freenode/#ecl - IRC Chatlog

9:48:46 borodust im using droid studio with llvm debugger

9:49:26 borodust i also found a hack how to redirect stdout/err to logcat

9:50:11 borodust thanks for the info!

11:02:18 ralt is there a library to use the android GUI framework on ECL?

11:03:57 ralt not really interested, just asking when seeing the conversation above

11:09:27 borodust that would imply calling java from C, so i dont think anyone has ever bothered

11:28:04 jackdaniel ralt: EQL5-Android

11:28:09 jackdaniel uses QT5 that has android port

11:38:58 ralt Nice, ok

11:40:08 ralt FWIW: it seems (asdf:make) is broken with wildcard-package-inferred-systems. I haven't yet figured out if it's ECL of w-p-i-s fault yet.

11:41:05 ralt It's also broken with cffi-toolchain:static-program-op. I haven't figured out whose fault it is either, yet.

11:44:39 jackdaniel both were written by fare; I suppsoe that he is the best person to ask, however he seems to not like cl anymore

11:45:17 jackdaniel you may try upstream asdf maintained by r.goldman

12:08:50 ralt I'm playing with it, just wanted to let you know

12:09:08 jackdaniel thanks

12:09:26 jackdaniel my suggestion to try upstream version was to hint a possible solution

13:22:40 ralt yeah I already use that, sorry

13:22:57 ralt wasn't clear

13:28:21 ralt oh, and w-p-i-s is not Fare's https://github.com/privet-kitty/wild-package-inferred-system

15:20:51 aeth_ ** NICK aeth

16:02:28 borodust jackdaniel: qt5 uses native droid ui? damn, im old

16:04:15 borodust actually, it seems like it has natively styled ui rather than using native droid things

16:04:18 borodust https://doc.qt.io/qt-5/qtgui-attribution-android-native-style.html

19:51:37 borodust mhm, actual assert being triggered is "Wrong DATASTART/END pair"

19:52:53 borodust within GC_register_data_segments

19:53:31 borodust i suspect in /* !OS2 && !Windows && !AMIGA && !OPENBSD */ branch of a macro

20:04:35 borodust alright, i guess it's time go hardcore again ;p

21:05:51 borodust okay, sad news is that sigabort is gone after commenting out DATAEND cruft for android+clang codepath

21:09:47 borodust basically, i've checked built .so (via NDK/clang) and it had _end alright

21:11:21 borodust and after commenting out __end__ hack, it all seems to work

21:11:29 borodust not crashes, at the least

21:20:41 borodust but cl_boot returns 1, which looks kinda not good

21:21:53 borodust anyway, gonna cross-compile some simple lisp lib and i'll try to load it into the thing