: do you have a lucid with libtorrent 0.14.10 available?
19:53 willkg zanoi: yup.
19:54 zanoi could you try to change self.session.start_dht() to self.session.start_dt(None) and see if that works?
19:54 willkg: ^
19:55 i'm just guessing so it might not work
19:56 bendk joined #miro-hackers
19:56 zanoi it's in dl_daemon/download.py
19:57 willkg zanoi: sure.  i'll try that now.
19:57 zanoi thanks
19:58 bendk: sorry about the error :-S
20:20 bendk1 joined #miro-hackers
20:21 bendk left #miro-hackers
20:31 zanoi afk for dinner
20:31 bendk1 wow, do I hate trying to make labels wrap right
20:31 willkg bendk1: i hear that.
20:31 bendk1 I've ran my head into this issue like 10 times and each time it kicks my butt
20:31 willkg zanoi: sorry this is taking so long--i had to do updates in my lucid vm.
20:32 bendk1: it's a non-trivial problem.
20:32 bendk1 I guess there's a extended layout branch that makes things work, but that won't help us for a long time
20:32 * willkg nods.
20:32 willkg i vaguely remember there being work done on this in gtk3.
20:33 bendk1 yeah, I think it's in gtk-unstable right now, so that's a start
20:34 willkg two things:
20:34 1) i love the "see crash report" button.  that makes life much easier.
20:35 2) i can reproduce the dht crash you're seeing, ben.
20:35 testing zanoi's fix now.
20:35 bendk1 cool
20:36 kcw left #miro-hackers
20:38 willkg that seems to fix it.  i can't tell if dht is "working" other than that it's not failing, though.
20:41 zanoi yay
20:42 willkg zanoi: your idea works.  i'm going to check that in.
20:42 checked in.
20:42 zanoi great, thanks
20:42 yay, i love submitting patches through irc :-)
20:43 willkg heh.  :)
20:43 also, i finally wrote a git alias for getting the short version of the sha for HEAD.
20:43 i've been meaning to do that for a while now.
20:44 afk a bit....
20:44 kcw joined #miro-hackers
21:11 mattjohnson left #miro-hackers
21:16 DGMurdockIII joined #miro-hackers
21:18 mattjohnson joined #miro-hackers
21:22 janet_ left #miro-hackers
21:53 maggie_s joined #miro-hackers
21:58 willkg z3p: where is the "USB Name" information in the Ubuntu disk utility?
21:58 z3p: is it the volume name?
22:00 z3p: and i don't see the vendor id or product id in the Ubuntu disk utility either.
22:02 ajonas hi maggie_s ! thanks for the round of testing yesterday
22:02 maggie_s: i think paypal is broken in a new way, so nothing new to test on that front, i think
22:02 maggie_s hi ajonas! my pleasure
22:02 ajonas: alas!
22:03 ajonas i think we'll eventually have to figure out what's a UI problem that we can help shore up against with help text, and what's a straight up bug. i suspect there will be both - the hard part is we have limited control over how it will work. But I think we're getting there!
22:13 bendk1 deardiary: working on 15972 and it's taking a bit longer than I thought, but I have a good plan now
22:13 taking off for a while, see you all later
22:13 bendk1 left #miro-hackers
22:14 willkg z3p: also, is there a way to disable video syncing?
22:17 zanoi willkg: is there a reasdon why we are using the fingerprint "MR 1.1" for torrents?
22:17 *reason
22:18 willkg zanoi: i don't know what that means.
22:19 zanoi according to libtorrents doc: The fingerprint class represents information about a client and its version. It is used to encode this information into the client's peer id.
22:20 and we use the string "MR", major version: 1 minor version: 1
22:20 does that mean we identify as miro 1.1?
22:21 willkg that line was written by nassar in 2008.
22:21 give me a few minutes to poke around.
22:21 zanoi sure
22:21 willkg oh...  he did a dos2unix pass.  hrm...
22:22 zanoi heh.
22:22 willkg ok.  that's a line from chris lahey in january 2008.
22:23 and that checkin says that the version of miro is represented in the torrent id.  so MR is for "miro" and 1.1 was the version of miro that we last updated that id.
22:23 and man the codebase is so much better now....  whew!
