**** BEGIN LOGGING AT Sat Dec 31 02:59:56 2022
Dec 31 11:02:09 <Tofe> Morning!
Dec 31 11:02:20 <Tofe> Herrie: interesting stack for the crash of calendar: https://bpa.st/EHYO2
Dec 31 11:02:39 <Tofe> looks like the "bad variant" hint during a LS2 call was the right one
Dec 31 11:04:24 <Herrie> Tofe: Just curious what gets passed there that causes such havoc
Dec 31 11:05:41 <Herrie> It shouldn't crash things ungracefully like that I would say? There should be some validation or sanity checking happening, no?
Dec 31 11:14:59 <Tofe> maybe a try...catch around here https://github.com/webosose/chromium94/blob/master/src/neva/injection/renderer/webosservicebridge/webosservicebridge_injection.cc#L59 would be nicer
Dec 31 11:15:37 <Tofe> this json API is quite fragile yes
Dec 31 11:15:51 <Tofe> (and I never liked exceptions :/ )
Dec 31 11:16:16 <Tofe> but I feel like we already had a patch for this kind of crash?
Dec 31 11:17:50 <Tofe> can't find anything in our logs, maybe it was something else
Dec 31 11:18:58 <Herrie> Tofe: Yeah I also recall something but might have been with webengine still?
Dec 31 11:20:00 <Tofe> I think so too
Dec 31 11:20:36 <Herrie> But maybe better fix the JSON if possible?
Dec 31 11:21:41 <Tofe> https://github.com/webOS-ports/meta-webos-ports/blob/a1dce7795902c8a66896bffad7ffc4a5904ba054/meta-luneos/recipes-webos/chromium/files/0006-WebOSServiceBridgeInjection-fix-JSON-input-test.patch
Dec 31 11:23:37 <Herrie> Quite sure I forward ported that one... Maybe did it wrong somehow?
Dec 31 11:23:43 <Tofe> very similar location, but we still have the patch yes
Dec 31 11:24:24 <Tofe> mmh wait a bit
Dec 31 11:24:56 <Tofe> isn't the patch reversed somehow ? it's a bit weird
Dec 31 11:26:20 <Tofe> https://github.com/webOS-ports/meta-webos-ports/commit/fe75de15a20c9bc12865b0e163c3fe44960d69d7#diff-04464469cf97fdf70a6b9ac93fd9d0b462dfad0456fdb3edebc6c5b2f5a536b8 yes, it was reversed, it should be "||" in the final code
Dec 31 11:27:01 <Tofe> maybe upstream already applied our fix, and our patch didn't apply anymore
Dec 31 11:27:25 <Tofe> anyway, I think it's very probably the issue here
Dec 31 11:29:08 <Tofe> I'm just testing by hand, then I'll propose a PR without this patch
Dec 31 11:30:41 <Herrie> Tofe: Well or I broke it, quite possible
Dec 31 11:30:44 <Tofe> yes, works
Dec 31 11:31:04 <Herrie> Rebasis patch on newer Chromium wasn't easy at times, so could be I slipped up somehow
Dec 31 11:34:12 <Tofe> https://github.com/webOS-ports/meta-webos-ports/pull/561 here it is, I'm now rebuilding webruntime
Dec 31 11:37:01 <Tofe> Well, the rebase was quite some work, so it's easy to miss some things
Dec 31 11:37:20 <Herrie> After you rebuild and it's OK, go ahead with merge
Dec 31 11:37:38 <Herrie> For sure and this was an "easy" one to port LOL
Dec 31 11:37:56 <Herrie> Just sometimes not clear when reading patch and manually applying
Dec 31 11:38:04 <Herrie> Good it got fixed upstream anyway
Dec 31 11:38:20 <Herrie> Seems we were not the only ones running into it :D
Dec 31 11:43:25 <Tofe> yup
Dec 31 14:14:13 <Tofe> also, it looks like the "nowindow" attribute isn't respected anymore
Dec 31 16:51:01 <Herrie> Tofe: Didn't we have a patch for that too?
Dec 31 17:11:12 <Herrie> It could be they removed noWindow handling from OSE side in one of the many component updates somehow
Dec 31 17:24:40 <Tofe> Could be
Dec 31 17:28:49 <Tofe> the main commit in WAM is still there, at least
Jan 01 00:50:32 <Herrie> Happy New Year!
**** ENDING LOGGING AT Sun Jan 01 02:59:56 2023