Multileader source block grip point moves?

[Mike.Giglio]

When using grips to move a multileader source block with the attachment defined as insertion point (the insertion point being the center of the source block) the grip point moves off of the block if the move is to the left of the multileaders end point.When the drawing is reopened the blocks move from where they were placed to the off insertion point grips. Anybody experienced this or know how to fix it?

[Mike.Giglio]

Still haven't resolved this issue. Attached a .jpg and a drawing with the mleader in it to help describe the problem. Anybody have any ideas?

[NEaton]

Looks like a bug. I tried modifying a mleader in a new drawing and came up with the same thing. If nothing else maybe you can get an answer back from Autodesk. http://usa.autodesk.com/adsk/servlet...linkID=9240617. You could try posting your question at the Autodesk forums and see if there is any response.
Nancy

Still haven't resolved this issue. Attached a .jpg and a drawing with the mleader in it to help describe the problem. Anybody have any ideas?

[Mike.Giglio]

Thanks Nancy, helpful to know that the problem is not unique to my block or computer. Will give the Autocad forum a whirl.

[autocart]

... Anybody experienced this or know how to fix it?

Yes to both.
My workaround is a command-reactor:
;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
;callback-function:
;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
(defun *GKMM_commandEnded* (#reatype #command-name / $ss $i $enm)
(cond
((wcmatch (car #command-name) "GRIP_STRETCH")
(setq $ss (ssget "_X" '((0 . "MULTILEADER"))))
(and $ss
(setq $i 0)
(while (setq $enm (ssname $ss $i))
(repeat 2 (vla-put-blockconnectiontype (vlax-ename->vla-object $enm) (rem (1+ (vla-get-blockconnectiontype (vlax-ename->vla-object $enm))) 2)))
(setq $i (1+ $i))
)
)
)
)
)

;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
;reactor-loading:
;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
(defun reaload ()
(foreach $vla_reactor (cdar (vlr-reactors :VLR-command-Reactor))
(if (= (vlr-data $vla_reactor) "*GKMM_command-reactor*")
(vlr-remove $vla_reactor)
)
)
(setq *GKMM_command-reactor*
(VLR-command-Reactor "*GKMM_command-reactor*"
'((:vlr-commandEnded . *GKMM_commandEnded*))
)
)
(princ "\*GKMM_command-reactor* activated.\n")
)

(reaload)
;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


I copied it out of context and therefore had to adapt a few lines. This result is therefore untested but I hope it should work.

It still behaves a little buggy. Sometimes you have to stretch it twice in order to achieve the desired result. But for my feeling it is much better than without the workaround.