Open 1.4.1: [1.5.1] Dragging a trigger to the trash in Steve's Factory Manager stops the server

[Sketchy Pyjamas]

Version:
1.5.1

What is the bug:
This is actually Infinity server version 1.5.1. Your version drop-down doesn't have that version yet.

I have a bunch of command groups with roughly 60 commands total. I have a input node in each group and a trigger attached to the group.

I wanted to put the triggers inside the groups instead, so I disconnected a trigger and then dragged the trigger into the trash. This causes the server to crash without registering any error. Don't see any exceptions in the server log (update - it's a ClassCastException in SFM - see below). Just shows the server stopping.

When I restart the server, I notice that at least the trigger was deleted.

(Apologies - I missed the crash log when I first posted this. Link below)

Here's an image after 3 of the of triggers were deleted:

UPDATE: This is probably related: https://github.com/Vswe/ModJam3/issues/87
Assuming that's still the authoritative github repository, there's been no activity there for a very long time. This seems to be under active development, though: https://github.com/hilburn/AdvancedSystemsManager

Mod & Version:
Steve's Factory Manager A93

Paste.feed-the-beast.com log:
http://paste.feed-the-beast.com/view/b1f6ff28

Crash Log:
http://paste.feed-the-beast.com/view/96c90eca

Can it be repeated:
Yes, but I've only tried it with the one inventory manager with a bunch of command groups in it.
I did find that it's not necessary to disconnect the trigger from the group before dragging it to the trash. Dragging it to the trash without disconnecting also causes the server to stop.

Known Fix:

 

[Sketchy Pyjamas]

My work-around. I replaced all of the nodes in my command groups with triggers and left the triggers undeleted. I notice that some of the triggers have lost their interval handle, so they can no longer be connected. Related to the server halt?

 

[Photoloss]

I believe I've managed to crash it even with non-Trigger components, in singleplayer 1.5.1. The crash may require the deleted block to once have had a connection (currently connected also counts), and/or being within a Group. A complete reboot of Minecraft allows me to perform the exact same action without any crash, let the game run for a while and it'll cause crashes again.

If deletion of a node doesn't crash the game, often an unrelated piece will vanish for me until I open and close the SFM interface, since both are related to deletion the two issues might be related.

 

[Sketchy Pyjamas]

Yes, I see the disappearing command issue, too. Closing and reopening the SFM GUI fixes it. Deleting a non-empty command block apparently does cause a very similar error according to the github issue I linked to. I suspect it's the same classcastexception problem. I haven't tried deleting the contents of the formerly linked command group to see if that prevents the exception.

On my server, the crash will happen *every* time. I was able to reproduce it right after restarting the server.

You'll note that the github project hasn't received an update for about 9 months, so I doubt we can anticipate a fix any time soon.

Sketch