Re: assist

[Michael Weghorn]

Hi Marisane, and welcome!

(More comments on specific aspects below.)


On 2024-06-01 19:27, Marisane Moruthanyana wrote:

I trust that I have come to the right list.

I have just installed LibreOffice after being advised to do so.

I am challenged to understand some of its features and the interface in 
general.


However, I have, to some extent, established how to navigate its interface.

I am a totally blind guy.


This mailing list is focused on topics related to LibreOffice 
development, i.e. meant primarily for LibreOffice (and extension) 
developers.


There is also another mailing list for discussing topics related to 
LibreOffice accessibility (accessibil...@global.libreoffice.org), where 
questions like yours about how to use LibreOffice with a screen reader 
are likely to reach more people who can provide helpful input. It is 
mentioned (and includes a link where you can subscribe) on this website:

https://www.libreoffice.org/get-help/mailing-lists

I've added that mailing list to the recipients now - and suggest to 
continue this discussion there instead of on the development mailing list.



My challenge is to know how to navigate the table.

1. I am struggling. JAWS or NVDA just says blank, blank, blank where it 
should be saying table with, for example, two rows and three columns.


What is going on?

Do I need to install screen reader scripts, for example JAWS's or NVDA's?

How do I do that?

Where do I get those?


The way that tables/table cells are presented by the screen reader is 
primarily up to the screen reader, so this might potentially need a 
change in the screen reader if it doesn't work as it should.
I'm less experienced with JAWS, but NVDA does provide a LibreOffice app 
module, which is however already part of the standard installation.


However, I just tested that with LibrOffice Writer 24.2.3 and NVDA 
2023.3.4. For me, NVDA announces e.g. "table with two rows, two columns" 
when keyboard focus moves into a table in a Writer document. So does the 
Orca screen reader on Linux. (I haven't explicitly checked JAWS.)


Note that the screen reader might allow configuring the behavior in its 
settings.


Did you also try that with Writer, or did you use Calc or another 
LibreOffice module? (I'm personally not sure whether it would 
necessarily be useful for a screen reader to announce the amount of rows 
and columns that a Calc spreadsheet has, as most of these will be unused.)




2. I am also struggling to add font face and size, let alone reading it.

I use JAWS+F or NVDA+F to read the font face and size, but I get nothing.

Each screen reader, depending on which is active, just echoes zero point.

What is going on?

How do I set or add and/or read the font face and size in LibreOffice?


This also works for me in a quick test with Writer with the 
above-mentioned LibreOffice and NVDA versions. What versions of NVDA and 
LibreOffice are you using?




My job is working with material production.

As a result, in my view, I should know how to navigate a table after 
creating it.


Tab and Shift+Tab can currently be used to navigate one cell 
forward/backward in tables.
Screen reader-specific table navigation currently don't work, will need 
work on both, LibreOffice and the screen reader to make it a reality.


Further, I should be able to add the font face and size in accordance 
with my visually impaired students in order to effectively serve their 
educational needs when enlarging their documents.


Changing the font and font size is possible either by defining that in a 
style and assigning that one (which is generally the preferred approach) 
or by manually setting the font and font size, e.g. in Writer's 
formatting toolbar, which can be reached by pressing F6 until the 
formatting toolbar gets focus. Once the toolbar has focus, you can move 
between its items using the Tab key, which should bring you to the "Font 
Name" and "Font Size" comboboxes that can be used to change those.


Are you running into any particular issues with these? Or is anything 
else missing?




I use LibreOffice after hours when I am at home as I do not have Ms 
Office on my home machine.


It is nearly the same as Word® and that makes my work easier.

I just need to understand how to deal with tables and fonts.

Others will come later as and when a need arises.

Please assist me.


Does the above help? Or is anything else still missing?

Regards,
Michael


OpenPGP_signature.asc
Description: OpenPGP digital signature

Re: "Could not find DWO CU" when collecting gdbtrace

[Stéphane Guillou]

On 31/5/24 16:17, Stéphane Guillou wrote:

Hi all

I have been getting a lot of "Could not find DWO CU" when trying to 
get backtraces with gdb 12.1 on Ubuntu 22.04, using the TDF *-dbg 
builds from https://dev-builds.libreoffice.org/daily/master/current.html


When gdb starts:

warning: Could not find DWO CU 
/home/tdf/lode/jenkins/workspace/lo_gerrit/tb/build_master/workdir/CObject/desktop/source/app/main.dwo(0x60838a0c9665b61c) 
referenced by CU at offset 0x0 [in module 
/home/stragu/lodev/lodbg/program/soffice.bin]


And a sample gdbtrace.log at: https://pastebin.com/GgjZqe49

Any idea? Will try with my own --enable-symbols build next.

No issue with a local --enable-symbols build. Trace has not warning 
about DWOs not found.


https://pastebin.com/FPL9gvnu

What's the difference with the daily TDF dbg builds?

--
Stéphane Guillou
Quality Assurance Analyst | The Document Foundation

Email: stephane.guil...@libreoffice.org
Matrix: @stragu:matrix.org
Fediverse: @str...@mastodon.indie.host
Web: https://stragu.gitlab.io/

[Bug 56756] FORMATTING: changes in gradients when formating a data series in Impress requires re-start of LiO

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=56756

--- Comment #14 from QA Administrators  ---
Dear bugquestcontri,

To make sure we're focusing on the bugs that affect our users today,
LibreOffice QA is asking bug reporters and confirmers to retest open, confirmed
bugs which have not been touched for over a year.

There have been thousands of bug fixes and commits since anyone checked on this
bug report. During that time, it's possible that the bug has been fixed, or the
details of the problem have changed. We'd really appreciate your help in
getting confirmation that the bug is still present.

If you have time, please do the following:

Test to see if the bug is still present with the latest version of LibreOffice
from https://www.libreoffice.org/download/

If the bug is present, please leave a comment that includes the information
from Help - About LibreOffice.

If the bug is NOT present, please set the bug's Status field to
RESOLVED-WORKSFORME and leave a comment that includes the information from Help
- About LibreOffice.

Please DO NOT

Update the version field
Reply via email (please reply directly on the bug tracker)
Set the bug's Status field to RESOLVED - FIXED (this status has a particular
meaning that is not 
appropriate in this case)


If you want to do more to help you can test to see if your issue is a
REGRESSION. To do so:
1. Download and install oldest version of LibreOffice (usually 3.3 unless your
bug pertains to a feature added after 3.3) from
https://downloadarchive.documentfoundation.org/libreoffice/old/

2. Test your bug
3. Leave a comment with your results.
4a. If the bug was present with 3.3 - set version to 'inherited from OOo';
4b. If the bug was not present in 3.3 - add 'regression' to keyword


Feel free to come ask questions or to say hello in our QA chat:
https://web.libera.chat/?settings=#libreoffice-qa

Thank you for helping us make LibreOffice even better for everyone!

Warm Regards,
QA Team

MassPing-UntouchedBug

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

Re: ESC meeting minutes: 2024-05-30

[Xisco Fauli]

Hello,

On 30/5/24 16:52, Miklos Vajna wrote:

+ RTL/CTL/CJK update (Jonathan)
    + tdf#81272: Slow CJK rendering due to font fallback
    + Per layout: O(nlogn) -> O(n)
    + tdf#161145: CJK characters have extra spacing since LO 7
    + Regression causing CJK characters to overflow cells on 
certain docx

    + Currently investigating to compare behavior against MSO
    + results from the vcl-level text layout changes?
  + will check (Xisco)
  + what doc corpus do you use? (Miklos)
  + a subset (Xisco)
  + not sure about the size off the top of the head 

    * docx: 2431 files
    * doc: 1006 files
    * rtf: 1006 files
    * ppt: 422 files
    * pptx: 1131 files

--
Xisco Faulí
LibreOffice QA Team
IRC: x1sc0

Re: ESC meeting agenda: 2024-05-23 16:00 CEST

[Samuel Mehrbrodt]

* Vertical tabpages state (Samuel)
  + review the state in 1 week


Here's the list of related bugs:

*tdf#161006, tdf#161020, tdf#161047, tdf#161049
*Fixed with https://gerrit.libreoffice.org/c/core/+/168081
*tdf#161026
*Unclear problem, NEEDINFO
*tdf#161030
*Fixed with https://gerrit.libreoffice.org/c/core/+/168251
So this should be in an acceptable state for 24.8 hopefully.
Regards Samuel

Re: ESC meeting agenda: 2024-05-30 16:00 CEST

[Julien Nabet]

Also, any idea when WASM TB will be upgraded?

Indeed, we need at least emscripten 2.0.32 to handle "as_handle()" method.

With 22ce8ed05be37d676739a578b05cc5217109fd87 "Emscripten: Unconditional
--enable-wasm-exceptions", we even need

emscripten 3.1.3 minimum

Julien
(Sorry, I had responded on the previous ESC).

Re: ESC meeting agenda: 2024-05-23 16:00 CEST

[Julien Nabet]

Also, any idea when WASM TB will be upgraded?

Indeed, we need at least emscripten 2.0.32 to handle "as_handle()" method.

With 22ce8ed05be37d676739a578b05cc5217109fd87 "Emscripten: Unconditional 
--enable-wasm-exceptions", we even need


emscripten 3.1.3 minimum

Julien

[Bug 86872] Word completion: Add ability to save word completion (aka "auto complete") lists and re-use them

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=86872

Shantanu  changed:

   What|Removed |Added

   See Also|https://bugs.documentfounda |
   |tion.org/show_bug.cgi?id=11 |
   |8309|

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

Re: Jan Iversen - license statement

[張雅菁]

Re: GSoC: Libreoffice Theme (week - 1 update)

[Sahil Gautam]

Hi,
https://gerrit.libreoffice.org/c/core/+/168016  updated patch trying to store 
the expert configuration
as a part of `StyleSettings`. Next step would be to discriminate between 
different settings based on
the scheme selected. Once that's done, I will start working on the widget 
toolkits to make the color
communication a "2-way-thing".

Also I am taking a 3-4 days break from "Libreoffice Themes" and will be working 
on other issues,
because by then I will have a fast computer. This would save me some time and 
frustration. Or
the pace I am working at now (credits to the 4 cores), it's like I change one 
line in settings.hxx
and it takes 2 hours or more to compile. Thanks to the comments, I remember 
what I changed :-).

Sahil Gautam
4 cores a day barely satisfies VCL

Re: GSoC: Libreoffice Theme (week - 1 update)

[Sahil Gautam]

Hi,
https://gerrit.libreoffice.org/c/core/+/168016  updated patch trying to store 
the expert configuration
as a part of `StyleSettings`. Next step would be to discriminate between 
different settings based on
the scheme selected. Once that's done, I will start working on the widget 
toolkits to make the color
communication a "2-way-thing".

Also I am taking a 3-4 days break from "Libreoffice Themes" and will be working 
on other issues,
because by then I will have a fast computer. This would save me some time and 
frustration. Or
the pace I am working at now (credits to the 4 cores), it's like I change one 
line in settings.hxx
and it takes 2 hours or more to compile. Thanks to the comments, I remember 
what I changed :-).

Sahil Gautam
4 cores a day barely satisfies VCL

Re: [GSoC] Comments in Sidebar Deck: Weekly Update

[Mohit Marathe]

Hi everyone,

This is my update of the past week:

- Created .ui files for "Thread" and "Comment", which will be loaded in the 
Comments panel.

- Added a method to add comments in the panel (I still need to link it so that 
it will be triggered when we want to add a comment).
- Researched on how to populate already existing comments in the panel

I couldn't get any work done in the first couple of days as my laptop was being 
repaired. I plan to be more productive this week.

Thanks,
Mohit

On Sunday, May 19th, 2024 at 10:57 PM, Mohit Marathe  
wrote:

> Hi everyone,
>
> I’ll be posting weekly updates about my progress on this project in this 
> thread.
>
> This week, I learned how to use Glade for making UIs.
> I used it to make .ui for “Comments” deck in the sidebar.
>
> This is how looks (for now, as there are lots of things to add):
>
> You can find the patch here: https://gerrit.libreoffice.org/c/core/+/167840
>
> Thanks,
> Mohit

Re: GSoC: Libreoffice Theme (week - 1 update)

[Michael Weghorn]

On 2024-05-27 06:48, Sahil Gautam wrote:

@Michael made a vaild point.

Haven't played around with the ColorConfig
much, (slow PC reasons), but this week I will go  through that.
Things will be more clear then.

This is my take on it now.
- Read the colors from the widget toolkits and store it somewhere.
   The entries in the `Color 
ColorConfig::GetDefaultColor(ColorConfigEntry eEntry)`
   are hard-coded, can be created on startup instead. Something like 
having a vector
   of structs with the default color values for different ENUMs 
associated to

   UI elements.

- Then the colors in the StyleSettings can be initialized based on theme 
selection.
   If it's "System Theme", then the colors saved in the point above can 
be used to
   initialize the StyleSettings. If it's some custom theme (extension), 
then the

   colors should be read from the expert configuration instead.


One more aspect to take into account is that the system theme can change 
while LO is running.


You can see that e.g. by starting LO with the gtk3 VCL plugin, then 
changing the GTK system theme:

The new theme is applied right away, not only after restarting LO.
(Similar with kf5 and changing the Qt style.)


OpenPGP_signature.asc
Description: OpenPGP digital signature

Re: GSoC: Libreoffice Theme (week - 1 update)

[Sahil Gautam]

@Michael made a vaild point.

Haven't played around with the ColorConfig
much, (slow PC reasons), but this week I will go  through that.
Things will be more clear then.

This is my take on it now.
- Read the colors from the widget toolkits and store it somewhere.
  The entries in the `Color 
ColorConfig::GetDefaultColor(ColorConfigEntry eEntry)`
  are hard-coded, can be created on startup instead. Something like 
having a vector
  of structs with the default color values for different ENUMs 
associated to

  UI elements.

- Then the colors in the StyleSettings can be initialized based on theme 
selection.
  If it's "System Theme", then the colors saved in the point above can 
be used to
  initialize the StyleSettings. If it's some custom theme (extension), 
then the

  colors should be read from the expert configuration instead.

On 5/27/24 9:15 AM, Michael Weghorn wrote:

On 2024-05-27 05:04, Sahil Gautam wrote:
The project is about not depending on the OS for colors and implement 
"Themes"
in Libreoffice. So as per my understanding (opinions invited), all 
the `Set...Color()`
calls in the `bool UpdateSettings(AllSettings& rSettings)` functions, 
which take
colors from the OS and put them into the `StyleSettings` are to be 
removed, and colors
should be passed to the widget toolkits, which were read from the 
expert configuration.


My take would be that colors should still be taken from the 
OS's/toolkit's current system theme/style by default, and only be 
overriden if explicitly configured so.

Re: GSoC: Libreoffice Theme (week - 1 update)

[Michael Weghorn]

On 2024-05-27 05:04, Sahil Gautam wrote:

The project is about not depending on the OS for colors and implement "Themes"
in Libreoffice. So as per my understanding (opinions invited), all the 
`Set...Color()`
calls in the `bool UpdateSettings(AllSettings& rSettings)` functions, which take
colors from the OS and put them into the `StyleSettings` are to be removed, and 
colors
should be passed to the widget toolkits, which were read from the expert 
configuration.


My take would be that colors should still be taken from the 
OS's/toolkit's current system theme/style by default, and only be 
overriden if explicitly configured so.


OpenPGP_signature.asc
Description: OpenPGP digital signature

Re: GSoC: Libreoffice Theme (week - 1 update)

[Sahil Gautam]

Hi, the coding period starts today.
The project is about not depending on the OS for colors and implement "Themes"
in Libreoffice. So as per my understanding (opinions invited), all the 
`Set...Color()`
calls in the `bool UpdateSettings(AllSettings& rSettings)` functions, which take
colors from the OS and put them into the `StyleSettings` are to be removed, and 
colors
should be passed to the widget toolkits, which were read from the expert 
configuration.

In GTK, it's done by adding a provider to the screen, and in QT/kf5, it's done 
by creating
a subclass of QStyle (QProxyStyle). For Windows and OSx, I don't know it yet.

So this would include defining UI elements from the theming perspective like 
(menubar),
(rest of the application - the notebook bar) etc.

First I will focus on getting the colors right, so I will add more entries like 
menubar etc
to the tools > options > Application Colors dialog, and once colors are done 
right for all
the platforms, creating a dedicated dialog for adding gradient support or 
background image
on the notebookbar etc shouldn't be that difficult.

I will be setting the milestones by the end of this week, as sometimes I feel 
that it's
not 350h of coding, more like 220h of coding and 130h of compiling ;). So in 
this week
probably, I will get a reality check of the scale (as my summer holidays have 
started :D).

Sahil Gautam

Re: Make failed due to multiple definitions in zxing

[Laurent Balland]

Without  --enable-pch=full option, it works fine. Thanks again :-)

Le 26/05/2024 à 09:54, Laurent Balland a écrit :


Thanks Caolán for your answer.

I build with --enable-pch=full since a long time. I do not remember 
when I activated it, but I did not modify my autogen.input since last 
November.


I'm going to try to build without this option.

Best regards

Laurent

Le 25/05/2024 à 22:38, Caolán McNamara a écrit :

On Sat, 2024-05-25 at 16:12 +0200, Laurent Balland wrote:

After "./g pull -r" and "make distclean && ./autogen.sh && make", I
get the following errors (see attached file).

Did you always build with
--enable-pch=full
before without error? I happen to not typically use it on Linux so I
wonder if that's the difference.

Re: Make failed due to multiple definitions in zxing

[Ilmari Lauhakangas]

On 25.5.2024 23.38, Caolán McNamara wrote:

On Sat, 2024-05-25 at 16:12 +0200, Laurent Balland wrote:

After "./g pull -r" and "make distclean && ./autogen.sh && make", I
get the following errors (see attached file).


Did you always build with
--enable-pch=full
before without error? I happen to not typically use it on Linux so I
wonder if that's the difference.


Related comment in an issue last year: 
https://github.com/zxing-cpp/zxing-cpp/issues/685#issuecomment-1868610667


'Only now did I realize that LibreOffice does not use the cmake build 
files provided by zxing-cpp but rolls it's own (ughh) and that was the 
"attack vector" for the precompiled headers change.'


It's weird that there would still be problems as upgrading to ZXing 
2.2.1 should have fixed the issue.


Ilmari

Re: Make failed due to multiple definitions in zxing

[Laurent Balland]

Thanks Caolán for your answer.

I build with --enable-pch=full since a long time. I do not remember when 
I activated it, but I did not modify my autogen.input since last November.


I'm going to try to build without this option.

Best regards

Laurent

Le 25/05/2024 à 22:38, Caolán McNamara a écrit :

On Sat, 2024-05-25 at 16:12 +0200, Laurent Balland wrote:

After "./g pull -r" and "make distclean && ./autogen.sh && make", I
get the following errors (see attached file).

Did you always build with
--enable-pch=full
before without error? I happen to not typically use it on Linux so I
wonder if that's the difference.

Re: Make failed due to multiple definitions in zxing

[Caolán McNamara]

On Sat, 2024-05-25 at 16:12 +0200, Laurent Balland wrote:
> After "./g pull -r" and "make distclean && ./autogen.sh && make", I
> get the following errors (see attached file).

Did you always build with
--enable-pch=full
before without error? I happen to not typically use it on Linux so I
wonder if that's the difference.

Re: GSoC: Libreoffice Theme (week - 1 update)

[Sahil Gautam]

Hi,the WIP prototype patch is linked in the sources below.
I created an extension and found that we can overwrite the expert configuration 
from
the extension. Then I searched on the web for what options do these toolkits 
provide in
terms of styling. For gtk, the only option that I found was to go via the CSS 
route, meaning
that one has to create either a CSS file, or a CSS string, load it into a 
GtkCssProvider(StyleProvider),
and add the provider to the GdkScreen. For QT/kf5 I am supposed to create a 
derived class of QStyle
and change the drawing behaviour for the widgets there. The youtube tutorial 
explains it well, but I
am yet to test it properly (my lsp was having some issues).

I added an entry under tools > options > application colors for changing the 
menubar color, and it works
in gtk. But it works only when one goes to tools > options > applic... and 
changes it on the spot.
On startup, the colors are washed away for some reason.

The various vclplugins communicate with the configuration via `AllSettings` 
which are passed to
the various `updateSettings(...)` functions. I tried using  `Color aColor = 
svtools::ColorConfig::GetDefaultColor(svtools::MENUBARBACKGROUND);`
but got linking errors. Meaning that the only way  of exchaning information 
with the toolkits is `AllSettings.StyleSettings().getcolor()`,
Please share if you know of any other ways to access the expert config colors 
in VCLs. I have been searching for the central AllSettings/
StyleSettings object, but didn't find it yet.

The `ColorConfig` class seems to be the central structure for accessing/modifying the 
tools > options > App Colors, but I found
not a single instance of it in the VCL. Also there are some questions that I am 
 trying to find the answer to first, like

- how is the central StyleSettings object (part of AllSettings) is 
communicating with expert configuration.
- why is my  lsp suddenly broken in qt source files :)


- A little frustrated --

Woring in VCL and colorcfg with a 4 core i5 laptop is a bad idea. And this was 
the major setback that I faced.
I couldn't just change something like add a new function and test it quickly. 
Compiling it took an hour or more,
and by then my brain's cache was cleared. So within this week, I am building a 
computer with 7950x.
I have ccache --max-size= 200G, still 4 cores!!

I had some college exams this month + the slow PC. I tried working on something 
else while the program compiles,
but that doesn't help as it takes 30 minutes to sink into the codebase and the 
constant context switching makes
it worst. Summer holidays for this year are starting day after tomorrow, and I 
hope the next week I can get more work done.

Sahil Gautam
still debugging with printf

Sources:
https://gerrit.libreoffice.org/c/core/+/168016  [PROTOTYPE PATCH]
https://wiki.documentfoundation.org/images/b/b0/LibreOffice_config_extension_writing.pdf
https://www.youtube.com/watch?v=Qn4HHbDxG0M  [A nice tutorial on Styling in 
QT/kf5]
https://discourse.gnome.org/t/setting-css-attributes-on-individual-widget-instances/13836
  [Styling in GTK]

On 5/14/24 2:03 PM, Printf Debugging wrote:
Hi, I am Printf from India, and I will be working on the "Libreoffice 
Theme" project this summer. First of all, many many thanks to Heiko 
and Rafael for mentoring the project, and to the community for being 
so welcoming! This week, I have been wrapping up some patches that I 
had assigned myself, along with some hunting and exploration in the 
VCL. I found that there are functions like `bool 
GtkSalGraphics::updateSettings(AllSettings& rSettings)` which extract 
the colors from the widget systems and save those in a local copy via 
the `Set.Color()` functions. The transaction seemed to be one way 
though, meaning that I didn't see any calls putting the updated colors 
back into the widget system, to modify the UI. For the colors like 
accent-color etc, I think all the getting/setting happens on the local 
copy after extraction. So I have been trying to understand how GTK 
does colors. Found that styling in GTK is built around 
`GtkStyleContext` to which we add StyleProviders, which have the 
styling information. There are a lot of GtkStyleContext pointers in 
`gtkgdi.hxx` for each major component. These are created at startup... 
(survey in progress) I will come up with a prototype patch to change 
the menubar colors probably in this week (counting the start of week 1 
from today) or at max next week. The idea is to get a thumbs up from 
the mentors, community and the devs on the approach/roadmap for the 
project, like shall we create a separate dialog for themes, or shall 
we repurpose the `Tools > Options > Application Colors/Personas` area. 
If we go the dialog route, then what should be the layout, considering 
that this project might extend beyond just plain colors during or even 
after the GSoC project. Is my styling approach scalable, meaning 
should I create a separate class etc etc. This is 

Re: GSoC: Adding native support for histogram chart

[Devansh Varshney]

Hi Everyone,

For this week, the columns/bars are now visible over the UI. I still have
to make proper changes to the model and controller. Now I need to make
changes to modify these changes eventually towards getting the Histogram
plot.
Here's the link to the PR: https://gerrit.libreoffice.org/c/core/+/167068/


[image: Screenshot from 2024-05-23 19-49-06.png]
[image: image.png]

Regards,
Devansh

On Wed, 8 May 2024 at 21:24, Devansh Varshney 
wrote:

> Hi everyone,
>
> For this week, my top priority is to make the new Histogram chart type
> selectable from the UI.
>
> Initially, it will be a working copy of the existing basic bar/column
> chart type.
>
> Then, I'll make the required changes with each passing iteration.
>
> Here's the link to the PR: https://gerrit.libreoffice.org/c/core/+/167068/
>
>
> Best regards,
>
> Devansh
>
> On Tue, 7 May 2024 at 18:05, Devansh Varshney <
> varshney.devansh...@gmail.com> wrote:
>
>> Hi everyone,
>>
>> I am Devansh and this year I have been selected for the* Google Summer
>> of Code (GSoC) program with LibreOffice* to *Add the support for the
>> Histogram Chart* - GSoC Idea list (
>> https://wiki.documentfoundation.org/Development/GSoC/Ideas#Histogram_.28and_other_missing.29_Chart_Type
>> ) and *Tomaž Vajngerl* is going to be* my mentor,* who has guided me
>> throughout the proposal process and will continue to do so until the
>> project's completion.
>>
>> I gravitated towards adding the support for the Histogram chart (and if
>> time permits other missing charts also from the MS Office 2016) stems from
>> my family background in science, engineering, and finance, where data
>> visualization plays a crucial role. Working on this project allows me to
>> connect with my family's work and explain my contributions in a meaningful
>> way.
>>
>> Second, while preparing the proposal report I came across that the *CERN*
>> also uses the LibreOffice for their work is doing a work around because of
>> no native support of these new chart types. (
>> https://indico.cern.ch/event/977476/contributions/4204938/attachments/2191304/3703670/EN-Plotting-a-histogram-in-libreoffice.pdf)
>> which was the point where I can see the impact of this task. Not only this
>> *IndiaMart*, a prominent B2B marketplace where my friends work told me
>> that it also uses LibreOffice.
>>
>> Seeing the potential impact of this project on real-world users was a
>> significant motivator for me. This project checks all the right boxes for
>> me.
>>
>> I have been impressed by the LibreOffice community and documentation,
>> which are among the best I've seen.
>> I'd like to extend my gratitude to the community members who have helped
>> me get started with my contributions, including Hossein, Heiko, Michael,
>> Marco, and especially Ilmari, who was the first to guide me.
>>
>> Regards,
>> Devansh 
>> (IRC: devansh)
>>
>

RE: VS 2022 version 17.9.1 has a bug preventing building cli_ure

[Vasily Melenchuk]

New version of Visual Studio 17.10.0 has exactly the same issue again. 
Do not hurry with update.

LINK : error LNK2034: metadata inconsistent with COFF symbol table: 
symbol '?   rtl_ustr_asciil_reverseEquals_WithLength@@$$J0YAEPEB_SPEBDJ@Z' 
(06000589) has inconsistent metadata with(0A000A14) in climaker_emit.o
climaker_emit.o : error LNK2020: unresolved token (0A000A14) "extern 
"C" unsigned char __cdecl  rtl_ustr_asciil_reverseEquals_WithLength(char16_t 
const *,char const *,long)" (?
rtl_ustr_asciil_reverseEquals_WithLength@@$$J0YAEPEB_SPEBDJ@Z)
D:\libreoffice\master\instdir\sdk\bin\climaker.exe : fatal error 
LNK1120: 1 unresolved externals



-Original Message-
From: LibreOffice  On Behalf Of 
Oliver Brinzing
Sent: Thursday, March 21, 2024 10:58 PM
To: libreoffice@lists.freedesktop.org
Subject: Re: VS 2022 version 17.9.1 has a bug preventing building cli_ure

Am 20.03.2024 um 06:01 schrieb Mike Kaganski:
> On 24.02.2024 21:50, Mike Kaganski wrote:
>> After upgrading my VS v.17.9.0 to v.17.9.1, building cli_ure started to fail 
>> like this:
>>
>>> LINK : error LNK2034: metadata inconsistent with COFF symbol table: 
>>> symbol 
>>> '?rtl_ustr_asciil_reverseEquals_WithLength@@$$J0YAEPEB_SPEBDJ@Z' 
>>> (06000579) has inconsistent metadata with (0A000A1A) in 
>>> climaker_emit.o climaker_emit.o : error LNK2020: unresolved token 
>>> (0A000A1A) "extern "C" unsigned char __cdecl 
>>> rtl_ustr_asciil_reverseEquals_WithLength(char16_t const *,char const 
>>> *,long)"
>>> (?rtl_ustr_asciil_reverseEquals_WithLength@@$$J0YAEPEB_SPEBDJ@Z)
>>> C:\lo\build\instdir\sdk\bin\climaker.exe : fatal error LNK1120: 1 
>>> unresolved externals
>>
>> I have filed a bug to Microsoft:
>>
>> https://deu01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fdev
>> elopercommunity.visualstudio.com%2Ft%2FLINK-%3A-error-LNK2034%3A-meta
>> data-inconsis%2F10599604%3Fport%3D1025%26fsid%3Dfdff4046-995a-4c06-8c
>> 64-1c14b9752c8f=05%7C02%7Cvasily.melenchuk%40cib.de%7C18f799eaae
>> bb4c97b38108dc49e142af%7C00dd3e38f5434e1a88b93dbb13d75c49%7C0%7C0%7C6
>> 38466479173686144%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjo
>> iV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C4%7C%7C%7C=5yAjhj
>> 9ym%2FhxgKpBUeYmnNftJ2md%2F8D82udjjOOKr%2BY%3D=0
>
> This problem is fixed in VS 2022 version 17.9.4:
> https://deu01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flear
> n.microsoft.com%2Fen-us%2Fvisualstudio%2Freleases%2F2022%2Frelease-not
> es%2317.9.4=05%7C02%7Cvasily.melenchuk%40cib.de%7C18f799eaaebb4c9
> 7b38108dc49e142af%7C00dd3e38f5434e1a88b93dbb13d75c49%7C0%7C0%7C6384664
> 79173695033%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMz
> IiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C4%7C%7C%7C=%2FvnkJSIPmEMe
> S4bWkRU%2B3sSDSJ0BzdEZ2ERpL8Lh%2F6Y%3D=0
>

yes, it works again with VS 2022

btw: i had a build break after a "clean build":

cd /cygdrive/d/sources/libo-core
./g checkout master
./g fetch --tags
./g pull -r

