Corrupted Calc Display of Excel File

users
#1

When I open a certain Excel spreadsheet which shows calendar dates and
programs for cumulative mileage, the information comes out corrupted when I
open it on my new computer but ok on my old computer. I have put the file
on a thumb drive with the same results, i.e. bad on the new computer, ok on
the old computer. I have opened other Excel created spreadsheets on the new
computer without an apparent problem (these other spreadsheets do not have
formulas.

An example of the corruption: In the date column, 31 shows up in all cells
vertically where the number should reflect day of the month. In other cells
where mileage should be cumulated for a month reflecting other daily
entries, there is a zero.

Any idea of the problem and solution?

#2

Yes, the new computer is using LibreOffice 4.2 and the old computer is using
4.1. I've installed 4.1.4 on the new computer and it works. Any idea why this would occur?
Makes no sense that an immediately newer version would be so fundamentally
corrupted.

Ken

#3

Hi ksc

ksc wrote

I've installed 4.1.4 on the new computer and it works. Any idea why
this would occur?
Makes no sense that an immediately newer version would be so fundamentally
corrupted.

The main reason is that the Calc spreadsheet had a major code rewrite which
enabled several enhancements including GPU calculations (when activated and
when the GPU is compatible)

LibreOffice 4.2.0.4 (notice the zero after 4.2) is the first release in a
new branch (the 4.2 branch). For Windows users usually a 4.2 version is
always better than a 4.1...

However LibreOffice is a multi-platform software and since most developers
are on Linux it follows a Linux logic
(https://wiki.documentfoundation.org/ReleasePlan). Therefore an x.x.0
release (and in fact any release before x.x.4) is considered potentially
buggy and not ready for production.

Therefore even if LO 4.2 is announced everywhere on the LO site, in fact it
is not recommended for general use or in other words "is targeted for early
adopters and private power users" (as stated in the not so evident Release
Notes page http://www.libreoffice.org/download/release-notes/)

Bottomline: stick with the trusted 4.1 branch (4.1.5 should be out within
the next days) or keep 4.2 if you can live with the bugs (and report them at
https://bugs.freedesktop.org/enter_bug.cgi?product=LibreOffice ) so that
they are eventually fixed.

Hope this helps...

#4

Thanks. That's a good explanation. Doesn't make sense that the unitiated
(like me, though I've been using computers for years) are led to download
4.2 with a serious flaw.

#5

Hi :slight_smile:
I just wish the main downloads page could make that clear instead of
pushing all noobs into using the "feature rich" branch.
Regards from
Tom :slight_smile:

#6

Now I see that numbers entered for February 1 forward (I tried a few dates)
result in Err:502 in the lower cell in the cumulative column. And I'm using
4.1.4. Another bug?

#7

Hi ksc

ksc wrote

Now I see that numbers entered for February 1 forward (I tried a few
dates)
result in Err:502 in the lower cell in the cumulative column. And I'm
using
4.1.4. Another bug?

Maybe it is...
Can you provide a step-by-step description on how to replicate this problem
and/or upload a sample file somewhere?

In any case, some of the bugs are already solved in the Master branch and
should hopefully be included in the 4.2 branch.

#8

Hi all

Pedro wrote

ksc wrote

Now I see that numbers entered for February 1 forward (I tried a few
dates)
result in Err:502 in the lower cell in the cumulative column. And I'm
using
4.1.4. Another bug?

Maybe it is...
Can you provide a step-by-step description on how to replicate this
problem and/or upload a sample file somewhere?

In case someone is following this thread, I received the file by mail and
the error described was a consequence of the functions used if some cells
were not filled in. Therefore it wasn't a bug in the 4.1 branch.

The file did reveal some serious mess in 4.2 but hopefully these regressions
will be fixed in the near future (maybe some of them are already fixed...)

#9

In case someone is following this thread,

​We are ;)​

I received the file by mail and
the error described was a consequence of the functions used if some cells
were not filled in. Therefore it wasn't a bug in the 4.1 branch.

The file did reveal some serious mess in 4.2 but hopefully these
regressions
will be fixed in the near future (maybe some of them are already fixed...)

Is it possible to reproduce the issue with a minimal, non sensitive sample
file ? It might be useful to have an example of what did go wrong in 4.2,
in case there is more report about this later.​​

#10

Hi :slight_smile:
Is your new computer using LIbreOffice 4.2.0? If so then it might be
a good idea to install LibreOffice 4.1.4 instead to replace it. You
can fidn out the version number by going to the "Help" menu and then
click on "About LibreOffice"

Are you using Windows on both machines? Knowing that much would make
a bug-report much easier for the devs to work at.
Regards from
Tom :slight_smile: