node after root node behaves inconsistently

[telephon]

Syntactically it is possible to add a node after the root node (group with id 0). Normally, nodes are added relative to the default group (id 1), which makes it possible to free them by sending a ´/freeAll´ message to it.
Now sometimes, one may want to keep certain nodes alive even with cmd-period (freeAll), or be able to specify one group where freeAll has no effect (e.g. keeping a recording running of a session).
Actually, as it turns out, nodes that are added after the root node behave in such a way. The problem is only, that they behave really inconsistently. E.g. they show up in the group count, but not in the tree dump of the server.

Some examples:

g = Group(RootNode(s), \addAfter).register;
g.isPlaying;

s.freeAll;
g.isPlaying; // true. OK - useful - groups after the root node survive cmd-period.

g = Group.new; // but obviously this breaks the node id allocator.


// alternative:
s.reboot;

g = Group.basicNew(s, s.nextPermNodeID).register;
s.sendMsg(*g.newMsg(RootNode(s), \addAfter).postcs);

g.isPlaying; 
s.freeAll;
g.isPlaying;

[mossheim]

Tested, still an issue. (macOS 10.11, 3.9-dev)

[vivid-synth]

There are no issues with how scsynth or supernova is handling these nodes though, correct?

[mossheim]

If I understand you correctly, no, I don't think there's an issue with how scsynth is handling these nodes, it seems to be all on the language side with node allocation. The /status.reply messages indicate that the group still exists consistent with the results of Julian's g.isPlaying queries above.