/opt/lo/bin/make clean gb_COLOR=1 gb_TITLES=1

/cygdrive/d/sources/libo-core/autogen.sh \ 
--with-external-tar=/cygdrive/d/sources/lo-externalsrc \ 
--with-junit=/cygdrive/d/sources/junit-4.10.jar \
--with-ant-home=/cygdrive/d/sources/apache-ant-1.9.5 \
--with-jdk-home=/cygdrive/d/Programme/Java/jdk-17 \ --enable-pch \ 
--disable-ccache \ --enable-debug \ --disable-odk \ --disable-online-update \
--with-visual-studio=2022

/opt/lo/bin/make build gb_COLOR=1 gb_TITLES=1 2>&1 | tee build.log

[...]
[PAT] icu
[BIN] dragonbox
[MOD] frozen
[BIN] glm
Configuring OpenSSL version 3.0.13 for target VC-WIN64A Using os-specific seed 
configuration
/cygdrive/d/sources/libo-core/workdir/UnpackedTarball/openssl/crypto/sha/../providers:
 No such file or directory at Configure line 3419.
make[1]: *** 
[D:/sources/libo-core/external/openssl/ExternalProject_openssl.mk:67:
D:/sources/libo-core/workdir/ExternalProject/openssl/build] Error 1
make[1]: *** Waiting for unfinished jobs

but folder "providers" is located at:

/cygdrive/d/sources/libo-core/workdir/UnpackedTarball/openssl/providers

workaround: copy folder "providers" to:

/cygdrive/d/sources/libo-core/workdir/UnpackedTarball/openssl/crypto

Re: where are published maven artifacts, after 7.6.4 ?

[sophi]

Hi,
Just for the list, I sent a link to the repository to Antoine who asked 
also on the fr lists.

Cheers
Sophie
Le 21/05/2024 à 14:15, antoine.pourc...@kelis.fr a écrit :

Hello,

Does anyone know why maven artifacts after 7.6.4 are no longer distributed
on  
https://mvnrepository.com/artifact/org.libreoffice/libreoffice?

Sincerely,

Antoine

scenario.software





--
Sophie Gautier so...@libreoffice.org
GSM: +33683901545
IRC: soph
Foundation coordinator
The Document Foundation

Re: Compiling nearly every file now compiles sqlbison.cxx and/or sqlflex.cxx

[Patrick Luby]

> On May 20, 2024, at 1:52 PM, Noel Grandin  wrote:
> 
> 
> So I saw the same thing on my ARM macbook, and I chased it for a while, 
> resetting/clearing ccache, rebooting, etc, but it just kept coming back, so I 
> gave up and updated to a newer version of ccache for my mac, which seems to 
> have fixed it (and as a bonus, is much newer and supports nice compression, 
> saving me disk space, etc)
> 
> Patch to update LODE here:
>https://gerrit.libreoffice.org/c/lode/+/167874
> 
> you will need to re-run LODE setup after applying that patch.
> 

Thanks Noel! Your LODE patch stopped the warnings in two clean builds (but no 
ccache clearing) for me.

Earlier today, like you I had found that if you interrupt the build after some 
point and run ccache -C, the warnings stop. But the problem would reappear 
after a make distclean or the next time some unknown code is recompiled (maybe 
in connectivity?).

Patrick

Re: Compiling nearly every file now compiles sqlbison.cxx and/or sqlflex.cxx

[Noel Grandin]

So I saw the same thing on my ARM macbook, and I chased it for a while,
resetting/clearing ccache, rebooting, etc, but it just kept coming back, so
I gave up and updated to a newer version of ccache for my mac, which seems
to have fixed it (and as a bonus, is much newer and supports nice
compression, saving me disk space, etc)

Patch to update LODE here:
   https://gerrit.libreoffice.org/c/lode/+/167874

you will need to re-run LODE setup after applying that patch.

Re: Compiling nearly every file now compiles sqlbison.cxx and/or sqlflex.cxx

[Caolán McNamara]

On Sun, 2024-05-19 at 20:26 -0400, Patrick Luby wrote:
> Any ideas? I’d like to just wrap a push/pop —implicit-fallthrough
> around whatever code compiles these sql* files.

Are you using ccache? Does blowing away that cache with ccache -C 
resolve it?

otherwise touch some file in vcl and what's the output of
make verbose=TRUE vcl.build?

Re: Build fail in mariadb on fedora 40

[Heiko Tietze]

On 18.05.24 2:51 PM, Dr. David Alan Gilbert wrote:

It looks like others started hitting it; see:
   https://gerrit.libreoffice.org/c/core/+/167806
Run into the same issue on a VM with Arch Linux. Latest master compiles now 
flawless.


OpenPGP_signature.asc
Description: OpenPGP digital signature

[Bug 135066] Frozen UI or Crash when interacting with embedded OLE object's Tabbed UI after re-entering edit mode if containing document uses classic toolbar UI

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=135066

Stéphane Guillou (stragu)  changed:

   What|Removed |Added

   See Also|https://bugs.documentfounda |
   |tion.org/show_bug.cgi?id=13 |
   |5213|

--- Comment #8 from Stéphane Guillou (stragu) 
 ---
*** Bug 135213 has been marked as a duplicate of this bug. ***

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

[Bug 135066] Frozen UI or Crash when interacting with embedded OLE object's Tabbed UI after re-entering edit mode if containing document uses classic toolbar UI

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=135066

Stéphane Guillou (stragu)  changed:

   What|Removed |Added

  Component|Impress |UI
 Whiteboard||target:7.2.0
   Severity|normal  |critical
 Status|NEW |RESOLVED
 CC||stephane.guillou@libreoffic
   ||e.org,
   ||szymon.k...@collabora.com
Crash report or||["libgobject-2.0.so.0.7200.
crash signature||4","SfxDispatcher::GetFrame
   ||","libc.so.6"]
Summary|Freeze UI and Crash: Insert |Frozen UI or Crash when
   |secondly table from Tab on  |interacting with embedded
   |writer document that is |OLE object's Tabbed UI
   |inserted impress|after re-entering edit mode
   ||if containing document uses
   ||classic toolbar UI
 Resolution|--- |FIXED
 Blocks||107237, 108589

--- Comment #7 from Stéphane Guillou (stragu) 
 ---
Tested in 7.0.0.3:

- all components with classic toolbar interface: not reproduced
- all components with tabbed UI: not reproduced
- only Impress with tabbed UI: not reproduced
- only Writer with tabbed UI: reproduced, with crash signature
"libgobject-2.0.so.0.7200.4" when killed after hang ->
https://crashreport.libreoffice.org/stats/crash_details/30b3ea38-d22f-4823-b86e-1804f3010db3
- the other way around (only Impress with tabbed UI, inserting Presentation OLE
object in Writer document, inserting two tables successively), I reproduce the
same thing.

Note that:
- you don't even need to insert a first table: even without editing the object
in the first instance, the hang happens after re-entering edit mode.
- also happens between other components (e.g. Calc + Writer), so not
Impress-specific
- also happens with other kinds of interaction with the tabbed UI. For example,
when inserting an image in a Presentation OLE object when Calc uses classic
toolbars, I get a crash signature "SfxDispatcher::GetFrame":
https://crashreport.libreoffice.org/stats/crash_details/5d8a8edb-f4a0-4455-a97c-63f679c0f555

In 7.1, I get direct crash with signature "libc.so.6":
https://crashreport.libreoffice.org/stats/crash_details/de4ad25d-56d1-402c-8f24-7cbaa697f9d5

Not reproduced in 7.2.0.4 nor in a recent daily build:

Version: 24.8.0.0.alpha1+ (X86_64) / LibreOffice Community
Build ID: 1b45ca1aa7d7cb8e7adcc07f8c60e26a413eca8c
CPU threads: 8; OS: Linux 6.5; UI render: default; VCL: gtk3
Locale: en-AU (en_AU.UTF-8); UI: en-US
Calc: CL threaded

Fix bibisected with linux-64-7.2 to first _good_ build
[7052746687af637a721d0d16c814a9769f887993] which points to:

commit  9f06c54ca1819b6d4df48f33772cfa5ceb90dccd
author  Szymon Kłos Tue Apr 06 07:08:30 2021 +0200
committer   Szymon Kłos Mon Apr 19 08:37:31 2021 +0200
notebookbar: minify for online
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/114154

Thanks Szymon!


Referenced Bugs:

https://bugs.documentfoundation.org/show_bug.cgi?id=107237
[Bug 107237] [META] Notebookbar Tabbed
https://bugs.documentfoundation.org/show_bug.cgi?id=108589
[Bug 108589] [META] OLE/Embedded object edit transition mode issues
-- 
You are receiving this mail because:
You are the assignee for the bug.

Re: Compiling nearly every file now compiles sqlbison.cxx and/or sqlflex.cxx

[Patrick Luby]

> On May 19, 2024, at 4:37 PM, Caolán McNamara  
> wrote:
> 
> On Sun, 2024-05-19 at 16:30 -0400, Patrick Luby wrote:
>> cd vcl && touch skia/gdiimpl.cxx && make
> 
> maybe adding -d to the end of the make might give a hint as to what
> could be going on

I did a make -d and no sql* files were listed as dependencies. Then I edited 
config_host.mk and added -E to clang++ everywhere, did "cd vcl && touch 
skia/gdiimpl.cxx && make”, looked at the .o file, and no sql* includes.

So, I just did a git bisect and found that commit 
0cf755adbb39591e9f6d273bfe5dce2469a86dbf causes this behaviour. Pretty wild 
that undefining _FORTIFY_SOURCE causes all these warnings.

Any ideas? I’d like to just wrap a push/pop —implicit-fallthrough around 
whatever code compiles these sql* files.

Patrick

Re: Compiling nearly every file now compiles sqlbison.cxx and/or sqlflex.cxx

[Caolán McNamara]

On Sun, 2024-05-19 at 16:30 -0400, Patrick Luby wrote:
> cd vcl && touch skia/gdiimpl.cxx && make

maybe adding -d to the end of the make might give a hint as to what
could be going on

Re: Compiling nearly every file now compiles sqlbison.cxx and/or sqlflex.cxx

[Patrick Luby]

> On May 19, 2024, at 3:33 PM, Caolán McNamara  
> wrote:
> 
> On Sun, 2024-05-19 at 13:22 -0400, Patrick Luby wrote:
>> A couple of weeks ago I started seeing my builds slow down and my
>> build log filled with compiler warnings in sqlbison.cxx and
>> sqlflex.cxx on macOS.
>> 
>> Is anyone else seeing this? I see these warnings even when compiling
>> files in vcl. Seems to me that something in the build is not working
>> right since I see one of these files is recompiled for each and every
>> other C++ file that is compiled.
> 
> That seems very odd. Checking a gerrit macos build log like
> https://ci.libreoffice.org/job/gerrit_mac/167726/consoleFull
> I just see the expected handful of sqlbison and sqlflex warnings as
> those individual files are compiled once.
> 
> I can't see any reason why compiling vcl would have any connection to
> something in connectivity.
> 
> What is your configure line? wondering if there is some
> --enable-mergelibs or --enable-pch thing that could be a trigger.

Maybe. I have neither of those enabled in my local build. But after a full 
build, I can get this behaviour to occur by the following steps:

cd vcl && touch skia/gdiimpl.cxx && make

And I see. Maybe there is a build tool getting rebuilt?:

/Volumes/LOBuilds/lode/dev/core/workdir/LexTarget/connectivity/source/parse/sqlflex.cxx:6117:14:
 warning: code will never be executed [-Wunreachable-code]
if ( ! (yy_did_buffer_switch_on_eof) )
^~~
/Volumes/LOBuilds/lode/dev/core/workdir/LexTarget/connectivity/source/parse/sqlflex.cxx:6445:14:
 warning: code will never be executed [-Wunreachable-code]
if ( ! (yy_did_buffer_switch_on_eof) )
^~~
/Volumes/LOBuilds/lode/dev/core/workdir/LexTarget/connectivity/source/parse/sqlflex.cxx:6440:5:
 warning: unannotated fall-through between switch labels 
[-Wimplicit-fallthrough]
case EOB_ACT_END_OF_FILE:
^
/Volumes/LOBuilds/lode/dev/core/workdir/LexTarget/connectivity/source/parse/sqlflex.cxx:6440:5:
 note: insert '[[fallthrough]];' to silence this warning
case EOB_ACT_END_OF_FILE:
^
[[fallthrough]];  
/Volumes/LOBuilds/lode/dev/core/workdir/LexTarget/connectivity/source/parse/sqlflex.cxx:6440:5:
 note: insert 'break;' to avoid fall-through
case EOB_ACT_END_OF_FILE:
^
break;  3 warnings generated.
[DEP] LNK:Library/libvcllo.dylib.d
[LNK] Library/libvcllo.dylib

Re: Compiling nearly every file now compiles sqlbison.cxx and/or sqlflex.cxx

[Caolán McNamara]

On Sun, 2024-05-19 at 13:22 -0400, Patrick Luby wrote:
> A couple of weeks ago I started seeing my builds slow down and my
> build log filled with compiler warnings in sqlbison.cxx and
> sqlflex.cxx on macOS.
> 
> Is anyone else seeing this? I see these warnings even when compiling
> files in vcl. Seems to me that something in the build is not working
> right since I see one of these files is recompiled for each and every
> other C++ file that is compiled.

That seems very odd. Checking a gerrit macos build log like
https://ci.libreoffice.org/job/gerrit_mac/167726/consoleFull
I just see the expected handful of sqlbison and sqlflex warnings as
those individual files are compiled once.

I can't see any reason why compiling vcl would have any connection to
something in connectivity.

What is your configure line? wondering if there is some
--enable-mergelibs or --enable-pch thing that could be a trigger.

Re: Build fail in mariadb on fedora 40

[Julien Nabet]

On 18/05/2024 14:51, Dr. David Alan Gilbert wrote:

...
It looks like others started hitting it; see:
   https://gerrit.libreoffice.org/c/core/+/167806
(I believe as a windows user)


Ok.

