Bug 79450

Hi @ap, thanks for reporting.

Please can you attach attach a screenshot to see what is bad for you and so verify if we see the same?.

Created attachment 100201 [details]
attachment-31675-0.html

Here is attached:
- the Calc file,
- what it looks like on my computer.

On this occasion, many thanks for your job on Open/LibreOffice!


2014-05-31 0:45 GMT+02:00 <bugzilla-daemon@freedesktop.org>:

>  m.a.riosv <mariosv@miguelangel.mobi> changed bug 79450
> <https://bugs.freedesktop.org/show_bug.cgi?id=79450>
>  What Removed Added  Status UNCONFIRMED NEEDINFO  CC
> mariosv@miguelangel.mobi  Ever confirmed   1
>
>  *Comment # 1 <https://bugs.freedesktop.org/show_bug.cgi?id=79450#c1> on
> bug 79450 <https://bugs.freedesktop.org/show_bug.cgi?id=79450> from
> m.a.riosv <mariosv@miguelangel.mobi> *
>
> Hi @ap, thanks for reporting.
>
> Please can you attach attach a screenshot to see what is bad for you and so
> verify if we see the same?.
>
>  ------------------------------
> You are receiving this mail because:
>
>    - You reported the bug.
>
>

Created attachment 100202 [details]
attachment-31675-1.dat

Created attachment 100203 [details]
Calc_centered_bubbles.ods

Created attachment 100204 [details]
Calc_centered_bubbles_Screen_Copy.jpg

Created attachment 100205 [details]
attachment-32654-0.html

PS: nearly the same, but with a little smaller offset (1.001 vs 1.005).


2014-05-31 0:45 GMT+02:00 <bugzilla-daemon@freedesktop.org>:

>  m.a.riosv <mariosv@miguelangel.mobi> changed bug 79450
> <https://bugs.freedesktop.org/show_bug.cgi?id=79450>
>  What Removed Added  Status UNCONFIRMED NEEDINFO  CC
> mariosv@miguelangel.mobi  Ever confirmed   1
>
>  *Comment # 1 <https://bugs.freedesktop.org/show_bug.cgi?id=79450#c1> on
> bug 79450 <https://bugs.freedesktop.org/show_bug.cgi?id=79450> from
> m.a.riosv <mariosv@miguelangel.mobi> *
>
> Hi @ap, thanks for reporting.
>
> Please can you attach attach a screenshot to see what is bad for you and so
> verify if we see the same?.
>
>  ------------------------------
> You are receiving this mail because:
>
>    - You reported the bug.
>
>

Created attachment 100206 [details]
attachment-32654-1.dat

Created attachment 100207 [details]
Calc_centered_bubbles_Screen_copy_bis.jpg

Created attachment 100235 [details]
Sample file to test

Reproducible with:
Win7x64
Version: 4.1.6.2 Build ID: 40ff705089295be5be0aae9b15123f687c05b0a
Version: 4.4.0.0.alpha0+ Win-x86@39, Branch:master, Time: 2014-05-29_10:21:12

Seems as it's needs a minimal difference in the position to sum the color,

But when we assign a different colour to the bubbles, seems to works, please try with the last three bubbles.

Created attachment 100333 [details]
attachment-27371-0.html

Yes, I did obtain the same result as you, when the color of each of the
three last bubbles are "manually" changed, and even if the color chosen is
the default color of the series [see lines 9-11].

One detail: when two bubbles are centered *without having changed their
default color*, the color do not sum, but also the label of the second
bubble does not appear [see lines 1-2] and the bubble cannot be selected in
the chart (for example by using the TAB button from another bubble
selected). All is like the line of the second bubble in the data table
would not be red when building the chart (the second bubble does not seem
to exist in the chart). In fact, the transparency of the color does not
change this behaviour [see the second chart, which has been obtained by
copying the first one and removing transparency of the color of the series].


2014-06-01 15:04 GMT+02:00 <bugzilla-daemon@freedesktop.org>:

>   *Comment # 9 <https://bugs.freedesktop.org/show_bug.cgi?id=79450#c9> on
> bug 79450 <https://bugs.freedesktop.org/show_bug.cgi?id=79450> from
> m.a.riosv <mariosv@miguelangel.mobi> *
>
> Created attachment 100235 [details] <https://bugs.freedesktop.org/attachment.cgi?id=100235> [details] <https://bugs.freedesktop.org/attachment.cgi?id=100235&action=edit>
> Sample file to test
>
> Reproducible with:
> Win7x64
> Version: 4.1.6.2 Build ID: 40ff705089295be5be0aae9b15123f687c05b0a
> Version: 4.4.0.0.alpha0+ Win-x86@39, Branch:master, Time: 2014-05-29_10:21:12
>
> Seems as it's needs a minimal difference in the position to sum the color,
>
> But when we assign a different colour to the bubbles, seems to works, please
> try with the last three bubbles.
>
>  ------------------------------
> You are receiving this mail because:
>
>    - You reported the bug.
>
>

Created attachment 100334 [details]
attachment-27371-1.dat

Created attachment 100335 [details]
Calc_centered_bubbles_v3.ods

Confirmed your last comments.

But, please don't reproduce the emails in the answer, that make the thread more difficult to follow.

This is related to a performance improvement hack. We check if two objects are at the same place and in that case don't render the second. This is a bit more problematic for bubbles where the bubble size is different between the two points.

I'll think a bit if we can remove this feature for bubble charts.

I also am encountering this bug; it is blocking use of translucent bubbles to display values that might repeat, since this bug means some of the data just is not displayed (making the chart inaccurate).  I had hoped to use uniform translucent bubbles to visually represent the clumping effect of data, but instead the bug eliminates the visual effect.

In addition, failure to render overlaid datapoints is unexpected result for scatterplots with data labels.  One would expect that overlaid data points would have overlaid data labels, which would prompt closer examination of visual representation on chart.  Taking that out removes a visual cue that LO is removing data points, and so I think that also is a bug.