# mark bugs fixed in 4:7.4.7-1+deb12u3 correctly, didn't get done
# because they were already archived
unarchive 1059158
fixed 1059158 4:7.4.7-1+deb12u3
unarchive 1069835
fixed 1069835 4:7.4.7-1+deb12u3
www.access2base.com/access2base.html into libreoffice-common as doc
-- Rene Engelhard Sun, 17 Nov 2019 13:34:51 +0100
As a minor note - there's a duplicate b-d on "libcairo2-dev, libcairo2-dev" and
"libglib2.0-dev, libglib2.0-dev"
Yeah, that's a side-effect of that being auto-gen
rchs anyway which we definitely
+ need to keep anyway (similar as libuuilo.so).
+- recommend kio >> 5.103.0-1 in -kf5
+
+ -- Rene Engelhard Fri, 24 May 2024 21:06:45 +0200
+
libreoffice (4:7.4.7-1+deb12u2) bookworm-security; urgency=high
* debian/patches/add-notify-for-script-use.d
1070887
<https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1070887> did...)
Regards,
Rene
.
Cleaning those bugs up and making an own one for kio.
Regards,
Rene
that one is backported to kf5 we'd get the fix then if we
shipped libreoffice-kf6 and KF6/Plasma6 is what is used.
(According to IRC it isn't sure whether this will be backported to kf5.
Let's see)
Regards,
Rene
Hi,
Am 25.04.24 um 18:37 schrieb Andreas B. Mundt:
On Thu, Apr 25, 2024 at 05:43:29PM +0200, Rene Engelhard wrote:
Am 25.04.24 um 17:03 schrieb Andreas B. Mundt:
For now, we traced the issue back to libreoffice-kf5. If this package
is removed, neither the document disappears on closing
either. In fact it caused 32bit builds
to fail[1] and I don't know what more regressions this caused. I would
be wary of "just" backporting it in a point release...
Regards,
Rene
[1] by relying on internal glibc/kernel types, see
https://salsa.debian.org/libreoffice-team/l
backport to 24.2.2+, but given it's
sstill stuck behind time_t...)
Regards,
Rene
{mysql,firebird}.ucf: add
* debian/control.in: add Breaks: -sdbc-{mysql,firebird} (<< 4:24.2.2-2)
to libreoffice-common
-- Rene Engelhard Sat, 30 Mar 2024 09:30:30 +
even though -2 was never attempted on armhf due to that build-dep (-3
was built then)
s: 1 Errors: 0
4:24.2.2-1 build failed with an orcus not rebult for time_t and after
that it succeeded (-3 forced an appropriate build-dep)
4:24.2.0-1 from testing now fails in that way after the orcus bin-NMU
(0.19.2-3+b2) migrated.
Regards,
Rene
Source: libreoffice
Version: 4:24.2.0-1
Severity: serious
Tags: trixie ftbfs
Hi,
Am 30.03.24 um 12:56 schrieb Rene Engelhard:
Am 30.03.24 um 08:49 schrieb Rene Engelhard:
That would mean a bin-NMU of liborcus would work and then a rebuild
of libreoffice (gb, but I need a new upload anyway
ord processor, and spacing is
something very very basic for them and incredibly necessary.
And again: That is the definition of important at most, not serious.
Read the bug severities.
Regards,
Rene
argue that it's not visible enough but
that's by no means important.
Or what space do you mean?
What are you aiming at with your non-bugs?
Regards,
Rene
Hi,
Am 06.04.24 um 11:43 schrieb Rene Engelhard:
Am 06.04.24 um 11:31 schrieb Rene Engelhard:
Am 06.04.24 um 11:11 schrieb Rene Engelhard:
See https://people.debian.org/~rene/libreoffice/1068479-works.png
Just for avoidance of doubt since the screenshot is in German:
[...]
And the first
Hi again,
Am 06.04.24 um 11:11 schrieb Rene Engelhard:
Am 06.04.24 um 11:03 schrieb Rene Engelhard:
Am 06.04.24 um 00:34 schrieb José Luis González:
The setting for spacing between paragraphs is missing in the spacing
and indentation tab of the paragraph dialog.
?
