https://bugs.documentfoundation.org/show_bug.cgi?id=149497

--- Comment #8 from Colin <[email protected]> ---
(In reply to Heiko Tietze from comment #7)
> We discussed this topic in the design meeting.
> 
> To rephrase the request, it should be possible to sort data in protected
> sheets (filtering works out of the box). Sorting may affect the data
> integrity, for example if a reference changes after sorting. We should also
> keep in mind that you can limit sort to ranges. So this option would be a
> bit dangerous, yet there might be scenarios where it is acceptable and
> needed.
> 

> > Not completely worthless, though, because such a design and functionality
> > can still be used as bad examples in textbooks to contrast with the
> > fundamental engineering principle of least astonishment. 
> :-)

Not sure if a procedure I currently use would be included as bad example but it
certainly obviates the known shortcoming of ranges incorporated into formulae
that are then sorted. It couldn't work over vast ranges but I find that it's
adequate for smaller selections.
it'a pointless having a formula referring to a "range" which can be disrupted
by a sort where those ranges tend to be relative to the current location, but
if the referenced "range" is only a smaller matrix then I find that instead of
something like =SUM(AA10:AA20) where the rows can end up all over the sheet -
it achieves the correct objective if the formula is modified to
=AA10+AA11+AA12..... or even =SUM(AA10;AA11;AA12;AA.......) where each
individual cell is then tracked to its new location and therefor correctly
summed.
Imagine having a few thousand lines which are really consistent daily events in
clusters of 8. These clusters can be replicated and retain both their
"internal" and "external" integrity so even =SUBTOTAL() works perfectly if some
of the daily events are filtered out.
Not sure if I'm adding value to the conversation or just a pain in the derriere
but it works for my analysis of "repeating" events closely related to each
other and without calculations traversing rows.

-- 
You are receiving this mail because:
You are the assignee for the bug.

Reply via email to