All times shown according to UTC.
| Time | Nick | Message |
|---|---|---|
| 00:08 | janetPCF joined #miro-hackers | |
| 00:10 | janet_ left #miro-hackers | |
| 00:18 | janet joined #miro-hackers | |
| 00:43 | ajonas left #miro-hackers | |
| 00:45 | Sergei left #miro-hackers | |
| 00:58 | ajonas joined #miro-hackers | |
| 02:10 | zanoi left #miro-hackers | |
| 02:11 | zanoi joined #miro-hackers | |
| 02:16 | zanoi | d'oh, it's obvious |
| 02:16 | ./run --unittest feedparser runs the module feeduparser which contrains two classes with each 2 tests | |
| 02:16 | ./run --unittest FeedParser runs only the class FeedParser which only contains 2 tests | |
| 03:32 | ajonas left #miro-hackers | |
| 04:29 | thrillERboy joined #miro-hackers | |
| 04:49 | DGMurdockIII left #miro-hackers | |
| 05:38 | CarlFK left #miro-hackers | |
| 07:18 | glee left #miro-hackers | |
| 10:38 | Dawnrell joined #miro-hackers | |
| 12:11 | thrillERboy left #miro-hackers | |
| 12:18 | Dawnrell left #miro-hackers | |
| 12:35 | z3p left #miro-hackers | |
| 13:43 | CarlFK joined #miro-hackers | |
| 14:01 | zanoi | is the EventLoopTest class necessary for testing classes that use the event loop or is it only necessary if the test class itself wants to use the event loop? |
| 16:44 | bendk1 | zanoi: hi there, sorry I've been away |
| 16:44 | you still wondering about EventLoopTest? | |
| 16:46 | zanoi: EventLoopTest gives you the option of running the eventloop, so it's good for testing code that uses the eventloop | |
| 16:52 | willkg joined #miro-hackers | |
| 17:39 | zanoi | bendk1: nevermind, i solved the problem |
| 17:43 | note to self: not calling eventloop.shutdown() in a callback will make a test fail :S | |
| 17:43 | bendk1 | zanoi: there's a method of EventLoopTest for that |
| 17:43 | self.stopEventLoop(abnormal=False) | |
| 17:44 | good to hear you figured it out, sorry I wasn't around before | |
| 17:44 | feel free to ask more questions as they come up | |
| 17:49 | zanoi | bendk1: thx |
| 17:57 | bendk1: so do you mean i can replace eventloop.shutdown() with self.stopEventLoop()? | |
| 17:57 | bendk1 | yeah, that should work |
| 17:57 | zanoi | because using self.stopEventLoop() instead of eventloop.shutdown gives me an error |
| 17:57 | bendk1 | you might need the abnormal=False parameter |
| 17:58 | zanoi | oops |
| 17:58 | right :) | |
| 17:58 | bendk1 | when you start the eventloop, we set a timeout so that if it's not stopped then the tests keep running, but we mark a failure |
| 17:58 | so you need to signal that that's not the case | |
| 17:59 | going to lunch, be back in a bit | |
| 17:59 | zanoi | yes, makes sense |
| 17:59 | kk | |
| 18:00 | is runEventLoop() is a blocking call which can only be stopped with a callback? | |
| 19:00 | Dawnrell joined #miro-hackers | |
| 19:08 | bendk1 | back |
| 19:08 | zanoi: yes it's blocking | |
| 19:09 | really what happens is that runEventLoop() starts processing the eventloop, so the thread that your unittest is running in is now just processing eventloop callbacks until the eventloop gets stopped somehow | |
| 19:14 | zanoi | yup, thats what i thought, took me quite a while to figure out it wasn't asynchronous though |
| 19:38 | DGMurdockIII joined #miro-hackers | |
| 19:49 | bendk joined #miro-hackers | |
| 19:50 | bendk1 left #miro-hackers | |
| 20:04 | willkg | zanoi: hi! :) |
| 20:05 | bah--i just discovered that get/set_main_window_dimensions are really just saving/loading from config. irritating. | |
| 20:29 | Odysimus left #miro-hackers | |
| 20:36 | zanoi | willkg: hey |
| 20:40 | Odysimus joined #miro-hackers | |
| 20:42 | paroneayea | paulproteus: ping |
| 20:56 | zanoi | could anyone explain me what the site_tabs parameter in Exporter.export_content() is meant to be? |
| 21:21 | willkg | zanoi: if you run Miro, in the sidebar on the left side, there are media tabs and site tabs. each site is a website. |
| 21:22 | zanoi: if you look at _write_site_entry, it's pretty much just saving the url and title of the web-site. | |
| 21:25 | zanoi | willkg: thx |
| 21:26 | willkg: i'm wondering though which class it expects to receive | |
| 21:27 | willkg | zanoi: so if you trace it back, it's looking at tabs.TabOrder.site_order. that's a class method that returns view_for_type(u'site'). |
| 21:29 | zanoi: and i'm pretty sure that's just a Tab of type site. | |
| 21:29 | er, a TabOrder of type site. | |
| 21:38 | zanoi | willkg: as in a just TabOrder("site") instance? because that's was my conclusion as well but I can't get it to work |
| 21:38 | i'm probably just doing it wrong though | |
| 21:38 | willkg | can't get what to work? |
| 21:39 | zanoi | creating a TabOrder and passing it as the site_tabs parameter |
| 21:40 | willkg | pastebin the code you're trying. |
| 21:43 | oh, whoops. looks like site_tabs is not a list of TabOrder objects, but rather the sites themselves. if you look at opml.Exporter.export_subscriptions, it's passing in actual sites. | |
| 21:43 | those are probably ChannelGuide instances. | |
| 21:44 | looking at lib/test/opmltest.py TestExporter.test_simple is probably a good example using feeds which get passed into media_tabs. | |
| 21:44 | zanoi | been staring at that for a while now :) |
| 21:45 | but ChannelGuide instances sounds great, I'll have a look! | |
| 21:46 | willkg | ChannelGuide class is defined in guide.py. a while back we called them "Channel Guides" in the interface, but then we renamed them to "sites" but didn't carry the renaming through the codebase. |
| 21:46 | similarly "feeds" == "channels". | |
| 21:46 | afk a bit.... | |
| 21:46 | zanoi | that explains why I couldn't find any class with Sites in its name :) |
| 21:47 | willkg | sorry about that. |
| 21:47 | legacy hoo-hah. | |
| 21:47 | zanoi | heh |
| 21:47 | willkg | afk for realz.... |
| 21:52 | zanoi | yay, ChannelGuides was the solution! |
| 22:02 | the site_tabs name for a ChannelGuide list is quite unlucky though... | |
| 22:42 | willkg | zanoi: it's not unlucky so much as it's just legacy stuff. |
| 22:46 | z3p joined #miro-hackers | |
| 22:56 | zanoi | willkg: right |
| 23:11 | Dawnrell left #miro-hackers | |
| 23:23 | zanoi | willkg: sorry, am having another problem |
| 23:23 | calling Exporter.export_subscriptions() in a unit test fails with a "Can't find singleton" exception.. | |
| 23:23 | do I have to create the singleton manually? other tests don't do that as far as I can see.. | |
| 23:36 | willkg left #miro-hackers | |
| 23:39 | willkg joined #miro-hackers | |
| 23:44 | willkg | rrr... hung my machine. |
| 23:45 | paroneayea | hrm |
| 23:47 | willkg: I'm getting an error in config.py from a commit you made today | |
| 23:47 | willkg | paroneayea: pastebin it? |
| 23:47 | paroneayea | yeah |
| 23:48 | this is the commit: | |
| 23:48 | http://dpaste.com/289979/ | |
| 23:48 | and this: | |
| 23:48 | http://dpaste.com/289980/ | |
| 23:48 | is the error | |
| 23:48 | willkg | did you do a fetch? |
| 23:48 | paroneayea | basically you've got the wrong number of arguments there |
| 23:48 | willkg | i fixed it in a subsequent checkin. |
| 23:48 | paroneayea | oh |
| 23:48 | :) | |
| 23:48 | I'll do one now :) | |
| 23:48 | willkg | it was bad code-fu on my part. |
| 23:49 | * paroneayea | a bad bug reporter! :) |
| 23:49 | willkg | oh no... you're doing fine. |
| 23:49 | totally my fault. it never should have made it to the repository. | |
| 23:50 | i was doing some work so that you could run miro on linux in a sandbox without affecting your configuration or your setup. | |
| 23:50 | so you can do this now: | |
| 23:50 | ./run.sh --home=/tmp/miro --gconf-name=mirotest | |
| 23:50 | makes my life a lot easier. | |
| 23:51 | paroneayea | willkg: oh nice :) |
| 23:51 | that's awesome | |
| 23:56 | willkg | sorry about that. |
| 23:57 | the other thing i just "fixed" is that miro in fullscreen only shows the overlay if you move the mouse to the bottom of the screen. |