It's definitely
Hi,
Am 06.04.24 um 11:31 schrieb Rene Engelhard:
Am 06.04.24 um 11:11 schrieb Rene Engelhard:
See https://people.debian.org/~rene/libreoffice/1068479-works.png
Just for avoidance of doubt since the screenshot is in German:
That is the second part "Absatz".
First is "above&quo
Hi,
Am 06.04.24 um 11:03 schrieb Rene Engelhard:
Am 06.04.24 um 00:34 schrieb José Luis González:
The setting for spacing between paragraphs is missing in the spacing
and indentation tab of the paragraph dialog.
?
It's definitely there. Format -> Paragraph has spacing "after/befor
Hi again,
Am 06.04.24 um 11:11 schrieb Rene Engelhard:
Am 06.04.24 um 11:03 schrieb Rene Engelhard:
Am 06.04.24 um 00:34 schrieb José Luis González:
The setting for spacing between paragraphs is missing in the spacing
and indentation tab of the paragraph dialog.
?
It's definitely
And it's unreproducible.
Regards,
Rene
but I don't buy this since as I said back then it
looked fine here.
And you didn't even reply to the mail. Ignoring the moreinfo tag.
Ii's also not the fine style to not CC the actual bug so the text gets
actually sent there.
Regards,
Rene
Hi,
Also see https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=977857
Regards,
Rene
reassign 1068023 libreoffice-gtk4
thanks
Hi,
Am 29.03.24 um 21:04 schrieb Rene Engelhard:
Package: libreoffice
if it happens only with gtk4, isn't it better suited at libreoffice-gtk4?
I use other gtk applications (e.g. evince, audacity, inkscape, firefox)
in this same environment
Package: libreoffice
Version: 4:24.2.0-1
Severity: normal
X-Debbugs-Cc: Daniel Kahn Gillmor
I have qtwayland5 5.15.10-2+b1 installed. I do not have XWayland
installed at all. I'm running from within a Wayland session, using sway
1.9-1 as a compositor.
When i try to launch libreoffice using
Hi again,
Am 10.03.24 um 19:59 schrieb Rene Engelhard:
BTW, What is the replacement for it? setuptools._distutils? As in the
following patch?
OK, so discussion on IRC gave:
19:51 < tumbleweed> _rene_: distutils is removed in 3.12
19:51 < tumbleweed> if you need distutils, tr
Hi,
Am 10.03.24 um 19:44 schrieb Rene Engelhard:
and similar stuff in upstreams configure.
python_include=`$PYTHON -c "import distutils.sysconfig;
print(distutils.sysconfig.get_config_var('INCLUDEPY'));"`
python_version=`$PYTHON -c "import distutils.sy
EFIX}/lib/python3.12/site-packages
checking for python3 extension module directory (pyexecdir)...
${PYTHON_EXEC_PREFIX}/lib/python3.12/site-packages
[...]
in upstreams configure.
I think this bugreport is too early.
Regards,
Rene
can re-enable it now and not have it
blocking builds.
But I need to do the Replaces: libreoffice-common in -evolution anyways now?
Regards,
Rene
forwarded 1065448 https://bugs.documentfoundation.org/show_bug.cgi?id=160033
thanks
Hi,
Am 04.03.24 um 21:58 schrieb Rene Engelhard:
Package: libreoffice-common
Version: 4:24.2.0-1
Severity: normal
Tags: upstream
Then you should have filed it upstream :). Didn't write the reportbug
text
Package: libreoffice-common
Version: 4:24.2.0-1
Severity: normal
Tags: upstream
X-Debbugs-Cc: pe...@pblackman.plus.com
Dear Maintainer,
When creating pdf files from odt files, soffice writes a CreationDate field
which contains the actual build date/time. This varies with every build.
For an
Source: libreoffice
Version: 4:24.2.0-1
Severity: serious
Control: close -1 4:24.2.0-3
Am 23.02.24 um 17:14 schrieb Rene Engelhard:
Hi,
Am 23.02.24 um 08:02 schrieb HIGUCHI Daisuke (VDR dai):
Sorry, resending to BTS, not to debian-openoffice.
No problem, that -1 redirects the bug
Hi,
Am 23.02.24 um 08:02 schrieb HIGUCHI Daisuke (VDR dai):
Sorry, resending to BTS, not to debian-openoffice.
No problem, that -1 redirects the bug reports to debian-openoffice is a bug.
(Fixed in later versions but those are stuck after the t64 transition.)
Regards,
Rene
ice-dictionaries/-/merge_requests/6>
Merged (and uploaded after fixing the changelog)
Regards,
Rene
Hi,
oops.
Am 21.01.24 um 15:35 schrieb Rene Engelhard:
Here the new libxml2 removes functions and symbol versions used by
gazillions of packages over the whole of the Debian archive.
And no, the exact point of Debian library package names is that they
HAVE to change on ABI changes
Hi,
Am 21.01.24 um 15:27 schrieb Eric Valette:
On 21/01/2024 14:49, Rene Engelhard wrote:
Exactly that is the point of #1059040. The binary packages have to be
renamed. (Then rebuild against libxml2-WHATEVERNEW). Then a rebuild
LO will have a proper dependency on libxml2-WHATEVERNEW.
I
with 2.12. And here libsdlo.so *does* link against libxml:
Missing dependency < dependency at least.
Yeah. But for that you need a palantir. For an unknown amount of
packages in the archive?
No. The bug is in libxml2.
Regards,
Rene
ped the ball on
it. Can resurrect it, though)
Regards,
Rene
libwoff1:amd64 1.0.2-2 amd64
library for converting fonts to WOFF 2.0
ii lmodern 2.005-1 all scalable
PostScript and OpenType fonts based on Computer Modern
ii tex-gyre 20180621-6 all scalable
PostScript and OpenType fonts based on URW Fonts
ii texlive-fonts-recommended 2022.20230122-3
all TeX Live: Recommended fonts
ii xfonts-100dpi 1:1.0.5 all 100
dpi fonts for X
ii xfonts-75dpi 1:1.0.5 all 75 dpi
fonts for X
ii xfonts-base 1:1.0.5+nmu1 all
standard fonts for X
ii xfonts-encodings 1:1.0.4-2.2 all
Encodings for X.Org fonts
ii xfonts-scalable 1:1.0.3-1.3 all
scalable fonts for X
ii xfonts-utils 1:7.7+6 amd64
Regards,
Rene
be closed.
Thanks for confirming. Let's keep it open and mark it as forwarded to
the bug James mentioned.
Regards,
Rene
libmergedlo.so in 64bit archs anyway
we can just include it here since as said i's included in 64bits
packages nevertheless.
Already in git (you got the mails).
No idea when this actually will end up in stable, though :)
Regards,
Rene
Hi,
Am 23.12.23 um 11:43 schrieb Rene Engelhard:
Hi,
Am 23.12.23 um 02:40 schrieb Bastian Germann:
graph_legend.dot should have quotes around the font name references.
Ah, thanks. Unfortunately this is a generated file...
And yes, I also noticed that the FreeSans.ttf is at fault. Indeed I
occurance archive-wise?
A workaround would be removing doxygen from Build-Depends
and the two doxgen output files from debian/libgraphite2-doc.docs
Thought abozt this, too. But that would loose the API documentation
Regards,
Rene
--disable-gui build they'd still be there, we explicitely
remove them, because otherwise we get all those libs and that makes the
difference size-wise GUI/nogui not really big..)
Regards,
Rene
an existing or new spreadsheet using
loadComponentFromURL
Sigh. All that stuff needing *ui stuff where they actually don't need ui
files because it's built with --disable-gui. That is the third one like
this.
See also http://bugs.debian.org/1058653
Regards,
Rene
Hi,
Am 14.12.23 um 07:57 schrieb Rene Engelhard:
Hi,
Am 14.12.23 um 06:01 schrieb tony mancill:
A FTBFS bug [1] cropped up in gpredict recently that appears to be to a
change in libreoffice-writer-nogui. The build step that fails is:
$ soffice --strace --writer --headless --convert-to pdf
ibreoffice-core (since you *can* mix them.) See also
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1052052
Regards,
Rene
retitle 1057434 libreoffice-numbertext: NUMBERTEXT() gives Err: 504
thanks
Am 05.12.23 um 18:56 schrieb Rene Engelhard:
When trying to use the =NUMBERTEXT(5) formula in Libreoffice Calc, you get an
'Err: 504' error in those cells.
Confirmed..
Cause is
$ debdiff libreoffice-numbertext_1.0.11
stable contains 1.0.11-1?
