There are examples of rna_foo_set()
that return and do WM_report(RPT_ERROR, ...)
, e.g. rna_ActionSlot_name_set
My understanding was that we would only set it readonly if it is "dirty" (so had modifications, painted on, ...)
Also noticed that sorting vertexgroups got broken between 3.6 LTS and 4.0 (will probably report this separately)
Reported in #130686
Also noticed that sorting vertexgroups got broken between 3.6 LTS and 4.0 (will probably report this separately)
Got it patched sorting bones, but needs a lot of fine-tuning/checking if this breaks anywhere, will put on my desk for now...
#107851 is related, also mentions the corresponding code comment (from ancient code)
/* Sorting rules; only object lists, ID lists, or deform-groups. */
Could we find the commit that fixed this (so we could have this in a corrective release?
However, there does seem to be an attribute propogation problem. ... I need to investigate the attribute propagation logic when I get some time.
Will confirm then for now