https://bugs.documentfoundation.org/show_bug.cgi?id=171795

ady <[email protected]> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
     Ever confirmed|0                           |1
            Summary|Autofilter reproduces part  |Autofilter shows old values
                   |of the first row as an      |as an extra "half height"
                   |extra "half height" last    |last row when
                   |row when View>ViewHeaders   |View>ViewHeaders is
                   |is "nulled"                 |"nulled" (STR comment 7)
             Status|UNCONFIRMED                 |NEW

--- Comment #7 from ady <[email protected]> ---
This seems to be a visual problem, and therefore, in order to replicate the
issue with attachment 206730, specific visual conditions must be met; which I
include in the following steps.

So, here are the STR:

1. Open attachment 206730 – macros are not relevant for this report.

2. In the AutoFilter of cell AC6, uncheck (i.e. filter out) the "(empty)"
entry; OK.

3. Reduce the zoom factor to 120%. If the last row of the table is not already
(fully) displayed on screen (without having to scroll down), that means that
the zoom factor needs to be reduced even more (in order to be able to replicate
the visual problem). Both the first and the last rows of the table must be
simultaneously displayed on screen (without having to scroll).

4. In the AutoFilter of cell AC6, Sort Descending. Please read the next step
(too) before actually performing this step.
4.1. Note the values being displayed on the right side of the last row of the
table: 1 4 2 3 18 28. Paying attention to these values is relevant to this
report.

5. There are several alternatives for this step:
5.A._ Click on the "Stats" worksheet tab, and then click (back) on the
"Results" worksheet tab; or,
5.B._ Minimize the Window, and restore it; or,
5.C._ Reduce the zoom factor.

5.1. Note the new values being displayed on the right side of the last row of
the table; they are different than before step 5.

Step 5 demonstrates that there is no "real" change of values, but rather a
visual issue. Any of the alternatives in step 5 triggers a visual "refresh".

If you happen to notice a "half last row" (also mentioned in comment 1) after
step 4 (of this comment), it will disappear after performing step 5; another
visual effect that is "solved" by step 5.

There is no need to modify any value in order to see the "new" results. Any
action that will trigger a visual refresh will allow you to see the "real" last
row. This is also the reason that a zoom factor of 200% will prevent you from
replicating the visual problem, because when you scroll down (in order to see
the last row on screen) the screen refresh "cleans" the visual artifacts.

Another alternative step 5 could had been:
5.D._ Hard Recalculate

...but it would not lead to the conclusion that there is no change in values
but rather a visual issue only.

Apparently, having the column and row headers displayed on screen somehow
avoids the visual issue. Perhaps their presence triggers a visual refresh
too(?). This seems to be a key factor.

In the steps above, we have first filtered out the "(empty)" entry and then we
sorted. Performing the sorting action first and then filtering out the
"(empty)" entry, would avoid the visual issue because the very same application
of the filter triggers a visual refresh.

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

Reply via email to