(Which does work, just tested).
Regards,
Rene
9~rc1-1
is the end of a branch.
Regards,
Rene
tag 1056104 + wonffix
thanks
Hi,
Am 17.11.23 um 18:30 schrieb Rene Engelhard:
So it would boil down to
rene@frodo:~/LibreOffice/git/master/debian$ git diff rules
diff --git a/rules b/rules
index e7152c60..ed7a8157 100755
--- a/rules
+++ b/rules
@@ -2247,6 +2247,7 @@ config_host.mk
f.
Which would need to be hacked in since it's only flagged by
--enable-wasm-strip which as the name says is for WebAssembly
(cross-)compile and thus can't be done generically (and strips out a
sh*load of stuff)...
So it would boil down to
rene@frodo:~/LibreOffice/git/master/debian$ git diff
Hi,
Am 16.11.23 um 21:14 schrieb Rene Engelhard:
-
Failed to install: No such file or directory at
/<>/solenv/bin/ooinstall line 102.
-
No, totally unrelated. (And I wonder how you got into this at all.)
And besides that, if it was that that it already built ever
f you are unable to reproduce this, please provide a build log and diff it
so that we can identify any relevant changes that may have occurred in the
meantime.
That routinely will be useless since my paths in the build log will
probably not match yours. And I am not doing sbuild which does
<>.
Regards,
Rene
is not)
Regards,
Rene
tag 1055048 + moreinfo
Thanks
Am 30. Oktober 2023 09:12:05 MEZ schrieb Nicolas Patrois
:
>Package: libreoffice-writer
>Version: 4:7.5.8~rc1-2
>Severity: normal
>Tags: upstream
>
>Dear Maintainer,
>
>I have problems when I try to create overlapping lines in tables.
(...)
Since 7.5.8 is
(or upstream) to answer.
Regards,
Rene
@@
include /usr/share/dpkg/architecture.mk
+export DEB_CXXFLAGS_MAINT_APPEND += -fexcess-precision=fast
+
%:
dh $@
Will do that one, thanks.
Regards,
Rene
eadable text and low contrast due to a gray background.
Well, yes, that's how LO looks without Gtk3 integration.
Regards,
Rene
it in task-mate-desktop... I prepared a
merge request for tasksel...
I don't know where to go to tweak this in the distro, but I suspect you do!
That would be tasksel.
Regards,
Rene
e: $(call
fetch_Download_item_unchecked,https://download.documentfoundation.org/libreoffice/src/$(shell
echo $(gb_LO_VER) | sed -e
"s/\([0-9]*\.[0-9]*\.[0-9]*\).*/\1/"),libreoffice-$(i)-$(gb_LO_VER).tar.xz))
Regards,
Rene
fixed 1045940 0.18.1-1
thanks
Am 14.08.23 um 20:09 schrieb Rene Engelhard:
Am 14.08.23 um 18:05 schrieb Rene Engelhard:
make[1]: Leaving directory '/<>'
dh_autoreconf_clean
dh_clean
rm: cannot remove './doc/_doxygen': Is a directory
rm: cannot remove './doc/tmp': Is a dir
-over but this is quite obviously not a
bug here.
Regards,
Rene
in KDE (X11 and wayland).
I see you have non-Debian packages installed. try with a clean Debian?
Regards,
Rene
he correct Replaces:
Regards,
Rene
-nogui (<< 4:7.6.0~rc1),
libreoffice-impress (<< 4:7.6.0~rc1),
libreoffice-impress-nogui (<< 4:7.6.0~rc1)
Leaves "just" sid.
Regards,
Rene
Hi,
Am 22.07.23 um 16:17 schrieb Andreas B. Mundt:
Hi Rene, all,
first, many thanks for maintaining libreoffice! We are running the
KDE plasma desktop environment in our school and observe probably the
same issue here:
Sure? The description in the original bug doesn't talk about mouse
just does it for stable and
security, maybe
"So there is simple solution:
apt-get purge libreoffice*
apt-get install libreoffice
And now it successfully starts. Sorry."
OK. :)
Regards,
Rene
Hi,
Am 20.07.23 um 18:42 schrieb Rene Engelhard:
To compare: You can't also break apache config or merge new config stuff
^ not
Regards,
Rene
s
because as we see stuff relies on keys present.