BTW, I could run a full build on Win10 with master sources updated today 
(but without enable-dbgutil).


I did a make clean and build again with enable-dbgutil.

Exception enable-dbgutil, the autogen.input contains only:

--with-visual-studio=2022

--with-parallelism=1

(this last one is mandatory for me because the build hangs regularly 
without it whereas I disabled Windows Defender on the whole cygwin 
directory).



Hmm I'm using gcc instead, but I don't see how the cast
would be valid:


I'm giving a try right now with gcc instead of clang and with a 
simplified autogen.input (compared with previous one):


-disable-compiler-plugins
--enable-dbus
--enable-werror
--enable-dependency-tracking
--enable-python=fully-internal
--enable-symbols
--with-referenced-git=/home/julien/lo/libreoffice
--with-external-tar=/home/julien/lo/libreoffice/external/tarballs
--enable-avahi
--with-lang=en-US fr



...
so those C_MODE_START/END make me think someone is doing something
funky with the compiler anyway.


Indeed!

Re: Build fail in mariadb on fedora 40

[Dr. David Alan Gilbert]

* Julien Nabet (serval2...@yahoo.fr) wrote:
> On 17/05/2024 15:47, Dr. David Alan Gilbert wrote:
> > Hi,
> >I'm getting:
> > 
> > workdir/UnpackedTarball/mariadb-connector-c/libmariadb/ma_client_plugin.c:87:207:
> >  error: initialization of ‘struct st_mysql_client_plugin *’ from 
> > incompatible pointer type ‘struct st_mysql_client_plugin_AUTHENTICATION *’ 
> > [-Wincompatible-pointer-types]
> > 87 |(struct st_mysql_client_plugin *)_socket_client_plugin, 
> > (struct st_mysql_client_plugin *)_sha2_password_client_plugin, 
> > (struct st_mysql_client_plugin *)_native_password_client_plugin, 
> > (struct st_mysql_client_plugin_AUTHENTICATION 
> > *)_gssapi_client_client_plugin,
> >|
> > 
> >^
> > /discs/fast/core/workdir/UnpackedTarball/mariadb-connector-c/libmariadb/ma_client_plugin.c:87:207:
> >  note: (near initialization for ‘mysql_client_builtins[3]’)
> > make[1]: *** [/discs/fast/core/solenv/gbuild/LinkTarget.mk:366: 
> > /discs/fast/core/workdir/GenCObject/UnpackedTarball/mariadb-connector-c/libmariadb/ma_client_plugin.o]
> >  Error 1
> > make[1]: *** Waiting for unfinished jobs
> > make: *** [Makefile:294: build] Error 2
> > 
> > Host: Fedora 40, x86-64
> > configured with:
> > ./configure --srcdir=/discs/fast/core --enable-option-checking=fatal 
> > --enable-debug
> > 
> > using --with-system-mariadb fixed it.
> 
> Hello David,

Hi Julien,
  Thanks for the reply.

> Sorry, I don't  know why you encounter this :-(

It looks like others started hitting it; see:
  https://gerrit.libreoffice.org/c/core/+/167806
(I believe as a windows user)

> If it can help, here's the content of my autogen.input:
> 
> CC=clang
> CXX=clang++

Hmm I'm using gcc instead, but I don't see how the cast
would be valid:

UnpackedTarball/mariadb-connector-c/include/mysql/client_plugin.h has:

#include 

struct st_mysql_client_plugin_AUTHENTICATION
{
  MYSQL_CLIENT_PLUGIN_HEADER
  int (*authenticate_user)(MYSQL_PLUGIN_VIO *vio, struct st_mysql *mysql);
};

although, hmm, that header also defines:
#define mysql_declare_client_plugin(X)  \
 struct st_mysql_client_plugin_ ## X\
_mysql_client_plugin_declaration_ = {   \
  MYSQL_CLIENT_ ## X ## _PLUGIN,\
  MYSQL_CLIENT_ ## X ## _PLUGIN_INTERFACE_VERSION,
#define mysql_end_client_plugin }

which is curiously different from the system 
/usr/include/mysql/server/mysql/client_plugin.h

#define mysql_declare_client_plugin(X)  \
 C_MODE_START MYSQL_PLUGIN_EXPORT_C \
struct st_mysql_client_plugin_ ## X\
_mysql_client_plugin_declaration_ = {   \
  MYSQL_CLIENT_ ## X ## _PLUGIN,\
  MYSQL_CLIENT_ ## X ## _PLUGIN_INTERFACE_VERSION,
#define mysql_end_client_plugin }; C_MODE_END

so those C_MODE_START/END make me think someone is doing something
funky with the compiler anyway.

Dave


> --enable-ld=lld
> --enable-online-update
> --enable-dbgutil
> --enable-evolution2
> --enable-gtk4
> --enable-qt5
> --enable-kf5
> --enable-gtk3-kde5
> --enable-skia=debug
> --enable-ext-nlpsolver
> --enable-ext-numbertext
> --enable-ext-wiki-publisher
> --enable-dbus
> --enable-werror
> --enable-dependency-tracking
> --enable-python=fully-internal
> --without-system-mariadb
> --enable-bundle-mariadb
> --enable-symbols
> --enable-avahi
> --enable-eot
> --enable-odk
> --with-lang=en-US de es fr hu it ja nl pt pt-BR ru nb nn
> --with-myspell-dicts
> 
> Julien
> 
-- 
 -Open up your eyes, open up your mind, open up your code ---   
/ Dr. David Alan Gilbert|   Running GNU/Linux   | Happy  \ 
\dave @ treblig.org |   | In Hex /
 \ _|_ http://www.treblig.org   |___/

[Bug 161143] Updated to version 24.2.3.2 (x86_64) and Calc autofilter format error when file re-opened to continue working on it

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=161143

--- Comment #4 from ady  ---
Was attachment 194181 created in LO? Or was this file created by some other
tool instead?

Was the file initially created using an ODS format? Or was it originally
created with some other file format and then converted to ODS?

Once those questions are answered... The next question is why in older versions
there was no visual problem (e.g. LO 7.5).

So, is this some modification in newer LO versions that is intentional and
there is no solution other than having to modify the file? Or, instead, is this
really a bug in newer LO versions?

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

[Bug 161143] Updated to version 24.2.3.2 (x86_64) and Calc autofilter format error when file re-opened to continue working on it

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=161143

m_a_riosv  changed:

   What|Removed |Added

 Ever confirmed|0   |1
 Status|UNCONFIRMED |NEEDINFO

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

[Bug 161143] Updated to version 24.2.3.2 (x86_64) and Calc autofilter format error when file re-opened to continue working on it

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=161143

--- Comment #3 from m_a_riosv  ---
Even simplest, for the header cells, select them, click on background color
button and click in No fill.

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

[Bug 161143] Updated to version 24.2.3.2 (x86_64) and Calc autofilter format error when file re-opened to continue working on it

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=161143

--- Comment #2 from m_a_riosv  ---
Created attachment 194181
  --> https://bugs.documentfoundation.org/attachment.cgi?id=194181=edit
Sample from the reporter

Please use this site to answer and attach files.

The file has a format that shows in white those header fields.

The default style has a green as background color, that makes their text showed
in white.
The text in first line have a default style with modifications.

Editing Default style and put their background to none.

For that first line, apply Default style, and set up the bold and the size you
like.

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

Re: Build fail in mariadb on fedora 40

[Julien Nabet]

On 17/05/2024 15:47, Dr. David Alan Gilbert wrote:

Hi,
   I'm getting:

workdir/UnpackedTarball/mariadb-connector-c/libmariadb/ma_client_plugin.c:87:207:
 error: initialization of ‘struct st_mysql_client_plugin *’ from incompatible 
pointer type ‘struct st_mysql_client_plugin_AUTHENTICATION *’ 
[-Wincompatible-pointer-types]
87 |(struct st_mysql_client_plugin *)_socket_client_plugin, (struct 
st_mysql_client_plugin *)_sha2_password_client_plugin, (struct 
st_mysql_client_plugin *)_native_password_client_plugin, (struct 
st_mysql_client_plugin_AUTHENTICATION *)_gssapi_client_client_plugin,
   |

   ^
/discs/fast/core/workdir/UnpackedTarball/mariadb-connector-c/libmariadb/ma_client_plugin.c:87:207:
 note: (near initialization for ‘mysql_client_builtins[3]’)
make[1]: *** [/discs/fast/core/solenv/gbuild/LinkTarget.mk:366: 
/discs/fast/core/workdir/GenCObject/UnpackedTarball/mariadb-connector-c/libmariadb/ma_client_plugin.o]
 Error 1
make[1]: *** Waiting for unfinished jobs
make: *** [Makefile:294: build] Error 2

Host: Fedora 40, x86-64
configured with:
./configure --srcdir=/discs/fast/core --enable-option-checking=fatal 
--enable-debug

using --with-system-mariadb fixed it.


Hello David,

Sorry, I don't  know why you encounter this :-(

If it can help, here's the content of my autogen.input:

CC=clang
CXX=clang++
--enable-ld=lld
--enable-online-update
--enable-dbgutil
--enable-evolution2
--enable-gtk4
--enable-qt5
--enable-kf5
--enable-gtk3-kde5
--enable-skia=debug
--enable-ext-nlpsolver
--enable-ext-numbertext
--enable-ext-wiki-publisher
--enable-dbus
--enable-werror
--enable-dependency-tracking
--enable-python=fully-internal
--without-system-mariadb
--enable-bundle-mariadb
--enable-symbols
--enable-avahi
--enable-eot
--enable-odk
--with-lang=en-US de es fr hu it ja nl pt pt-BR ru nb nn
--with-myspell-dicts

Julien

[Bug 161143] Updated to version 24.2.3.2 (x86_64) and Calc autofilter format error when file re-opened to continue working on it

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=161143

m_a_riosv  changed:

   What|Removed |Added

 CC||miguelangelrv@libreoffice.o
   ||rg

--- Comment #1 from m_a_riosv  ---
Please attach a sample file.

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

Re: tdf#157003: interactive Java unit test debugging

[Hossein Nourikhah]

Hello Martin,

If you remember from the previous problem with this test, I could run it 
this way on Windows, as displayed on the command line:


java -Xmx64M -classpath 
"$W/JavaClassSet/JunitTest/sw_complex;C:/cygwin64/home/user/lode/opt/share/java/junit.jar;$I/program;$W/Jar/OOoRunner.jar;$I/program/classes/libreoffice.jar;$W/Jar/test.jar" 
-Dorg.openoffice.test.arg.soffice="path:$I/program/soffice" 
-Dorg.openoffice.test.arg.env=PATH="$PATH" 
-Dorg.openoffice.test.arg.user=file:///$W/JunitTest/sw_complex/user 
org.junit.runner.JUnitCore 
complex.indeterminateState.CheckIndeterminateState


You can find similar command on your platform from the make output, and 
then set up a Java debugger/IDE to debug this Java command line.


Btw, one of the goals of porting complex test from Java to Python/C++ is 
to ease the debugging.


Regards,
Hossein

On 17.05.2024 12:05, Martin Gube wrote:

Hi Developers,

I'd like to interactively debug the Java unit test
JunitTest_sw_complex (CheckIndeterminateState.java) for further work
on tdf#157003. Can anyone help me on how to do this?

In case of failing unit tests, the make process output after invoking
"make JunitTest_sw_complex" suggests the following:
"Or to do interactive debugging, run two shells with:
make debugrun
make gb_JunitTest_DEBUGRUN=T JunitTest_sw_complex"

However "make gb_JunitTest_DEBUGRUN=T JunitTest_sw_complex" results in
a virtually never ending make process for me. I tried this few times
and canceled it after several hours without visible progress.

Also I tried to run JunitTest_sw_complex inside GDB in different ways,
without success. However, I'm wondering if GDB is able to debug Java
[1].
I also tried to use JDB instead of GDB but got stuck with making the
unit test run inside JDB.

Can anyone help me how to interactively debug Java unit tests for
LibreOffice?

Many thanks for your help + best Regards,
Martin
P.S.: please excuse me in case you got this mail twice. I already sent
it yesterday, however it appeared like not properly sent.
 [1] https://www.sourceware.org/gdb/


--
Hossein Nourikhah, Ph.D., Developer Community Architect
Tel: +49 30 5557992-65 | Email: hoss...@libreoffice.org
The Document Foundation, Winterfeldtstraße 52, 10781 Berlin, DE
Gemeinnützige rechtsfähige Stiftung des bürgerlichen Rechts
Legal details: https://www.documentfoundation.org/imprint

Re: Merging feature/windows-cross-build

[Don Grabar]

[Bug 161143] New: Updated to version 24.2.3.2 (x86_64) and Calc autofilter format error when file re-opened to continue working on it

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=161143

Bug ID: 161143
   Summary: Updated to version 24.2.3.2 (x86_64) and Calc
autofilter format error when file re-opened to
continue working on it
   Product: LibreOffice
   Version: 24.2.3.2 release
  Hardware: Other
OS: Windows (All)
Status: UNCONFIRMED
  Severity: normal
  Priority: medium
 Component: Calc
  Assignee: libreoffice-bugs@lists.freedesktop.org
  Reporter: grinaho...@aol.com

Description:
Updated to version 24.2.3.2 (x86_64) and Calc autofilter data error when file
opened to continue working on it. The heading row of the autofilter had the
cell info but was invisible in the cell. You can see the data in the top
equation line but not in the cell. However, it only did this for four of the
columns. This issue happened on every sheet in the workbook.

Steps to Reproduce:
1. n/a
2.
3.

Actual Results:
n/a - no other computers to upgrade version

Expected Results:
n/a


Reproducible: Didn't try


User Profile Reset: Yes

Additional Info:
Version: 24.2.3.2 (X86_64) / LibreOffice Community
Build ID: 433d9c2ded56988e8a90e6b2e771ee4e6a5ab2ba
CPU threads: 8; OS: Windows 10.0 Build 22621; UI render: Skia/Raster; VCL: win
Locale: en-US (en_US); UI: en-US
Calc: threaded

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

[Bug 89519] re-export in PPT - some drawings are lost (but not in PPTX) (WPS)

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=89519

--- Comment #5 from QA Administrators  ---
Dear Ljiljan,

To make sure we're focusing on the bugs that affect our users today,
LibreOffice QA is asking bug reporters and confirmers to retest open, confirmed
bugs which have not been touched for over a year.

There have been thousands of bug fixes and commits since anyone checked on this
bug report. During that time, it's possible that the bug has been fixed, or the
details of the problem have changed. We'd really appreciate your help in
getting confirmation that the bug is still present.

If you have time, please do the following:

Test to see if the bug is still present with the latest version of LibreOffice
from https://www.libreoffice.org/download/

If the bug is present, please leave a comment that includes the information
from Help - About LibreOffice.

If the bug is NOT present, please set the bug's Status field to
RESOLVED-WORKSFORME and leave a comment that includes the information from Help
- About LibreOffice.

Please DO NOT

Update the version field
Reply via email (please reply directly on the bug tracker)
Set the bug's Status field to RESOLVED - FIXED (this status has a particular
meaning that is not 
appropriate in this case)


If you want to do more to help you can test to see if your issue is a
REGRESSION. To do so:
1. Download and install oldest version of LibreOffice (usually 3.3 unless your
bug pertains to a feature added after 3.3) from
https://downloadarchive.documentfoundation.org/libreoffice/old/

2. Test your bug
3. Leave a comment with your results.
4a. If the bug was present with 3.3 - set version to 'inherited from OOo';
4b. If the bug was not present in 3.3 - add 'regression' to keyword


Feel free to come ask questions or to say hello in our QA chat:
https://web.libera.chat/?settings=#libreoffice-qa

Thank you for helping us make LibreOffice even better for everyone!

Warm Regards,
QA Team

MassPing-UntouchedBug

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

Re: [libreoffice-design] Re: behavior of new functions XLOOKUP and XMATCH

[Wol]

On 15/05/2024 08:06, Heiko Tietze wrote:

If you want, we can run a poll what users prefer.


If you can't understand what's wanted - DON'T GUESS. Especially don't 
guess and hide that fact from your users!


An error of "#value" or similar seems a pretty good option ...


Cheers,

Wol

Re: [libreoffice-design] Re: behavior of new functions XLOOKUP and XMATCH

[Heiko Tietze]

Re-adding the dev ML; Regina's full reply at
https://listarchives.libreoffice.org/global/design/2024/msg00045.html

On 15.05.24 2:03 AM, Regina Henschel wrote:
The problem is, what should be the output, when the user sets Match_mode 2 
together with Search_mode 2 or -2.

Excel returns #VALUE! for the combination =XLOOKUP(C3;A$1:A$3;B$1:B$3;"INV";2;2)
while we do some processing.

On 14.05.24 5:04 AM, Mike Kaganski wrote:

Emitting error is the only sane way here.Absolutely.


It is not some kind of interruptive error box but a typical way to give feedback 
on typos or for unspecified content. Silently switching to some other way of 
processing is almost impossible to understand for users. And I cannot follow the 
arguments that no error message makes the handling easier.


If you want, we can run a poll what users prefer.



OpenPGP_signature.asc
Description: OpenPGP digital signature

Re: behavior of new functions XLOOKUP and XMATCH

[Thorsten Behrens]

Hi *,

Mike Kaganski wrote:
> Emitting error is the only sane way here. No error message means doing
> something different than user requested, without user knowing that, with no
> practical benefit (the wildcards / regexes won't be used, so the task is not
> done anyway).
>
My understanding is, the regex _would_ be used, but in a much slower,
linear search.

But from a user-perspective, I'd prolly agree with Mike. To add to
that, for practical matters and importing XLSX files, we would very
likely need a compatibility mode for XLOOKUP/XMATCH anyway, with the
same error handling as Excel...

Cheers,

-- Thorsten


signature.asc
Description: PGP signature

Re: behavior of new functions XLOOKUP and XMATCH

[Heiko Tietze]

Correct me but

=XLOOKUP(C3;A$1:A$3;B$1:B$3;"INV";_3_;1) returns #VALUE! (Excel) or Err:504 (LO)
=_Y_LOOKUP(C4;A$1:A$3;B$1:B$3;"INV";0;1) returns #NAME? (both)

What exactly will change?

PS: ux-advice@ is the forward mailing list from Bugzilla and I'm not sure how 
many people have registered. Rather use design@ for input from the UI/UX group.


On 14.05.24 12:23 AM, Regina Henschel wrote:

Hi UX-experts,

the new function XLOOKUP and XMATCH have a Match_mode with values 0, 1, -1, 2 
and a Search_mode with values 1, -1, 2, -2.
The Match_mode 2 means wildcard mode, that is ? * search or regular expressions 
search. The Search_modes 2 and -2 mean binary search in a sorted array.


Combining parameter values for wildcard mode and binary search is possible. But 
there exist no way to actually do it. Excel gives an error messages in such 
cases. LibreOffice silently switches to linear search.


The ODF TC is currently working on the specification for these new functions. 
The TC could either follow Excel's way or LibreOffice's way. From a developer 
state of view it is no large effort to change the current behavior of 
LibreOffice. The TC does not know yet whether Microsoft would be willing to 
switch to LibreOffice's behavior.


I write to you for to get your opinion what will be the best solution from a 
_user_ point of view.


Some arguments so far:
* No error message would make use of the functions in macros and forms easier.
* No error message might be better for using the functions in automatically 
calculations without UI.
* Using linear search can be unexpected slow on a huge Lookup_array. An error 
message makes the problem visible to the user.


I would love to hear your opinion.

Kind regards,
Regina







--
Dr. Heiko Tietze, UX-Architect and UI-Designer
Tel: +49 30 5557992-63 | Mail: heiko.tie...@documentfoundation.org
The Document Foundation, Winterfeldtstraße 52, 10781 Berlin, DE
Gemeinnützige rechtsfähige Stiftung des bürgerlichen Rechts
Legal details: https://www.documentfoundation.org/imprint


OpenPGP_signature.asc
Description: OpenPGP digital signature

Re: behavior of new functions XLOOKUP and XMATCH

[Mike Kaganski]

On 14.05.2024 3:23, Regina Henschel wrote:
Combining parameter values for wildcard mode and binary search is 
possible. But there exist no way to actually do it. Excel gives an error 
messages in such cases. LibreOffice silently switches to linear search.


...

Some arguments so far:
* No error message would make use of the functions in macros and forms 
easier.
* No error message might be better for using the functions in 
automatically calculations without UI.
* Using linear search can be unexpected slow on a huge Lookup_array. An 
error message makes the problem visible to the user.


Emitting error is the only sane way here. No error message means doing 
something different than user requested, without user knowing that, with 
no practical benefit (the wildcards / regexes won't be used, so the task 
is not done anyway). Calculations without UI means either prepared 
document (and there a sane set of options must be used), or 
programmatically defined formulas (and there a sane set of options must 
be used).


--
Best regards,
Mike Kaganski

Re: Windows build broken.

[Ilmari Lauhakangas]

On 13.5.2024 11.09, Regis Perdreau wrote:


Hi,

I follow carefully again windows build instruction wiki, but the windows 
build does not work again. Could you we expect that it will work for 
everybody not only on developper's computer ? (i know it's sometimes 
difficult, but if we could have some news, it's better)


Please share the text output showing how the build fails.

Ilmari

Re: OSS-fuzzing

[Caolán McNamara]

On Sun, 2024-05-12 at 21:00 +0300, Даблус wrote:
> Hello. I want to try to run oss-fuzz tests by myself, but i found no
> info about it. I found some scripts in the "bin" folder, like oss-
> fuzz-setup.sh and oss-fuzz-build.sh, and found that I should use
> clang, but I found nothing about the environment. I can build
> LibreOffice with clang, but if i run oss-fuzz-build, i get a lot of
> errors after postprocess step
> 
> What environment should i use and how should i run fuzzing correctly?

The fuzzing stuff is organized for the oss-fuzz environment so that's
the case that gets used and is known to work so:

git clone https://github.com/google/oss-fuzz.git
cd oss-fuzz
python infra/helper.py shell libreoffice
compile

should successfully build the default fuzzer config in the oss-fuzz
provided docker images

https://google.github.io/oss-fuzz/advanced-topics/debugging/#debugging-build-scripts

[Bug 140411] Calc constantly re-calculating (bug in the sheet?)

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=140411

--- Comment #5 from QA Administrators  ---
Dear Dan,

This bug has been in NEEDINFO status with no change for at least
6 months. Please provide the requested information as soon as
possible and mark the bug as UNCONFIRMED. Due to regular bug
tracker maintenance, if the bug is still in NEEDINFO status with
no change in 30 days the QA team will close the bug as INSUFFICIENTDATA
due to lack of needed information.

For more information about our NEEDINFO policy please read the
wiki located here:
https://wiki.documentfoundation.org/QA/Bugzilla/Fields/Status/NEEDINFO

If you have already provided the requested information, please
mark the bug as UNCONFIRMED so that the QA team knows that the
bug is ready to be confirmed.

Thank you for helping us make LibreOffice even better for everyone!

Warm Regards,
QA Team

MassPing-NeedInfo-Ping

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

Re: tdf132599_frames_* tests needs hu-HU hyphenation for english text

[Rene Engelhard]

Hi,

Am 10.05.24 um 18:58 schrieb Rene Engelhard:
I think the files should be fixed to set the text language to English 
(USA) instead?



https://gerrit.libreoffice.org/c/core/+/167486.

FTR that one didn't work (as written in my mail)but Laszlo did 
https://gerrit.libreoffice.org/c/core/+/167512 which fixes it (and got 
just merged).


Thanks for the quick fix.


Regards,

Rene

Re: tdf132599_frames_* tests needs hu-HU hyphenation for english text

[Rene Engelhard]

Hi again,

Am 10.05.24 um 17:27 schrieb Rene Engelhard:


Am 10.05.24 um 17:18 schrieb Rene Engelhard:

[...]
The test code sets en_US as locale, and the text files have english 
text.


Well, sets is wrong, by bad.

But it checks for it and skips the test if en_US isn't there.

But LibreOffice suggests to install hu-HU hyphenation patterns when 
opening the .fodt. Indeed, the text is formatted as Hungarian.


but as said it expects hu-HU since the text is formatted as Hungarian.


I think the files should be fixed to set the text language to English 
(USA) instead?



https://gerrit.libreoffice.org/c/core/+/167486.

(Probably bigger than needed since I used my workstations Debian stable 
LO for it.


Feel free to do it less invasive to the .fodt diff if possible :)

I don't know how minimal the diff was is if one edited it with a newer LO.)


But I got only down to

Test name: (anonymous 
namespace)::testTdf132599_frames_on_same_page_hyphenation::TestBody

equality assertion failed
- Expected: cept that it
- Actual  : space, ex
- In <>, attribute 'portion' of 
'/root/page/body/txt/anchored/fly[2]/txt/SwParaPortion/SwLineLayout[1]' 
incorrect value.


xmltesttools.cxx:170:Assertion
Test name: (anonymous 
namespace)::testTdf132599_frames_on_spread_hyphenation::TestBody

equality assertion failed
- Expected: cept that it
- Actual  : space,
- In <>, attribute 'portion' of 
'/root/page[3]/body/txt/anchored/fly/txt/SwParaPortion/SwLineLayout[1]' 
incorrect value.


Failures !!!
Run: 95   Failure total: 2   Failures: 2   Errors: 0
make: *** 
[/home/rene/LibreOffice/git/master/solenv/gbuild/CppunitTest.mk:130: 
/home/rene/LibreOffice/git/master/workdir/CppunitTest/sw_layoutwriter3.test] 
Error 1


and yes, I see that in that (obviously, since this is with 7.6.7) when I 
open  the document. But shouldn' t  that be a non-issue if tested with a 
"correct" LO?



Interestingly Jenkins' builds work fine here even with my new documents..


Do I miss something here?


Regards,


Rene

Re: tdf132599_frames_* tests needs hu-HU hyphenation for english text

[Rene Engelhard]

Hi,

Am 10.05.24 um 17:18 schrieb Rene Engelhard:

[...]
The test code sets en_US as locale, and the text files have english text.


Well, sets is wrong, by bad.

But it checks for it and skips the test if en_US isn't there.

But LibreOffice suggests to install hu-HU hyphenation patterns when 
opening the .fodt. Indeed, the text is formatted as Hungarian.


but as said it expects hu-HU since the text is formatted as Hungarian.


I think the files should be fixed to set the text language to English 
(USA) instead?




(I just work around it by adding hyphen-hu as a build dependency, but 
that should be fixed).



https://salsa.debian.org/libreoffice-team/libreoffice/libreoffice/-/commit/1330f842bec975ba616359d9a0cc05a2f37e7a00


Regards,


Rene

Re: ESC meeting minutes: 2024-05-09

[Timur]

On 10.5.2024. 07:53, Stephan Bergmann wrote:

On 5/9/24 16:53, Miklos Vajna wrote:
 + MAR-based updater will be enabled by default on Windows, but 
not Linux (Hossein)

   + rpm/deb update is handled by distro packages
   + for the ones taking binaries from TDF
   + TDF to provide rpm/deb repos?
   + the files are hosted already
   + just the repo defs are missing
   + suggesting to provide these repos
   + proposal is to povide repos and ignore MAR on Linux for now
   + interested in doing the scripting work? (Miklos)
 + the important part is the description / dependencies 
(Hossein)

 + willing to help
   + linux downloads are 2% of TDF downloads (Cloph)
 + so not a huge amount, but not objecting
 + only the the ones which are covered by rpm/deb binaries
   + what is the lag between Ubuntu-provided packages / TDF 
releases? (Olivier)

 + the package from Ubuntu follows TDF quickly
 + similar with Debian, thanks to Rene (Cloph)
   + with actively supported versions
 + see LibreOffice 7.3.7 for Ubuntu 22.04 
https://packages.ubuntu.com/jammy/libreoffice (Hossein)


And, just to mention it again, for Linux there's also always the 
Flatpak version on Flathub, 
.  I'm still 
providing builds there on a somewhat voluntary basis, usually as soon 
as a new version of LO gets released.



No wonder that Linux downloads are small, not only because there may be 
more Windows users, but also because LO is already in distro, and for 
Ubuntu-flavor users there is PPA (1). Link would be useful at download 
page (2).


If "TDF to provide rpm/deb repos" is done, it is still of benefit to 
other Linux users, and link is also welcome.


https://launchpad.net/~libreoffice/+archive/ubuntu/ppa

https://www.libreoffice.org/download/download-libreoffice/

Re: ESC meeting minutes: 2024-05-09

[Stephan Bergmann]

On 5/9/24 16:53, Miklos Vajna wrote:
     + MAR-based updater will be enabled by default on Windows, but not 
Linux (Hossein)

   + rpm/deb update is handled by distro packages
   + for the ones taking binaries from TDF
   + TDF to provide rpm/deb repos?
   + the files are hosted already
   + just the repo defs are missing
   + suggesting to provide these repos
   + proposal is to povide repos and ignore MAR on Linux for now
   + interested in doing the scripting work? (Miklos)
     + the important part is the description / dependencies (Hossein)
     + willing to help
   + linux downloads are 2% of TDF downloads (Cloph)
     + so not a huge amount, but not objecting
     + only the the ones which are covered by rpm/deb binaries
   + what is the lag between Ubuntu-provided packages / TDF 
releases? (Olivier)

     + the package from Ubuntu follows TDF quickly
     + similar with Debian, thanks to Rene (Cloph)
   + with actively supported versions
     + see LibreOffice 7.3.7 for Ubuntu 22.04 
https://packages.ubuntu.com/jammy/libreoffice (Hossein)


And, just to mention it again, for Linux there's also always the Flatpak 
version on Flathub, 
.  I'm still 
providing builds there on a somewhat voluntary basis, usually as soon as 
a new version of LO gets released.

Re: ESC meeting minutes: 2024-05-09

[Stephan Bergmann]

On 5/9/24 16:53, Miklos Vajna wrote:
     + MAR-based updater will be enabled by default on Windows, but not 
Linux (Hossein)

   + rpm/deb update is handled by distro packages
   + for the ones taking binaries from TDF
   + TDF to provide rpm/deb repos?
   + the files are hosted already
   + just the repo defs are missing
   + suggesting to provide these repos
   + proposal is to povide repos and ignore MAR on Linux for now
   + interested in doing the scripting work? (Miklos)
     + the important part is the description / dependencies (Hossein)
     + willing to help
   + linux downloads are 2% of TDF downloads (Cloph)
     + so not a huge amount, but not objecting
     + only the the ones which are covered by rpm/deb binaries
   + what is the lag between Ubuntu-provided packages / TDF 
releases? (Olivier)

     + the package from Ubuntu follows TDF quickly
     + similar with Debian, thanks to Rene (Cloph)
   + with actively supported versions
     + see LibreOffice 7.3.7 for Ubuntu 22.04 
https://packages.ubuntu.com/jammy/libreoffice (Hossein)


And, just to mention it again, for Linux there's also always the Flatpak 
version on Flathub, 
.  I'm still 
providing builds there on a somewhat voluntary basis, usually as soon as 
a new version of LO gets released.

Re: Which fonts are available on the build bots?

[Caolán McNamara]

On Thu, 2024-05-09 at 20:16 +0200, Regina Henschel wrote:
> Which fonts are available on all build bots?

Only the fonts that are bundled with LibreOffice, and we try and
basically force the same restriction to those fonts as best we can
given platform restrictions so we're all running the tests with the
same fonts.

> Is there a font in place, which covers "CJK Unified Ideographs"?

AFAIK there isn't a bundled font that covers CJK.

*but* what have done in similar case before is documented in:
vcl/qa/cppunit/data/tdf153440.ttf.readme and
vcl/qa/cppunit/data/tdf107718.otf.readme
so maybe the little font of
vcl/qa/cppunit/data/tdf107718.otf
might even be sufficient for your testing needs. See
testMixedCJKLatinScript_glyph_advancements for how to use such a "temp"
user font.

[Bug 92375] Impress loses cropped image ratio on saving and re-opening in different LO versions.

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=92375

--- Comment #45 from Christian Assig  ---
Heiko and pranav: My Python script should be able to repair your files with
size problems, see Comment 41.

Heiko: The difference between your three PCs is most probably the display
scaling settings.

To see how your display scaling is configured and how to change it on Windows
11, see for example the following documentation:

https://community.microcenter.com/kb/articles/746-how-to-adjust-display-scaling-in-windows-11
(in English)

https://basic-tutorials.de/ratgeber/software/windows-11-so-laesst-sich-die-dpi-skalierungsstufe-der-anzeige-aendern/
(in German, the setting you are looking for is called "Skalierung" in German)

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

Re: GSoC: Adding native support for histogram chart

[Devansh Varshney]

Hi everyone,

For this week, my top priority is to make the new Histogram chart type
selectable from the UI.

Initially, it will be a working copy of the existing basic bar/column chart
type.

Then, I'll make the required changes with each passing iteration.

Here's the link to the PR: https://gerrit.libreoffice.org/c/core/+/167068/


Best regards,

Devansh

On Tue, 7 May 2024 at 18:05, Devansh Varshney 
wrote:

> Hi everyone,
>
> I am Devansh and this year I have been selected for the* Google Summer of
> Code (GSoC) program with LibreOffice* to *Add the support for the
> Histogram Chart* - GSoC Idea list (
> https://wiki.documentfoundation.org/Development/GSoC/Ideas#Histogram_.28and_other_missing.29_Chart_Type
> ) and *Tomaž Vajngerl* is going to be* my mentor,* who has guided me
> throughout the proposal process and will continue to do so until the
> project's completion.
>
> I gravitated towards adding the support for the Histogram chart (and if
> time permits other missing charts also from the MS Office 2016) stems from
> my family background in science, engineering, and finance, where data
> visualization plays a crucial role. Working on this project allows me to
> connect with my family's work and explain my contributions in a meaningful
> way.
>
> Second, while preparing the proposal report I came across that the *CERN*
> also uses the LibreOffice for their work is doing a work around because of
> no native support of these new chart types. (
> https://indico.cern.ch/event/977476/contributions/4204938/attachments/2191304/3703670/EN-Plotting-a-histogram-in-libreoffice.pdf)
> which was the point where I can see the impact of this task. Not only this
> *IndiaMart*, a prominent B2B marketplace where my friends work told me
> that it also uses LibreOffice.
>
> Seeing the potential impact of this project on real-world users was a
> significant motivator for me. This project checks all the right boxes for
> me.
>
> I have been impressed by the LibreOffice community and documentation,
> which are among the best I've seen.
> I'd like to extend my gratitude to the community members who have helped
> me get started with my contributions, including Hossein, Heiko, Michael,
> Marco, and especially Ilmari, who was the first to guide me.
>
> Regards,
> Devansh 
> (IRC: devansh)
>

[Bug 76131] Existing pinned icon on Win7/8/10/11's taskbar is invalid after re-installation/update and start menu icons are swapped (W10/W11)

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=76131

--- Comment #89 from vistama...@web.de ---
indeed, I agree.

This issue looks unimportant to the developer, but as a closer look it is
crucial for the whole community.

I belong to those thousands of users, who do not install newer versions any
more, because I do not want to do this pin-work over and over again. T

In other words: IHO this issue reduces the count of people installing the most
recent version remarkable.

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

[Bug 92375] Impress loses cropped image ratio on saving and re-opening in different LO versions.

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=92375

--- Comment #44 from Heiko  ---
I am also experiencing this issue in LO draw. However, it seems not to be
correlated to the version of LO, but rather to the environment:
I am working with three PCs (all windows 11 enterprise).
Two of the PCs are working in sync (LO 7.5.2.2 & LO 7.5.5.2). The third one
opens odg-files with cropped images differently, independent of the LO-version
(tested with 7.6.6, 7.5.2.2 & LO 7.5.5.2). Different from the two other PCs,
but consistent across the three tested versions.
I have sadly no idea, what is the key difference with PC #3.

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

[Bug 76131] Existing pinned icon on Win7/8/10/11's taskbar is invalid after re-installation/update and start menu icons are swapped (W10/W11)

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=76131

V Stuart Foote  changed:

   What|Removed |Added

Version|24.2.3.2 release|4.2.2.1 release

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

[Bug 76131] Existing pinned icon on Win7/8/10/11's taskbar is invalid after re-installation/update and start menu icons are swapped (W10/W11)

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=76131

To  changed:

   What|Removed |Added

Version|4.2.2.1 release |24.2.3.2 release

--- Comment #88 from To  ---
Can confirm this is still the case with 24.2.3.2 would like to see this fixed,
thanks!

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

[Bug 76991] Writer destroys aspect ratio when scaling graphics in re-opened dialog (when "Keep ratio" and "Relative to" are both used)

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=76991

Stéphane Guillou (stragu)  changed:

   What|Removed |Added

Summary|Writer destroys aspect  |Writer destroys aspect
   |ratio when scaling graphics |ratio when scaling graphics
   ||in re-opened dialog (when
   ||"Keep ratio" and "Relative
   ||to" are both used)
   See Also||https://bugs.documentfounda
   ||tion.org/show_bug.cgi?id=55
   ||480
   Hardware|Other   |All

--- Comment #20 from Stéphane Guillou (stragu) 
 ---
Still current in recent daily build:

Version: 24.8.0.0.alpha0+ (X86_64) / LibreOffice Community
Build ID: 7a895ec4205659038aa95941b65715fed1a3e7be
CPU threads: 8; OS: Linux 6.5; UI render: default; VCL: gtk3
Locale: en-AU (en_AU.UTF-8); UI: en-US
Calc: CL threaded

Seems very much related to bug 55480, which also mixes "keep ratio" and
"relative to", and requires re-opening the dialog to reproduce the issue.

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

Re: GSoC project to add CPDB support to the LibreOffice print dialog

[Michael Weghorn]

On 2024-05-06 19:22, Biswadeep Purkayastha wrote:> I am Biswadeep 
Purkayastha, a GSoC 2024 contributor to OpenPrinting. In
the upcoming months I'll be working on getting CPDB support into the 
LibreOffice print dialogs. I had previously reached out to the community 
asking for a mentor to help me from the LibreOffice side and I am happy 
to announce that Michael Weghorn will be mentoring me from the 
LibreOffice side on this project. I look forward to working on this 
project with my mentors and being a part of the open-source community.


Thanks for the update and welcome again!
I'm looking forward to working together and hope you'll have a great 
GSoC experience!


OpenPGP_signature.asc
Description: OpenPGP digital signature

Re: Create a config extension to change a string-list value

[Stephan Bergmann]

On 5/6/24 23:19, Juan C. Sanz wrote:
There is a document in the wiki 
 explaining how to create a configuration extensions.


In this document it is said that the valid property values are string, 
boolean, int, short, long, double, binary.


Also in the Chapter 15 - Configuration Management 
 of the Developers Guide that the same types are supported.


But if you search for some config values in Expert configuration, the 
type may be different of the supported types, for example you con see 
string-list (or depend on version []string).


The defining document for the configuration XML format is 
, and also see the 
officecfg/registry/*.dtd files in the LO core repo.


I use something like this, but changing oor:type="xs:string-list" for 
every of the valid types, and with or without  label:


oor:name="Common" oor:package="org.openoffice.Office" 
xmlns:oor="http://openoffice.org/2001/registry; 
xmlns:xs="http://www.w3.org/2001/XMLSchema;>oor:name="Security">oor:name="SecureURL" oor:type="xs:string-list" 
oor:finalized="false">"$(work)/BD"


Three notes:

* The type is called oor:string-list, not xs:string-list.

* Your string list has one element with literal value "$(work)/BD".  You 
probably don't want surrounding quotes as part of the element's value.


* Whether or not something like $(work) is expanded in some way does not 
depend on the configuration machinery itself, but on the code that 
consumes that specific property.  I have no idea whether that value 
actually means what you would expect it to me.

Re: Which GUI toolkit should I learn so that I can "Hack the UI"

[Pierre Vacher]

I have extensions which display dialog windows in LibreOffice and which use
neither GTK nor QT but only the UNO API.

see: https://prrvchr.github.io/

One of the advantages is that the UNO API allows me to manage all my
dialogs in Python...

Le lundi 6 mai 2024, Michael Weghorn  a écrit :
> On 2024-05-03 19:15, Printf Debugging wrote:
>>
>> **Opinions invited**: I want to work on the UI, beyond small issues,
like creating custom widgets, notebookbar, tabbed UI etc. I feel that I
need to know a GUI toolkit well, like gtk or qt for that. So some days ago,
I started playing around with GTK. My question is that which toolkit will
be more useful? I picked GTK because we use .ui files for the UI, and being
good at GTK will somehow help when working on LO UI, but again I don't
know, that's just an idea.
>
> Starting with GTK generally sounds reasonable to me.
> Concepts are usually very similar in other toolkits (like Qt), so
learning GTK will also help understanding those better.
>
> In any case, I recommend explicitly testing the non-GTK case when doing
UI changes (with SAL_USE_VCLPLUGIN=gen and qt5/qt6/kf5/kf6 on Linux or by
testing on a different platform).
>

Re: Which GUI toolkit should I learn so that I can "Hack the UI"

[Michael Weghorn]

On 2024-05-03 19:15, Printf Debugging wrote:
**Opinions invited**: I want to work on the UI, beyond small issues, 
like creating custom widgets, notebookbar, tabbed UI etc. I feel that I 
need to know a GUI toolkit well, like gtk or qt for that. So some days 
ago, I started playing around with GTK. My question is that which 
toolkit will be more useful? I picked GTK because we use .ui files for 
the UI, and being good at GTK will somehow help when working on LO UI, 
but again I don't know, that's just an idea.


Starting with GTK generally sounds reasonable to me.
Concepts are usually very similar in other toolkits (like Qt), so 
learning GTK will also help understanding those better.


In any case, I recommend explicitly testing the non-GTK case when doing 
UI changes (with SAL_USE_VCLPLUGIN=gen and qt5/qt6/kf5/kf6 on Linux or 
by testing on a different platform).


OpenPGP_signature.asc
Description: OpenPGP digital signature

Re: Which GUI toolkit should I learn so that I can "Hack the UI"

[Miklos Vajna]

Hi,

On Fri, May 03, 2024 at 10:45:01PM +0530, Printf Debugging 
 wrote:
> **Opinions invited**: I want to work on the UI, beyond small issues, like
> creating custom widgets, notebookbar, tabbed UI etc. I feel that I need to
> know a GUI toolkit well, like gtk or qt for that. So some days ago, I
> started playing around with GTK. My question is that which toolkit will be
> more useful? I picked GTK because we use .ui files for the UI, and being
> good at GTK will somehow help when working on LO UI, but again I don't know,
> that's just an idea.

gtk is probably a good idea, given that these days much of the UI is
welded and the LO weld API is designed after gtk.

Regards,

Miklos

Re: [TDF Community] [Board Discuss] LibreOffice - peer2peer collaboration bits

[Heiko Tietze]

On 05.05.24 12:51 PM, Julien Nabet wrote:
Before talking about technical architecture, is there a real will about this? 
Are there enough financial resources for this?
There is will by at least some, see 
https://community.documentfoundation.org/t/budget-request-for-a-p2p-libreoffice-project/12016


The numbers are purely random to me as long we have no idea of the scope.

And I believe it's an average "Eve" end-user scenario since users with free and 
open LibreOffice can connect to business environment such as COOL using the 
desktop or mobile application.


OpenPGP_signature.asc
Description: OpenPGP digital signature

Re: [TDF Community] [Board Discuss] LibreOffice - peer2peer collaboration bits

[Julien Nabet]

Hello,

I'm not against the idea of a collaboration Office but who's gonna work 
on this?
It's not a simple feature where 1 or 2 benevolents can work on it for 
some months and that's all.
There are security issues, integrity of the produced documents, perhaps 
tracking who modified which parts and when, conflict resolving, etc.


Even if the feature is developped by private company, I expect 
bugtrackers would be on TDF. Dealing with these bugs will be more 
complicated:

how to reproduce easily the fact to be behind a company firewall?
how to simulate 2 or more people working on the same doc?

IMHO it required dedicated/paid persons from TDF or from some private 
companies (or both) to work on this. There may be some replacement of 
these people with the time of course, but they always must be identified 
so if there's any problem, we know how to contact and they'll respond in 
a reasonable delay (less than 1 week).
I mean, I expect the target of this feature is mainly companies and 
associations and these can be more picky.
Before talking about technical architecture, is there a real will about 
this? Are there enough financial resources for this?


Julien

Re: [TDF Community] [Board Discuss] LibreOffice - peer2peer collaboration bits

[Heiko Tietze]

On 04.05.24 12:05 PM, Thorsten Behrens wrote:

Heiko wrote:

How to connect two or more individuals? It requires routing

...

What could be achievable on TDF infrastructure?


Given what I've said above - let's try to make this completely
independent of TDF infrastructure. Either with no switching-server
at all or with something minimal that hopefully might not even need
TDF continuously maintaining a server. Note that maintenance by us
also has privacy implications, much more so than third-party-less
P2P.


Yup. At any rate, requiring any kind of centralized server
infrastructure has inevitable scalability challenges. It would still
be useful if TDF could help with bootstrapping whatever server
infrastructure will be needed, though.


I believe it's crucial for the success to have this networking open and free for 
everyone. And I cannot think of many alternatives to TDF, hosting the discovery 
and handshake around LibreOffice Technology. While we clearly should not route 
the actual content through the server for privacy and performance reason, it 
needs to be some kind of registration to allow connections at a random end-point 
and to join one and more other to this address.


Probably not a big deal to do the connection and to establish some kind of 
server-client networking, and maybe existing protocols provide all the necessary 
stuff but I wonder how well it scales. Is average customer hardware capable to 
keep track of 1M users at the same time? Or 100M?


Given all that is possible, are there non-technical implications that requires 
us to govern the process?


OpenPGP_signature.asc
Description: OpenPGP digital signature

Re: [TDF Community] [Board Discuss] LibreOffice - peer2peer collaboration bits

[Eyal Rozenberg]

Moving the follow-up here, as Thorsten suggested.

So, I'm arguing that we can side-step the need for "discovery" in the
P2P sense. Or even more fundamentally: P2P has several meanings. There
is "peer-to-peer networking", where nodes build an ad-hoc network in
which they need such peer discovery; and there is "peer to peer" on top
of an existing network in which, as the opposite of "through a central
server". I believe our interest should be in the second kind, as I'll
elaborate below.

On 04/05/2024 13:05, Thorsten Behrens wrote:

How to connect two or more individuals? It requires routing


I would opt for a simple protocol which does not take on problems
more complex than it has to, at least for a preliminary
implementation. Specifically - the Internet routes for us.


Connecting people for peer2peer communication is conventionally called
discovery.

There's ample writeups on the topic, here's two pointers:

* quite an accessible summary: https://jsantell.com/p2p-peer-discovery/
* scientific survey paper: 
https://www.inets.rwth-aachen.de/wp-content/uploads/2022/07/service_discovery_survey.pdf




Anyway - a server would not necessarily be required. That is, P2P
connection will happen when two users want to connect; but in our
case, they have already "connected" in some other way to agree on
making the P2P connection. So I suggest, in light of my previous
point, that we assume that the two (or more) users have another,
independent means of communication over which they can send some
data for bootstrapping the LibreOffice P2P. And this could be made
easy, UI-wise, so that the user just needs to press a copy button,
and paste some string so that the other user can see it. The other
user copies the string and pastes it into an appropriate area in
their own running LO instance. Then the connection is set up.


So in a word, piggy-backing on another, existing communication
channel?


Sort of, but not exactly. Think about a (non-Internet) phone
conversation. It's a "peer-to-peer" protocol, in the sense that there
isn't some "conversation server" which is active while the two
conversants converse: You dial the other person's number and connect to
them. But this is on top of an established network, which routes things
for you, without you having to "discover" anybody or anything. Also -
the phone number is apriori magic, which in ad-hoc peer-to-peer networks
you typically don't have. You need to know the person or organization
you're calling and obtain a numeric destination handle - not through the
use of your phone for the actual phone call. Maybe you've met and
exchanged numbers, maybe you got a card or saw an ad etc.

So, the "piggy-backing" would the interpersonal/social communication
which made the two (or more) people want to collaboratively edit a
document in the first place, plus the regular IP protocol and its
network structure.



What kind of data? Basically, I assume that would be a tuple of (IP,
port number, public key). I will admit that this doesn't cater to
the case of two firewalled users; that's a situation I'm not
experienced enough in handling, but I do know there are [many
approaches](https://en.wikipedia.org/wiki/NAT_traversal) (Wikipedia)
to handling it. Some may require a third-party "switching server",
some may not. But such a server can probably be very minimal and
hopefully not even aware of what protocol it's being used to allow
connections for.


Or taking the idea one step further: re-using the other, existing
comms channel, also for all of the collaboration traffic!


Well, I'm not sure it would be beneficial to create a strong tie-in
between our comm protocol and a specific choice of how the two users
communicate otherwise. But - that is a possibility. It may make things
easier technically I suppose - especially when it comes to more-than-two
collaborators.




Heiko wrote:

What could be achievable on TDF infrastructure?


Given what I've said above - let's try to make this completely
independent of TDF infrastructure. Either with no switching-server
at all or with something minimal that hopefully might not even need
TDF continuously maintaining a server. Note that maintenance by us
also has privacy implications, much more so than third-party-less
P2P.


Yup. At any rate, requiring any kind of centralized server
infrastructure has inevitable scalability challenges. It would still
be useful if TDF could help with bootstrapping whatever server
infrastructure will be needed, though.


Agreed; let's just think of such a bootstrapping as something to avoid
if possible, even at the cost of some tradeoffs. This relates
specifically to my two points of motivation for this feature (which I'll
quote for those who follow this list but not forum): This

"1.Undermines the paradigm Microsoft, Google and others are pushing,
of your work as a user going through them, visible and

Re: [TDF Community] [Board Discuss] LibreOffice - peer2peer collaboration bits

[Thorsten Behrens]

Hi Eyal, all,

this came up on board-discuss [1], but I believe the best place to
have this discussion, is on the LibreOffice developer list. Let's
please follow-up here.

A few comments in-line -

Eyal Rozenberg wrote:
> Heiko wrote:
> > How to connect two or more individuals? It requires routing
> >
> I would opt for a simple protocol which does not take on problems
> more complex than it has to, at least for a preliminary
> implementation. Specifically - the Internet routes for us.
>
Connecting people for peer2peer communication is conventionally called
discovery.

There's ample writeups on the topic, here's two pointers:

* quite an accessible summary: https://jsantell.com/p2p-peer-discovery/
* scientific survey paper: 
https://www.inets.rwth-aachen.de/wp-content/uploads/2022/07/service_discovery_survey.pdf

> Heiko wrote:
> > given that users not want to fiddle around with ports and
> > firewalls neither to share IP addresses I presume this requires a
> > server.
> But then it would not really be P2P, would it?
> 
> Anyway - a server would not necessarily be required. That is, P2P
> connection will happen when two users want to connect; but in our
> case, they have already "connected" in some other way to agree on
> making the P2P connection. So I suggest, in light of my previous
> point, that we assume that the two (or more) users have another,
> independent means of communication over which they can send some
> data for bootstrapping the LibreOffice P2P. And this could be made
> easy, UI-wise, so that the user just needs to press a copy button,
> and paste some string so that the other user can see it. The other
> user copies the string and pastes it into an appropriate area in
> their own running LO instance. Then the connection is set up.
> 
So in a word, piggy-backing on another, existing communication
channel?

> What kind of data? Basically, I assume that would be a tuple of (IP,
> port number, public key). I will admit that this doesn't cater to
> the case of two firewalled users; that's a situation I'm not
> experienced enough in handling, but I do know there are [many
> approaches](https://en.wikipedia.org/wiki/NAT_traversal) (Wikipedia)
> to handling it. Some may require a third-party "switching server",
> some may not. But such a server can probably be very minimal and
> hopefully not even aware of what protocol it's being used to allow
> connections for.
>
Or taking the idea one step further: re-using the other, existing
comms channel, also for all of the collaboration traffic!

> Heiko wrote:
> > What could be achievable on TDF infrastructure?
> >
> Given what I've said above - let's try to make this completely
> independent of TDF infrastructure. Either with no switching-server
> at all or with something minimal that hopefully might not even need
> TDF continuously maintaining a server. Note that maintenance by us
> also has privacy implications, much more so than third-party-less
> P2P.
>
Yup. At any rate, requiring any kind of centralized server
infrastructure has inevitable scalability challenges. It would still
be useful if TDF could help with bootstrapping whatever server
infrastructure will be needed, though.

> Heiko wrote:
> > Isn’t it better to share UNO commands and parameters?
> >
> Mmm... maybe... but - what about showing the other party's cursor
> and mouse movements? You can't do that with UNO commands.
>
Starting off from Collabora Online - which is a production-ready
implementation of LibreOffice collaboration, that uses both low-level
key & mouse events, as well as UNO commands - I guess the answer is
'both'? ;)

> Heiko wrote:
> > How do we solve the situation when one participant enters text and
> > another deletes the same paragraph?
>
> It doesn't have to be a great solution, as long as it is
> consistent. i.e. if users know that two people on a laggy connection
> editing the same sentence is likely to get them making changes in
> wrong positions etc., they will naturally limit the extent to which
> they do this - like we know from Etherpad. Consistency of behavior
> and "principle of least astonishment" would be more important than
> perfect coordination/synchronization of inputs.
>
With a dedicated server, you don't even have that problem. All input
will get serialized through this instance, so there's a strict
temporal ordering for all edits. Whatever package reaches the server
first, will 'win' in an edit war. A fully distributed solution (which
is way harder to implement!) has no such strict global ordering per
se, but there's algorithms such as CRDTs[2], which guarantee eventual
consistency in all peers. But you're right, the Etherpad experience
shows that under bad network connectivity, user experi

Re: About tdf#150082 gssapi

[Julien Nabet]

Thank you Michael and Mike for your feedback!

About sspi, I've submitted https://gerrit.libreoffice.org/c/core/+/167057

(just for the record, I didn't put "sspi_common" because with it, it 
fails to build complaining about missing "sspi_common.c")

Re: About tdf#150082 gssapi

[Mike Kaganski]

On 03.05.2024 14:20, Michael Stahl wrote:

On 03/05/2024 09:27, Julien Nabet wrote:
1) what libs to install on Windows to have one of these files 
available ? (and without installing all Kerberos stuff)


GSSAPI doesn't exist on Windows, but there is SSPI (in the Windows SDK) 
which serves the same purpose.


And mariadb-connector-c seems to include SSPI support - at least, its 
plugins/auth [1] contains sspi_client.c, sspi_errmsg.c, and 
sspi_common.h. Possibly it makes sense to check if adding these .c to 
generated cobjects would work?


[1] 
https://github.com/mariadb-corporation/mariadb-connector-c/tree/b323b5462db3751ed29fc3db83c855154f5f5ab4/plugins/auth


--
Best regards,
Mike Kaganski

Re: About tdf#150082 gssapi

[Michael Stahl]

On 03/05/2024 09:27, Julien Nabet wrote:

Hello,

I pushed a patch concerning tdf#150082 (LO Base MySQL connector don't 
accept auth via gssapi (missing auth_gssapi_client.dll?) but it 
concerned all OSes except Windows.


Indeed, on Windows, when I enable the build of gss in mk files, I got: 
"gssapi.h file missing"

...
1) what libs to install on Windows to have one of these files available 
? (and without installing all Kerberos stuff)


GSSAPI doesn't exist on Windows, but there is SSPI (in the Windows SDK) 
which serves the same purpose.

Re: [Bug 160876] user.documentfoundation.org account generation fails

[Stéphane Guillou]

Hi Donald

On 3/5/24 03:38, Donald Baker wrote:
I did not see any place to report a problem with your web site on your 
web site.  Most places have that.

Maybe not you.
It accepted all my account creation information and then called it 
invalid.  I had to set up a documentfoundation account on a different 
sign-in page.
It is possible that my Safari browser used a forbidden character in 
the generation of a random password.  But if so, your site should have 
shown which special characters were valid.
Professionalism is always appreciated - I really miss Quattro Pro and 
Lotus 1-2-3.


You can ask about account issues on the Website mailing list: 
webs...@global.libreoffice.org


If there is a bug on our infrastructure, you can open a report on 
Redmine: https://redmine.documentfoundation.org/


The first bug I was trying to report involved the persistent refusal 
of Calc to allow me to change the format of numbers on a graph axis. 
 Then I noticed a “Source Format” box checked, which I have not seen 
on other spreadsheet programs.  Unchecking the box fixed the problem. 
 But this kind of over-programmed “feature” is not a feature to me. 
 It’s just an unwelcome impediment.  It might have been cool if Calc 
had automatically used the cell format on the axis.  But it should 
never have refused to allow the use to change it.
You can open an enhancement request to improve the UI: 
https://bugs.documentfoundation.org/


Furthermore, I recall that in the old Quattro Pro one could set up a 
graph and then get the same setup as preferences to succeeding graphs 
in the same spreadsheet.  This saved a lot of annoying busywork.


Similar enhancement requests already exist, the preferred solution 
likely being implementing Chart styles. Please feel free to comment on 
those. See for example:


 * allow setting more defaults than just colours:
   https://bugs.documentfoundation.org/show_bug.cgi?id=62860
 * allow multi-selection of chart objects to change them at once:
   https://bugs.documentfoundation.org/show_bug.cgi?id=62860
 * chart style implementation / support:
 o https://bugs.documentfoundation.org/show_bug.cgi?id=62925
 o https://bugs.documentfoundation.org/show_bug.cgi?id=93597
 * once chart styles are available, spreadsheet themes could make use
   of them. See:
   https://bugs.documentfoundation.org/show_bug.cgi?id=48733#c11
 * we also have a wiki page on the topic of chart styles:
   https://wiki.documentfoundation.org/Chart_styles

I hope that helps.


Donald L. Baker
4203 S 109th E Ave #237
Tulsa  OK  74146

On May 2, 2024, at 11:51 AM, 
bugzilla-dae...@bugs.documentfoundation.org wrote:


Regina Henschel  changed bug 160876 


WhatRemoved Added
CC  rb.hensc...@t-online.de


*Comment # 1 
 on 
bug 160876 
 from 
Regina Henschel  *

That is a problem with the infra structure not with LibreOffice. Bugzilla is
the wrong place to report such problem.

For problems with your account please contact Christian Lohmaier or Guilhem
Moulin, e.g. by
IRChttps://web.libera.chat/?chan=#tdf-infra
or mailing listlibreoff...@lists.freedesktop.org.


You are receiving this mail because:

  * You reported the bug.




--
Stéphane Guillou
Quality Assurance Analyst | The Document Foundation

Email:stephane.guil...@libreoffice.org
Matrix: @stragu:matrix.org
Fediverse: @str...@mastodon.indie.host
Web:https://stragu.gitlab.io/

Re: [Bug 160876] user.documentfoundation.org account generation fails

[Donald Baker]

I did not see any place to report a problem with your web site on your web 
site.  Most places have that.  
Maybe not you.  

