Question on offsets (tetrahedral numbers)

[Marc LeBrun]

 >=Gene Smith
 >>=Franklin T. Adams-Watters
 >> So, should one submit a correction for a case like this?  Just ignore it?
 >> Or ask this group (which I guess is what I am doing for this case)?
 > I'm new on the list, but it seems to me  to be something you could 
submit a correction for.

I concur.

It's good to see concern with this.  Many transforms (eg Mobius) are 
exquisitely
sensitive to offset.

Having an incorrect offset will prevent programs (say, superseeker) 
from finding
valuable results.

As far as padding and getting the "right" offset, I'd suggest that the best
choice is the one that gives the most "natural" formula (or similar mechanical
definition) for a(n).

But I'd use great care in changing existing sequences, unless it is to correct
an outright error, because direct and indirect references to the original may
have accreted, both inside and outside of the OEIS.

If there's a more aesthetic version, or an equally-valid variant form, it's
better to submit it as well, cross-referenced, rather than thrash an existing
entry.

This issue can crop up, for example, when extending a sequence "backwards" to
include an a(0) value--sometimes there are plausible alternatives 
with different
rationales (eg one might give a nicer formula, another might have a more
intuitive concrete combinatorial interpretation).

If "the" right answer is multi-valued, then there should be multiple sequences
in the OEIS.