#pyqt | Logs for 2018-11-28
Back
[00:08:00] -!- Dave_Elec has quit [Ping timeout: 268 seconds]
[00:08:46] -!- Dave_Elec has joined #pyqt
[00:16:27] -!- Dave_Elec has quit [Ping timeout: 240 seconds]
[00:22:37] -!- Dave_Elec has joined #pyqt
[00:31:38] -!- Dave_Elec has quit [Ping timeout: 250 seconds]
[02:46:50] -!- kallesbar has joined #pyqt
[02:52:00] -!- Lauxley has joined #pyqt
[05:40:45] <ysangkok> The-Compiler: please do... i don't know how to elaborate on this
[05:48:03] <The-Compiler> ysangkok: https://www.riverbankcomputing.com
[05:50:14] -!- mintograde has joined #pyqt
[06:02:27] <ysangkok> ah thank you, you're a saint
[06:41:16] -!- mintograde has quit [Ping timeout: 250 seconds]
[10:04:02] <The-Compiler> ysangkok: "Should be fixed in tonight's sip snapshot. It's a sip module change - no need to re-build PyQt."
[10:10:36] -!- Siecje has joined #pyqt
[10:11:11] -!- beeman has joined #pyqt
[10:18:11] <ysangkok> awesome
[10:38:44] -!- Lauxley has quit [Ping timeout: 250 seconds]
[11:01:38] -!- Lauxley has joined #pyqt
[11:41:26] -!- at1as has joined #pyqt
[11:42:13] <at1as> hello all. i realize that PyQt5.5 changed the exception-handling behavior, to cause unhandled exceptions to abort().
[11:42:52] <at1as> is there a way to reverse this behavior to the previous behavior? perhaps by replacing sys.excephook() or some other method?
[11:44:51] <at1as> the users of my tool appreciate that their work is not lost because they run into a bug. i've upgraded the tool to use PyQt5, and believe i have handled all the important exceptions... but i'd hate to be wrong and have people's work be lost because we've become used to this safety-net.
[11:46:47] <ysangkok> you can LD_PRELOAD and replace abort
[11:46:53] <ysangkok> but that's the last resort hack i guess
[11:53:51] <at1as> lol
[11:54:04] <at1as> yup, i'd like something i can do within the Python context.
[11:56:22] <The-Compiler> at1as: exactly what you propose, setting sys.excepthook
[12:00:24] <at1as> thanks The-Compiler! any tips for what i need to do differently in excepthook? does PyQt5 rewrite sys.excepthook() to call qAbort? so simply by replacing it with some handling code, i avoid the abort?
[12:00:37] <at1as> sorry, qFatal
[12:01:07] <at1as> (http://pyqt.sourceforge.net/Docs/PyQt5/incompatibilities.html)
[12:02:37] <The-Compiler> at1as: dunno, a good thing to do would probably be to save some kind of backup and quit with an error
[12:06:57] -!- altendky has joined #pyqt
[12:07:34] <altendky> The-Compiler: so you've been going through the type hinting stuff. did you figure out anything about why mypy could handle the module reference for some things, but not others (such as qutebrowser)?
[12:10:39] <The-Compiler> altendky: I don't understand the question - do you mean the issues amosbird was mentioning in #python? No issues here.
[12:10:51] <The-Compiler> (but I invoke mypy via tox which installs everything properly as a package)
[12:11:06] <altendky> The-Compiler: yeah, i got it as well. let me try again instead of trying to remember
[12:12:01] <altendky> i do intend to have things installed when i fiddle with them...
[12:12:12] <The-Compiler> I'm mainly trying to figure out how to get useful stubs for mypy without having to rewrite everything by hand like https://github.com does (or at least reducing the rewriting needed, if there's a solution which works for PyCharm and mypy)
[12:12:31] <The-Compiler> because I figured out that those stubs are for Qt 5.7 and updating them is a pain
[12:12:53] <altendky> yeah, manual is silly at best (even if sometimes what we do)
[12:15:57] <altendky> https://gist.github.com
[12:16:26] <altendky> The-Compiler: so line 54 shows the error, let me see if i can get it to work with the path specified though
[12:17:15] <altendky> line 58 seems to be suggesting mypy wants it to be _not_ installed
[12:18:09] <altendky> oh, and line 52 shows it not complaining about attr
[12:20:06] -!- at1as has quit [Ping timeout: 244 seconds]
[12:21:44] <altendky> https://gist.github.com added y.py where i clone the qutebrowser source. so if i specify the not-installed source path i get an analysis. if i cd in and `-m qutebrowser` i get no output (like attr did)
[12:25:19] <The-Compiler> altendky: the PyQt5 stub files won't work anyways, so you'll need to install PyQt5-stubs (or just use "tox -e mypy")
[12:25:34] <The-Compiler> anyways, afk for a bit, should go on writing my research paper :)
[12:26:50] <altendky> i would think the stub files would relate to the long output of complaints, not the module not found stuff. i assumed the complaints meant it actually found the thing to analyse and did and has issues with it. i like that :] but yes, work away.
[13:34:26] -!- TechSmurf has joined #pyqt
[15:20:25] -!- beeman has quit [Quit: Connection closed for inactivity]
[15:49:27] -!- beeman has joined #pyqt
[17:24:39] -!- mintograde has joined #pyqt
[17:27:32] -!- Lauxley has quit [Remote host closed the connection]
[17:59:02] -!- Siecje has quit [Quit: Leaving.]
[18:29:49] -!- richar_d has joined #pyqt
[20:40:25] -!- beeman has quit [Quit: Connection closed for inactivity]
[20:54:52] -!- richar_d has quit [Quit: richar_d]
[21:39:34] -!- mintograde has quit [Ping timeout: 250 seconds]
[21:56:27] -!- Stinger7 has joined #pyqt
[22:02:33] -!- Stinger7 has quit [Quit: leaving]