I have not reproduced this crash using GHC 7.8.2. Doesn't narrow the issue down, but does mean that I have very little motivation to chase this. haha
Cheers, Corey -Corey O'Connor [email protected] http://corebotllc.com/ On Fri, Jan 31, 2014 at 2:00 PM, Corey O'Connor <[email protected]>wrote: > Reproduced the crash without performing output using vty. The UI was still > using vty for input and evaluating the state into a Picture value. The UI > was just not performing the actual blit of the Picture to the terminal > device. This, at least, eliminates much of the vty output layer. > > I would like to eliminate vty entirely from this process. To do so I'm > putting together something like what Mateusz recommends with the Cmcq > keymap. Instead this is at the UI level. The UI will simulate the threading > and evaluation behavior of Vty without actually using any vty library code. > The input sequence is specified at launch; probably using the Cmcq keymap. > The text required for display is fully evaluated on the correct thread. > > Cheers, > Corey > > > -Corey O'Connor > [email protected] > http://corebotllc.com/ > > > On Wed, Jan 29, 2014 at 1:21 PM, Corey O'Connor <[email protected]>wrote: > >> I'm wondering if the core lint error is unrelated. >> The core lint error disappears if the strictness annotation is removed >> from FBufferData's hlCache data member. However, the crash still occurs. :-\ >> >> There wasn't much evidence tying the two together in the first place. So >> eh. >> >> Still investigating... >> >> Cheers, >> Corey >> >> -Corey O'Connor >> [email protected] >> http://corebotllc.com/ >> >> >> On Wed, Jan 29, 2014 at 12:49 PM, Mateusz Kowalczyk < >> [email protected]> wrote: >> >>> On 29/01/14 20:24, Corey O'Connor wrote: >>> > I just tried with the ghc-7.8 branch on both Linux and Mac 64bit. The >>> issue >>> > still occurs. >>> > The ticket states x86-64 architecture and Unknown/Multiple for >>> platform. >>> > >>> > >>> > -Corey O'Connor >>> > [email protected] >>> > http://corebotllc.com/ >>> > >>> > >>> >>> Right, I saw the platform only after I sent the e-mail. I suppose that >>> if this is in fact the bug causing all the trouble and I can't replicate >>> with 32-bit install, it'd explain why I couldn't replicate the Yi test. >>> Hopefully this gets solved quickly and we can move into Vty 5.0 swiftly. >>> >>> >>> -- >>> Mateusz K. >>> >>> -- >>> -- >>> Yi development mailing list >>> [email protected] >>> http://groups.google.com/group/yi-devel >>> --- >>> You received this message because you are subscribed to the Google >>> Groups "yi.devel" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to [email protected]. >>> For more options, visit https://groups.google.com/groups/opt_out. >>> >> >> > -- -- Yi development mailing list [email protected] http://groups.google.com/group/yi-devel --- You received this message because you are subscribed to the Google Groups "yi.devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
