Re: [fossil-users] db errors after Ctrl-C

Unfortunately, the bug is a proper Heisenbug, and refuses to manifest at all under the debugger. I read up on the behaviour of Ctrl-C under Windows, here: https://msdn.microsoft.com/en-us/library/windows/desktop/ms682541%28v=vs.85%29.aspx and it's supposed to work like this: 1) If no handler is

Re: [fossil-users] db errors after Ctrl-C

On 12/13/16, rosscann...@fastmail.com wrote: > SNIPPET 1 > C:\Users\Ross\src>fossil commit -m "created" > ./math/src/transforms.rs contains mixed line endings. Use --no-warnings > or the "crnl-glob" setting to disable this warning. > Commit anyhow (a=all/c=convert/y/N)?

Re: [fossil-users] db errors after Ctrl-C

OK, I tried installing an interrupt handler that ignores the interrupt. This fixes my original problem, but of course means that you can never interrupt a long-running operation! Not really viable. It looks to me like main() needs an interrupt handler that works like shell's: sets some kind of

Re: [fossil-users] db errors after Ctrl-C

I think shell.c is used for "fossil sqlite3" mode, not for what I'm doing. For normal operations, the main() in main.c is invoked, and that main() doesn't install any kind of interrupt handler. It looks to me like main.c needs to install a sigint handler, like shell.c does. I'm going to

Re: [fossil-users] db errors after Ctrl-C

On Dec 13, 2016, at 4:20 AM, rosscann...@fastmail.com wrote: > > I noticed a minor bug in fossil. Yeah, I think I agree. > I can do some detective work here Try this: add #error Yo! on line 891 of src/shell.c, immediately after the ifdef for SIGINT. I think you’ll find that it builds

[fossil-users] db errors after Ctrl-C

I noticed a minor bug in fossil. Observed on Windows 10 with all of: 1) fossil 1.35 and 1.36 (downloaded executables): 2) built from source (MSVC build, latest code as of 2016-12-11) I am in the open checkout of a local (file) repository. I have added a large number of files. Some have mixed line