BugSplat view report details doesn't work on Windows 10?

When one gets a BugSplat, the dialog has a button at the bottom to View Report Details. On Mac this opens a view of the crash report. On Windows the button is inactive (won’t take focus) and does nothing if clicked. Does anyone know whether BugSplat keeps a crash log after the report on Windows, and if so, where it lives on the disk?

I found an old discussion that you took part in. The answer given then was %temp%. I tried using Task Manager to simulate a crash, but didn’t get a bugsplat, or see a new file in that folder. But, if you did get a bugsplat, you should see something new in %temp%.

In that topic, I identified where the crash reports are kept on Mac. I believe @eneroth3 said to look in %temp% on Windows but then that it didn’t work?

I’m definitely getting BugSplats from SU 2020 on Windows 10. It’s not just because of the Parallels VM I use, a regular Windows user sees it while testing one of my extensions too. Alas, my last deep dive into Windows was about 8 years ago! How do I get to %temp% to look for crash files there?

Edit: OK I found it. I see .dmp files and .log files from SketchUp. The .dmp files are binary that I don’t know how to read. The .log files don’t give a clue what caused the BugSplat.

Someone else mentioned %temp%, and I do see a lot of log files in there.

On Mac you know that ~/Library gets you into the user’s library without needing to know the user name, it’s handy for when giving paths to people. With Windows there are a number of shortcuts on those lines. The most useful for SketchUp paths are %AppData% and %LocalAppData%.

Those, and %temp%, can be typed into the path field in File Explorer. %LocalAppData%\temp would be the same as %temp%.

I found that .dmp files open Visual Studio. Not that I know what I’m looking at when I get there!

Yeah, I think I’m out of my pay grade trying to analyze the dump on Windows. I submitted two BugSplats, so hopefully somebody from the team can give me some insight into what I’m doing wrong.

I see the bugsplat reports, and your notes should be enough to go on.

This topic, that you were in on, may be relevant. The link is to Dan’s Ruby work around: