**** BEGIN LOGGING AT Sat Oct 10 02:59:59 2015
Oct 10 09:47:29 <Tofe> Morning
Oct 10 10:08:10 <Herrie> Tofe: morning
Oct 10 10:32:26 <Tofe> Herrie: do you know LS2 well ? I will need debugging it a bit, because I send an LS2 message but never get the answer...
Oct 10 10:33:27 <nizovn> maybe wrong bus?
Oct 10 10:35:43 <Tofe> Well, it's PalmServiceBridge, and it seems subsribed to private and public bus
Oct 10 10:36:30 <Tofe> I'm trying a wrong call willingly, in order to get an error response
Oct 10 10:37:32 <Tofe> something like call("palm://luna.test.com/display", "{}") should trigger a response, right?
Oct 10 10:37:49 <Tofe> "the URI doesn't exist", for instance
Oct 10 10:39:11 <Tofe> doing it with luna-send does trigger an answer, at least.
Oct 10 10:39:22 <nizovn> i think it should
Oct 10 10:40:58 <Tofe> oh, wait, my identifier is empty, maybe that's the problem here
Oct 10 10:43:34 <Herrie> Tofe: Yeah I think you should get an error
Oct 10 10:43:50 <Herrie> But I recall DougReeder reported a bug that he doesn't always get an expected reply as well
Oct 10 10:43:53 <Herrie> Let me look it up
Oct 10 10:45:01 <Tofe> Ok, the identifier thing was just because my debug was not complete; now I have one, I'm even priviledged, and it doesn't answer either
Oct 10 10:45:27 <Tofe> I only have a listener token
Oct 10 10:50:48 <Herrie> Tofe: This is what DougReeder raised: http://issues.webos-ports.org/issues/719
Oct 10 11:02:02 <Tofe> ok thanks
Oct 10 11:33:39 <Herrie> Tofe: Doesn't really seem to be related but might be good to keep in mind
Oct 10 12:54:31 <Tofe> Herrie: I think I understood the issue
Oct 10 12:56:06 <Tofe> Webkit is based on GLib, and uses the GLib event loop methods; LS2 also uses GLib and its event methods to process events
Oct 10 12:56:28 <Tofe> but Chromium has his own event loop, not based on GLib, so we never pump the LS2 messages
Oct 10 12:56:49 <Tofe> Not sure I'm correct, but it's worth checking.
Oct 10 13:38:18 <Herrie> Tofe: Not sure never looked to much into the event loop methods tbg
Oct 10 13:38:22 <Herrie> s/g/h
Oct 10 17:35:26 <Herrie> Hey dkirker!
Oct 10 18:01:14 <DougReeder> My originial bug title was inaccurrate; a better title would be "attempting to launch non-existant app does not return same errorText as webOS"
Oct 10 18:03:04 <DougReeder> When there’s an error return, we should always return *something* in errorText.
Oct 10 18:10:32 <Tofe> I've inserted a GLib event loop within Chromium's event loop, and it seems to do the trick. Now I crash when interpreting the response, but I get one at least :)
Oct 10 19:13:17 * elvispre has a new LuneOS build VM. This time with a 200 GB disk!
**** ENDING LOGGING AT Sun Oct 11 02:59:58 2015