measured values issue

[RussL]

We save our programs.
This allows us to re-align the part after the fact (after it's gone on to the next operation).
I've found that sometimes it works & sometimes it does not.
Say we want to shift a program .010 in "X"
We add an alignment with a X +.010 shift and say no to prompt.
This keeps nominals unchanged, but actuals should all change -.010.
Off-line (on different PC) sometimes it works, sometimes it does not.
Today it did not.
I took the same program and did the same procedure on the machine where the part was originally inspected (off-line).
It worked as it should - nominals remained the same, measured values changed by the shift.
Why would it work off-line on the machine the part was inspected on and not off-line on a different PC?
V3.7 MR2 - XP SP2

[Matthew D. Hoedeman]

We save our programs.
This allows us to re-align the part after the fact (after it's gone on to the next operation).
I've found that sometimes it works & sometimes it does not.
Say we want to shift a program .010 in "X"
We add an alignment with a X +.010 shift and say no to prompt.
This keeps nominals unchanged, but actuals should all change -.010.
Off-line (on different PC) sometimes it works, sometimes it does not.
Today it did not.
I took the same program and did the same procedure on the machine where the part was originally inspected (off-line).
It worked as it should - nominals remained the same, measured values changed by the shift.
Why would it work off-line on the machine the part was inspected on and not off-line on a different PC?
V3.7 MR2 - XP SP2

Funny thing about that.....

Pre V3.5, that is EXACTLY the way it worked, ALL THE TIME. Then with the advent of V3.5, they changed the way the alignments tie into features/dimensions/TABLE ZERO POSITION/etc. Pre V3.5, you could change the X offset by 0.010" and after telling it NO, DO NOT UPDATE NOMINALS TO NEW ALIGNMENT, the actuals WERE all shifted based upon the change you made to the program. This was VERY handy when trying to balance a new fixture between part datums and the call-outs on the base of the fixture. You SHOULD still re-measure after the shift as anything with more than a single vector will NOT completly update correctly, but it sure let you know if you were going in the right direction.

[jimfly]

I also has the same question once time.
Measured online,then modified offline at another PC,but can't work when return to online .