Quote from: gavrilo77 on November 10, 2017, 06:14:18 PM

Quote from: gjhiggins on November 10, 2017, 06:05:35 PM

Quote from: gjhiggins on November 09, 2017, 11:28:23 PM

Quote from: ylpkm on November 09, 2017, 09:20:27 PM

if you need it compiled for windows I can try if the dependency list isnt crazy to get working.

Thanks for the offer but I'm okay - I've updated the candidate Windows binary on minkiz: ~~https://minkiz.co/noodlings/slm/slimcoin-qt-v0.5.0-win32.zip~~
I haven't been able to check specifically whether the fault persists.

Welllll, I'm pleased to report that it was another case of moar ells ...

https://i.imgur.com/EJOwVFX.png

and the zip file with updated Windows binary is:

https://minkiz.co/noodlings/slm/slimcoin-qt-v0.5.1-win32.zip

i.e. the previous link was to a zip file with a current datestamp but unchanged contents. I had omitted to clean out the release subdirectory before recompiling, sigh. I nearly missed that, wouldn't have noticed had it not been for the fact that I'd corrected a typo in the print string immediately above my edit and the debug log was still showing the typo. I did it twice (cross-compiled the binary on the old Linux laptop, uploaded it to the server, downloaded it to the Windows VM), omitted to clean the release directory twice, typo showed up twice. WTF? It was only then that I started backtracking, checking my assumptions - finally realising that the zip file content was remaining unchanged. Hence the 0.5.1, for clearer differentiation.

Sorry if I wasted anyone's time with that boo-boo.

Cheers

Graham

Lovely, crash not any more!!! Thank you Graham!!!

Edit: Once again, the new update works well, PoB and PoW. No crashing. Still testing

I am very happy to inform you that after compiling the daemon again in a raspberry pi 3 with Raspbian with the last changes that gjhiggins has introduced in the code, the PoB finally works without closing the application. Since the version of Slimcoin that eddycurrent has been compiling for the different Arm devices was not fully functional I have thought, if the community agrees, that the bounty offered by the functional Arm versions will be divided into two, 5000SLM for gjhiggins to resolve the failure of the PoB and 5000SLM for eddycurrent when recompile the Arm v5 and v7 versions again and provide the v8 all with the new code. I wait for your comments.

Cheers.