That's why this uses ucf. It has even a three-way merge option...
Regards,
Rene
Regards,
Rene
instead?
Yes.
That answer doesn't fit the question :)
Regards,
Rene
a bug
Rationale:
mipsel is completely broken, long story at [1].
Upstream who added it (RedHat) confirmed no interest to do anything on
it.
Regards,
Rene
[1] see
https://buildd.debian.org/status/package.php?p=libreoffice=experimental:
[build MOD] testtools
S=/<> && I=$S/instdir &
Hi,
Am 13.07.23 um 18:08 schrieb Florian Ernst:
there is a new upstream release available at
<https://dev-www.libreoffice.org/src/libexttextcat/>, adding support for
some more languages / variants.
Oops, indeed. November 2021.. Bummer...
Regards,
Rene
Hi,
Am 10.07.23 um 22:45 schrieb Rene Engelhard:
[build UIT] cui_tabpages
[...]
Tests run: 2
Tests failed: 0
Tests errors: 0
Tests skipped: 0
That said, before that a make uicheck in cui or a make uicheck toplevel
gets:
[build ECH] CustomTarget/instsetoo_native/setup/setuprc
/bin/sh: 1
Hi,
forgot one sentence.
Am 10.07.23 um 22:45 schrieb Rene Engelhard:
So I had the idea to try a export DEB_BUILD_OPTIONS="noopt" build
(resulting into -O0) and it looks better:
So I guess I'll workaround this by building with -O0. (riscv64 isn't
alone there, see
https://salsa.
tag 1039906 - moreinfo
thanks
Hi,
Am 29.06.23 um 17:29 schrieb Rene Engelhard:
It seems that the exception only throws in current Debian package. If
I build LibreOffice (both 7.6 and 7.5 version) with external tarballs
and upstream compiler flags, see below
(that means, I pass parameters
Hi,
Am 29.06.23 um 17:29 schrieb Rene Engelhard:
Though that does not really matter, see [1]. We are supposed not to use
Forgot to add that.
https://www.debian.org/doc/debian-policy/ch-source.html#embedded-code-copies
Regards,
Rene
would need to
build with that python, too. And there's stuff in Debian using pyUNO.
I am building a debug version of LibreOffice with sbuild to debug uitest.
Yes, please.
Or try with a newer python? Unfortunately LibreOffice is not yet ready
for python 3.12...
Regards,
Rene
Hi,
Am 18.06.23 um 13:49 schrieb Changwoo Ryu:
2023년 6월 18일 (일) 오후 8:00, Rene Engelhard 님이 작성:
Should I add a Breaks: hunspell-ko (<< 0.7.94) here? Strictly speaking
it's not needed since it's a test failing not the dict being unusable,
but we could go the "better safe than s
Hi,
Am 18.06.23 um 12:52 schrieb Changwoo Ryu:
2023년 6월 18일 (일) 오후 7:39, Rene Engelhard 님이 작성:
But I saw you uploaded hunspell-dict-ko 0.7.94 which incidentally has
someting hunspell 1.7.2-related (did I mention I don't like the stuff
being in Korean instead of English).
At least the tests
Hi,
Am 13.06.23 um 18:49 schrieb Rene Engelhard:
Am 12.06.23 um 19:32 schrieb Changwoo Ryu:
This failure is not from hunspell-dict-ko. hunspell just failed to
call iconv() in that environment.
Yeah, see the code point in question, just wondering how that happens.
Probably because
Hi again,
Am 14.06.23 um 14:47 schrieb Rene Engelhard:
S=/<> && I=$S/instdir && W=$S/workdir && mkdir -p
$W/GenCxxObject/UnpackedTarball/skia/src/gpu/ganesh/vk/
$W/Dep/GenCxxObject/UnpackedTarball/skia/src/gpu/ganesh/vk/ && cd
/<&g
unstable task-german-desktop et. al. unstable qemu VM.
(And current git unfortunately doesn't change the situation).
Guess I need to report a bug at hunspell then.
