Git weirdness in settings branch

[Edward K. Ream]

 It looks like all the work at rev f990eb0c277c5 in the settings branch has somehow vanished after innocuous merges of devel into settings.

In the settings branch, git shows the merge, but doesn't show that the any code has disappeared.  But it has, as shown in Leo, and as shown by looking at leoConfig.py in scite.

Happily, the code is still accessible in various ways.  I'm investigating.

Edward

[Edward K. Ream]

Apparently a merge wasn't innocuous.  I'll see what I can do.

Edward

[Edward K. Ream]

On Thursday, September 5, 2019 at 9:12:51 AM UTC-5, Edward K. Ream wrote:

Apparently a merge wasn't innocuous.  I'll see what I can do.

Indeed not. Somehow merging devel into settings was highly toxic.  Two revs in the settings branch restored the proper versions of class ActiveSettingsOutline and c.config.getSource.

Here are the checkin logs:

Rev 597996106:

Restored the proper version of the ActiveSettingsOutline class that somehow got reverted during a "routine" merge.
Restored by copying the class from rev f990eb0c277c5 and pasting it into the tip.
Naturally, this changes gnx.
Note that Leo crashed trying to cut the old class, so something seriously wrong happened.

Rev 18887a56b:

Restored c.config.getSource. So weird

Summary

All seems well again.  In particular, merging devel into settings is once again a do-nothing.

The next step, in the settings branch, will be to complete c.config.settingIsActiveInPath.

Edward