22:24 zanoi so nowadays we have variables representing the version numbers i hope?
22:24 willkg we did back then, too.  no clue why they didn't use them.
22:24 zanoi ah, ok
22:24 willkg i'll write up a bug for this and fix it.
22:24 zanoi great
22:24 willkg zanoi: good catch.
22:25 zanoi :-)
22:25 btw, if i'm working on something that doesn't affect anyone else would you prefer me to write up a bug for it?
22:26 so that it's documented
22:26 willkg: ^
22:27 willkg can you give me an example?
22:27 zanoi right now i'm working on saving/restoring the torrent session state
22:28 willkg is that related to the magnet uri stuff?
22:28 zanoi well i wanted to add saving/restoring of DHT and that was related to magnet
22:29 but then i noticed we don't even have saving/restoring for sessions
22:29 and ben agreed that i should add both
22:29 willkg yes, that should get a bug.
22:29 zanoi ok
22:29 willkg any enhancements like that should get a bug.  your initial instinct was right on.
22:30 zanoi it's not really a notable enhancement though
22:30 willkg i have a script that looks at all the stuff in bugzilla and generates the release notes.  so, having bugs for everything makes that a lot easier.  otherwise i have to go back-fill bugs.
22:30 well, it is.  it might not be apparent to the user, though.
22:31 generally, if there isn't a compelling reason (i.e. it isn't notable) reason to do something, we should be hesitant to do it at this stage.
22:31 i'm not entirely sure i know what the ramifications of saving/restoring session state is, though.
22:31 zanoi i wonder if anyone has collected statistics of torrent clients  in the past years and wondered why so many people are using Miro 1.1 :-)
22:31 willkg i do know that when implementing the new fast resume data saving, i made a list of a bunch of things we should improve and pushed them all off to 4.1 because we're just way too late in the dev cycle to be making big changes.
22:32 zanoi well i could drop it if you prefer
22:32 it's not really needed
22:33 willkg pretend i don't really know what it entails.  why is it useful to save/restore session state?
22:35 maggie_s ajonas: I have a couple of Duke-related tickets stuck in my testing list (http://bugzilla.pculture.org/s[…]_bug.cgi?id=15075, http://bugzilla.pculture.org/s[…]bug.cgi?id=15926). Do they need any additional work?
22:35 zanoi i thought it would improve performance, but i barely know anything about it
22:35 willkg man...  this is so irritating.  i haven't a clue how to fill out this "miro device support" form and i'm a developer.
22:35 zanoi: i don't know either.
22:35 ajonas maggie_s: you should just ignore them - they're the only ones who can really verify them. i'll take a look
22:35 willkg i guess my gut instinct is that if you don't need it for magnet uris, i'd push it off until later.
22:36 maggie_s ajonas: OK
22:36 zanoi willkg: ok, i'll push it off
22:38 willkg having said that, torrent performance sucks on osx, so it's something that's worth looking into after the codefreeze.
22:40 zanoi just on osx? that's odd
22:41 anyway, i'll write up a bug and keep the work i've done so far but focus on more essential stuff for now
22:41 willkg yeah, it's weird.  i haven't looked into it and don't really know much about it.  pretty sure npr keeps saying it sucks, though.
22:41 and it might just be at starting up.
22:41 ok.  sounds super.
23:06 ajonas maggie_s: i can't reproduce paypal stuff right now (i just get errors), but one thing i wanted to mention: when you downgrade, you shouldn't get your previous payment back, since you've paid for the month ahead
23:07 maggie_s ajonas: I see, but how do we ensure that the payment stops being recurrent and that the admin won't be charged the next month?
23:07 ajonas maggie_s: we have to manually cancel the previous payment.
23:08 maggie_s: er, wait
23:08 when you downgrade, there should be a form that allows you to modify your paymetn
23:08 so paypal changes the recurring amount
23:08 maggie_s ajonas: yes
23:08 ajonas when you upgrade (from another paid tier), you have to create a new payment and we manually cancel your old one
23:09 maggie_s: does that answer your question?
23:09 maggie_s ajonas: frankly, I still cannot visualize how that should work
23:10 ajonas maggie_s: which part?
23:10 downgrade or upgrade?
23:10 maggie_s ajonas: does that mean that the user should be in control of paypal transactions?
23:11 ajonas i'm not sure what you're asking, can you say more?
23:12 maggie_s ajonas: did you mean that MC will be only creating new transactions for the current tier, and the user should remember to cancel older payments manually?
23:12 ajonas maggie_s: no. when someone upgrades, we'll receive a notification, and we (me) will cancel the previous payment
23:13 maggie_s: though they can also cancel their payment from paypal, in which case we'll email them and go, "hey, what's that about?" and if they don't change it, we'll move them down to basic
23:14 maggie_s ajonas: ah, I see - as paypal store owners, we'll have access to transactions and will be authorized to stop payments, right?
23:14 ajonas maggie_s yeah, exactly
23:14 maggie_s: it's not elegant, but it should work
23:16 maggie_s ajonas: now I understand. So, the only things to be tested at this stage are that the tier actually changes and that a new recurring payment is created accordingly
23:17 CarlFK left #miro-hackers
23:17 ajonas maggie_s: yes! well, and that a notification is sent. but i'm not sure the last part is complete to be tested yet.
23:17 maggie_s: and right now, i can't get anything to work in paypal, so, we'll need to deal with that first
23:18 maggie_s: the ustream bug should be testable, with new embeds
23:19 maggie_s ajonas: but those folks who switch from one paid tier to another - they must have already paid once at the beginning of this month and now must pay again with the new tier. So we'll need to return the unused part?
23:19 ajonas maggie_s: *considers*
23:20 maggie_s: i think the way it should work, is that the next payment wouldn't happen until the end of the cycle
23:20 maggie_s ajonas: I believe Asheesh is making the new payment start with the current date
23:21 ajonas maggie_s:i'm not sure that's right - but it's different for upgrades and downgrades. examples:
23:21 if i sign up for max on day 1, pay $75
23:21 then on day 15, downgrade to plus
23:21 the paypal form just changes my recurring amount, so that on day 30 i'm charged $15
23:22 maggie_s: if i sign up for plus on day 1, pay $15
23:22 then on day 15, upgrade to max
23:22 maggie_s ajonas: it doesn't change the payment, actually - it just creates a new one
23:22 ajonas the correct behavior is for there to be a new payment created with a prorated amount
23:22 maggie_s ajonas: at least, now
23:23 ajonas maggie_s: are you able to access it right now? when i tested it yesterday, the form was to modify the payment (which is the correct behavior)
23:24 maggie_s: that said
23:24 paypal is unclear in the way it shows this
23:24 after the fact
23:24 maggie_s ajonas: for me, yesterday it was adding new payments. let me see now...
23:24 ajonas maggie_s: yeah, then that would definitely be a bug
23:24 for http://bugzilla.pculture.org/s[…]_bug.cgi?id=16323
23:25 but i think that's covered in your comment there
23:27 maggie_s: sorry it's so convoluted! figuring out the complicated parts is helpful, though.
23:27 maggie_s: i've gotta run, but i'll be on later
23:27 maggie_s: asheesh is on a plane, so likely nothing will change until tomorrow
23:28 maggie_s ajonas: I've just tried it again - looks like it's still behaving in the same way
23:28 ajonas: creating new payments effective of the current date
23:28 ajonas maggie_s: weird - but also good, because then your comment still stands. for downgrade, right?
23:29 maggie_s ajonas: both ways, I think
23:29 ajonas maggie_s: for upgrade, it should create a new payment as of the current
23:30 maggie_s ajonas: at least, for dalmatia - the other site upgrades for free *smiles*
23:30 ajonas with a pro-rated amount, so if it's the same day just the new amount minus the old
23:30 maggie_s: let me give you admin on anne2 - i think i can't on the others because they're stuck, since i get paypal errors
23:31 maggie_s: but this is good to know.
23:32 maggie_s: actually, that'd likely mess things up, because you'll have a different paypal than me
23:32 maggie_s: i've gotta run, but i would say leave the paypal stuff for now, unless you see new things
23:32 maggie_s: have a good night!
23:32 ajonas left #miro-hackers
23:34 ajonas joined #miro-hackers
23:34 ajonas left #miro-hackers

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