Regards,
Rene
Hi,
Am 12.06.23 um 18:58 schrieb Changwoo Ryu:
2023년 6월 11일 (일) 오후 7:34, Rene Engelhard 님이 작성:
This blocks (well, theoretically not, but then it'd block in unstable)
the upload of hunspell 1.7.2 which has many bugfixes and improvements
(and which the new version of r-cran-hunspell - due
spell/issues/903)
) but that one is fixed, before that the failures in autopkgtest was
even longer.
I'd file a bug at github but it seems everything is korean there (which
I dislike, open source/free software stuff has to be done in english
imho, and I am not a english native speaker either)
Regards,
Rene
- which you here obviously did not and just posted a
hilaroioud "patch" link to all regardless of their nature.
Regards,
Rene
ions(+), 4481 deletions(-)
The whole diff is 17M, so I am not attaching it here. It is at
https://people.debian.org/~rene/libreoffice/7.4/7.4.7.debdiff
I know this is completely bending the rules, but so is #1035056, too
If you wish I can make this also -0deb12u1, but as there will be no
7.4.7-1
t; which imho is bad.
Regards,
Rene
e bookworm upgrade...)
+ * debian/patches/CVE-2023-2255.diff:
+fix CVE-2023-2555 ("Remote documents loaded without prompt via IFrame")
+ * debian/changelog:
+- mention CVE-2022-38745 in 1:7.3.1-1s changelog
+- fix typo in last upload (s/choosen/chosen/), thanks lintian
+
+ --
Hi,
Am 20.05.23 um 23:52 schrieb Thorsten Glaser:
Rene Engelhard dixit:
You as DD should know that stable won't get any updates for this anymore. (And
it will be oldstable in a short time anyway, even.)
So, who cares?
me :)
So reporting this against stable does not make any sense at all
been abvious.
Admittedly, I shouldn't have just answered Adrian on his last reset *on
the day it was supposed to be removed* but just also here, by bad there.
Still...
I have removed the upload.
No, you haven't. I had already.
Regards,
Rene
freeze anyway with medium urgency
so it's a total waste of time.
Regards,
Rene
7.5 (which also is
in "important bugfixes" mode only now, though[1]) - or even 7.6.0
alpha1[1][2].
I think I saw commits for font subsetting passing by when I checked the
commit log for other things..)
(And I still think such bugs should be reported upstream directly)
Regards,
Hi,
Am 01.05.23 um 00:41 schrieb Rene Engelhard:
It *seems* to be this in your nogui-strace:
94919 21:08:29.439478 openat(AT_FDCWD,
"/usr/lib/libreoffice/share/config/soffice.cfg/modules/simpress/ui/tabviewbar.ui", O_RDONLY) = -1 ENOENT (No such file or directory)
94919 21:08
Hi,
Am 12.01.23 um 17:38 schrieb Rene Engelhard:
This one also works but only with the binary in the instdir-nogui, *not*
if I install the resulting .debs:
Ah, hmm, so that really points to a packaging problem or a problem which
only exhibitx due to packaging.
From looking at the initial
? Or, is there another more tenable solution for this case?
Yes, maybe it's just an oversight when -nogui was introduced.
I would at least think that if scripting some mailmerge having the
actual driver for the database (if it's postgres) might make sense...
Regards,
Rene
ncy=medium
+
+ * debian/patches/avoid-empty-java.class.path.diff: apply upstream patch
+avoiding empty -Djava.class.path= (CVE-2022-38745)
+
+ -- Rene Engelhard Sat, 25 Mar 2023 14:04:55 +0100
+
libreoffice (1:7.0.4-4+deb11u5) bullseye; urgency=medium
* debian/patches/hrk-euro-default.dif
into bookworm
given the freeze schedule.
/7.5.0 might have worked if one was really courageos but then we'd have
a serious problem on any bigger surprise. Usually .0s are not really
good anyway, .1 woul bd a given and that really misses the deadline
completely)
Regards,
Rene
Hi,
No, it does not.
In a clean cowbuilder testing chroot:
Also not inside sbuild (in a freshly created sbuild testing chroot).
Regards,
Rene
clone 1031578 -1
retitle -1 ru/hu fail to build with python 3.11: re.error: global flags
not at the start of the expression at position 28
block -1 by 1031578
tag -1 - unreproducible
tag -1 + help
thanks
Hi,
Am 19.02.23 um 08:49 schrieb Rene Engelhard:
The real issue though seems
1 - 100 of 12438 matches
Mail list logo