It accepted all my account creation information and then called it invalid.  I 
had to set up a documentfoundation account on a different sign-in page.  
It is possible that my Safari browser used a forbidden character in the 
generation of a random password.  But if so, your site should have shown which 
special characters were valid.
Professionalism is always appreciated - I really miss Quattro Pro and Lotus 
1-2-3.  

The first bug I was trying to report involved the persistent refusal of Calc to 
allow me to change the format of numbers on a graph axis.  Then I noticed a 
“Source Format” box checked, which I have not seen on other spreadsheet 
programs.  Unchecking the box fixed the problem.  But this kind of 
over-programmed “feature” is not a feature to me.  It’s just an unwelcome 
impediment.  It might have been cool if Calc had automatically used the cell 
format on the axis.  But it should never have refused to allow the use to 
change it.  

Furthermore, I recall that in the old Quattro Pro one could set up a graph and 
then get the same setup as preferences to succeeding graphs in the same 
spreadsheet.  This saved a lot of annoying busywork.  


Donald L. Baker 
4203 S 109th E Ave #237
Tulsa  OK  74146  

> On May 2, 2024, at 11:51 AM, bugzilla-dae...@bugs.documentfoundation.org 
> wrote:
> 
> Regina Henschel  changed bug 160876 
> 
> What  Removed Added
> CCrb.hensc...@t-online.de 
> 
> Comment # 1  
> on bug 160876  
> from Regina Henschel 
> That is a problem with the infra structure not with LibreOffice. Bugzilla is
> the wrong place to report such problem.
> 
> For problems with your account please contact Christian Lohmaier or Guilhem
> Moulin, e.g. by
> IRC https://web.libera.chat/?chan=#tdf-infra
> or mailing list libreoffice@lists.freedesktop.org 
> .
> 
> You are receiving this mail because:
> You reported the bug.

Re: Uploading external tarballs to ?

[Stephan Bergmann]

On 4/25/24 23:02, Stephan Bergmann wrote:
(Shortly after I had submitted 
 "src/boost_1_85_0.tar.xz", a local test build of  "Upgrade external/boost to latest Boost 1.85.0" scarily failed with


cd /home/sberg/lo0/tar/tmp && echo fetching boost_1_85_0.tar.xz && 
bash -c '/usr/bin/curl -L -O 
https://dev-www.libreoffice.org/src/boost_1_85_0.tar.xz 2>&1 | tee -a 
/home/sberg/lo0/tar/fetch.log && [ $PIPESTATUS -eq 0 ]' && SUM=`shasum 
-a 256 boost_1_85_0.tar.xz | sed "s/ .*//"` && if test "$SUM" != 
"4e23218ff5036d57afd20f7cdab2e94cdbf6ba9c509d656ace643a81c40a985a"; 
then echo ERROR: expected checksum for boost_1_85_0.tar.xz is 
4e23218ff5036d57afd20f7cdab2e94cdbf6ba9c509d656ace643a81c40a985a 2>&1 
| tee -a /home/sberg/lo0/tar/fetch.log; false; fi && mv 
boost_1_85_0.tar.xz ../ fetching boost_1_85_0.tar.xz
  % Total    % Received % Xferd  Average Speed   Time    Time 
Time  Current
 Dload  Upload   Total   Spent
Left  Speed
100   153  100   153    0 0   1131  0 --:--:-- --:--:-- 
--:--:--  1141
ERROR: expected checksum for boost_1_85_0.tar.xz is 
4e23218ff5036d57afd20f7cdab2e94cdbf6ba9c509d656ace643a81c40a985a
make: *** [/home/sberg/lo1/core/Makefile.fetch:102: 
/home/sberg/lo1/core/workdir/download] Error 1


while a second attempt a little later succeeded.)


so  "Fail early when 
external tarball can't be fetched" to make such failure output less 
misleading

Re: Uploading external tarballs to ?

[Stephan Bergmann]

On 4/25/24 03:22, Guilhem Moulin wrote:

See https://git.libreoffice.org/extern#introduction .


"The new file will then shortly be available at 
https://dev-www.libreoffice.org/src/my-new-file.tar.gz once the 
changeset is submitted by a committer."


Is there a reason why there is apparently that temporal gap between 
submitting the change in Gerrit and the file becoming available for 
download?


(Shortly after I had submitted 
 
"src/boost_1_85_0.tar.xz", a local test build of 
 "Upgrade external/boost 
to latest Boost 1.85.0" scarily failed with


cd /home/sberg/lo0/tar/tmp && echo fetching boost_1_85_0.tar.xz && bash -c '/usr/bin/curl -L -O https://dev-www.libreoffice.org/src/boost_1_85_0.tar.xz 2>&1 | tee -a /home/sberg/lo0/tar/fetch.log && [ $PIPESTATUS -eq 0 ]' && SUM=`shasum -a 256 boost_1_85_0.tar.xz | sed "s/ .*//"` && if test "$SUM" != "4e23218ff5036d57afd20f7cdab2e94cdbf6ba9c509d656ace643a81c40a985a"; then echo ERROR: expected checksum for boost_1_85_0.tar.xz is 4e23218ff5036d57afd20f7cdab2e94cdbf6ba9c509d656ace643a81c40a985a 2>&1 | tee -a /home/sberg/lo0/tar/fetch.log; false; fi && mv boost_1_85_0.tar.xz ../ 
fetching boost_1_85_0.tar.xz

  % Total% Received % Xferd  Average Speed   TimeTime Time  Current
 Dload  Upload   Total   SpentLeft  Speed
100   153  100   1530 0   1131  0 --:--:-- --:--:-- --:--:--  1141
ERROR: expected checksum for boost_1_85_0.tar.xz is 
4e23218ff5036d57afd20f7cdab2e94cdbf6ba9c509d656ace643a81c40a985a
make: *** [/home/sberg/lo1/core/Makefile.fetch:102: 
/home/sberg/lo1/core/workdir/download] Error 1


while a second attempt a little later succeeded.)

Re: Minutes from the UX/design meeting 2023-Apr-24

[Stéphane Guillou]

On 25/4/24 09:06, Eyal Rozenberg wrote:

  * Make possible 2 or more impress in fullscreen each on a dedicated
    monitor AND each seekable independently with user-defined hotkeys
    per each file
    + https://bugs.documentfoundation.org/show_bug.cgi?id=160242


First of all, I really think this bug report should have been split up
into several ones - and the issues were sort of mapped on the bug page:

A. run two presentations full-screen at same time (on two monitors)
B. Keyboard-navigate in two running full-screen presentation (i.e.
different shortcuts)
C. Same as B, but arbitrary number of presentations
D. Support per-presentation navigation shortcut choices, persisted in
the ODP file

Suggest breakup instead then INVALID (or MOVED) on this one. I would say

A - NEW
B - UNCONFIRMED (or LATER?)
C, D - WONTFIX


Thanks Eyal! Please go ahead and open a new report for A if you think 
that's the way to go. Let's hold off on B.


Cheers

--
Stéphane Guillou
Quality Assurance Analyst | The Document Foundation

Email: stephane.guil...@libreoffice.org
Matrix: @stragu:matrix.org
Fediverse: @str...@mastodon.indie.host
Web: https://stragu.gitlab.io/

Re: Uploading external tarballs to ?

[Guilhem Moulin]

On Wed, 24 Apr 2024 at 19:24:40 +0200, Guilhem Moulin wrote:
> On Wed, 24 Apr 2024 at 19:19:04 +0200, Stephan Bergmann wrote:
>> Is there a new mechanism in place to do so?  The relevant directory appears
>> to only be root-writable now?
>
> 2024-04-19 17:09:22 <@guilhem> i'm freezing 
> dev-www.libreoffice.org/{src,external}
> on gimli while the data is being relocated.  there will be a change in
> workflow for those who upload there, will send an announce to the dev
> list in due time (hopefully tonight, otherwise during the week-end).
> shout out if something urgently needs to be uploaded in the meantime
>
> Which didn't happen yet due to the regressions caused by the last gerrit
> upgrade.  Writing docs for the new workflow now.

See https://git.libreoffice.org/extern#introduction .

-- 
Guilhem.


signature.asc
Description: PGP signature

Re: Minutes from the UX/design meeting 2023-Apr-24

[Eyal Rozenberg]

Sorry for missing a meeting with a few of "my own" bug reports...


On 24/04/2024 21:58, Heiko Tietze wrote:

  * Want indication of whether a master slide is in use or not
    + https://bugs.documentfoundation.org/show_bug.cgi?id=160403
    + MSO Powerpoints shows the number of slides in a tooltip (Stephane)
    + additional boolean indication would be nice (Eyal)
    + tooltip sounds good and is not harmful (John)
    + not much useful since "Delete Master" is only available if the
  slide master is not in use (Cor)


... but you only see that if you right-click the master. The boolean
indication suggestion would make you aware of this about all visible
masters at once. But I agree with the low priority.


    => no objection, low priority

  * When right-clicking a font family combo-box, offer font meta-data
    + https://bugs.documentfoundation.org/show_bug.cgi?id=152487
    + unclear use case and much likely off-topic being available in
  font management tools (Heiko)


I will have to make the use case clearer on the bug page. Again, sorry
for not being around to make that case.



    + info dialog sounds good but rather via special characters or
  in the character properties dialog (Stuart)
    + only use case is compatibility with some other font (John)


That's an interesting use case, but not what I had in mind.



  + likely not part of the meta data
    + the Internet returns all search results (Cor)


The point is to help the user select an item on a list, while the list
is open - so that the user doesn't have to search for information
elsewhere (e.g. the Internet).

>     => idea does not find support; WF>

  * Make possible 2 or more impress in fullscreen each on a dedicated
    monitor AND each seekable independently with user-defined hotkeys
    per each file
    + https://bugs.documentfoundation.org/show_bug.cgi?id=160242


First of all, I really think this bug report should have been split up
into several ones - and the issues were sort of mapped on the bug page:

A. run two presentations full-screen at same time (on two monitors)
B. Keyboard-navigate in two running full-screen presentation (i.e.
different shortcuts)
C. Same as B, but arbitrary number of presentations
D. Support per-presentation navigation shortcut choices, persisted in
the ODP file

>     + start Impress twice (Stephane)

When you start Impress twice you're still left with just the one
instance running. I guess you can force a second instance (with a
different profile folder perhaps?) but even then, IIANM, Impress is
"imperialistic" when in Full-Screen mode. Op claimed this doesn't work
right now.


workflow is supported with a third   monitor (John)


1. Is it? Can you elaborate on the bug page?
2. What about two monitors?

>     + too niche, better suited for an extension (Stephane, Sahil)

I partially disagree, considering the breakdown above:

A: Not so niche, in my opinion. Occasionally you want to show slides
from different presentations on different monitors - without having
combined them into a multi-monitor-targeting presentation (do we even
have that?). Happened to me already.

B: If Alt-Tab worked (which we can't know, since A is not possible yet),
then this would be kind of niche.

C, D: That's the niche part.


>     => WF

Disagree!

It looks like (some of) the ask changes app behavior rather than
enhancing it. The per-file shortcuts sounds like extension material to
me though.

Suggest breakup instead then INVALID (or MOVED) on this one. I would say

A - NEW
B - UNCONFIRMED (or LATER?)
C, D - WONTFIX

Re: Uploading external tarballs to ?

[Guilhem Moulin]

On Wed, 24 Apr 2024 at 19:19:04 +0200, Stephan Bergmann wrote:
> Is there a new mechanism in place to do so?  The relevant directory appears
> to only be root-writable now?

2024-04-19 17:09:22 <@guilhem> i'm freezing 
dev-www.libreoffice.org/{src,external}
  on gimli while the data is being relocated.  there will be a change in
  workflow for those who upload there, will send an announce to the dev
  list in due time (hopefully tonight, otherwise during the week-end).
  shout out if something urgently needs to be uploaded in the meantime

Which didn't happen yet due to the regressions caused by the last gerrit
upgrade.  Writing docs for the new workflow now.

-- 
Guilhem.


signature.asc
Description: PGP signature

Re: Windows installer errors and warnings

[Eike Rathke]

Hi,

On Tuesday, 2024-04-23 19:11:58 +0200, Jakub Kościelak wrote:

> > > ICE03 ERROR   Invalid identifier; Table: Directory, Column: 
> > > Directory, Key(s): LO_progr_resou_ca@va_LCMES_c8cb7e943
> > > ICE03 ERROR   Invalid identifier; Table: Directory, Column: 
> > > Directory, Key(s): LO_progr_resou_kmr@l_LCMES_c75f6f5fa
> > where ca@va seems to stem from ca-valencia or kmr@l from kmr-Latn.
> 
> I now know where these "@" signs come from. We use this script
> C:\24.2.3.1\solenv\bin\localestr: #!/bin/sh case "$1" in ca-valencia) echo
> ca@valencia ;; kmr-Latn) echo kmr@latin ;; sr-Latn) echo sr@latin ;;
> hu-hung) echo hu@hung ;; *) echo $1 | tr '-' '_' ;; esac in these places:
> C:\24.2.3.1\scp2\macros\macro.pl
> C:\24.2.3.1\solenv\gbuild\AllLangMoTarget.mk
> C:\24.2.3.1\solenv\gbuild\extensions\post_PackageInfo.mk and now out of
> curiosity I have 2 questions: 1. What was/is the reason that for 4 languages
> we replace "-" with "@" instead of underscore? 2. Is it obsolete or is it
> still needed?

The ca@valencia and the like include the glibc @modifiers. This is
correct. Whatever later scrambles those to ca@va_LCMES_c8cb7e943 for
Windows would need to adjust that to something like
ca_va_LCMES_c8cb7e943

  Eike

-- 
GPG key 0x6A6CD5B765632D3A - 2265 D7F3 A7B0 95CC 3918  630B 6A6C D5B7 6563 2D3A


signature.asc
Description: PGP signature

Re: Windows installer errors and warnings

[Julien Nabet]

You're welcome! :-)

Unless I missed it, don't forget license part! (see 
https://wiki.documentfoundation.org/Development/GetInvolved#License_statement).


Julien

Re: Windows installer errors and warnings

[Jakub Kościelak]

Thank you Julien,
I sent patches to gerrit for review,
Jenkins has already commented +1 ;-)

https://gerrit.libreoffice.org/c/core/+/166567

https://gerrit.libreoffice.org/c/core/+/166568


Jakub Kościelak

W dniu 23.04.2024 o 15:42, Julien Nabet pisze:

Hello

I can't tell about your comment but since you seem to have some 
knowledge on this part and you even attached a patch, I propose you to 
follow https://wiki.documentfoundation.org/Development/GetInvolved.


You'll be able to submit the patch directly on the gerrit so it'll be 
easier for review.


Julien

Re: Windows installer errors and warnings

[Jakub Kościelak]

ICE03   ERROR   Invalid identifier; Table: Directory, Column: Directory, 
Key(s): LO_progr_resou_ca@va_LCMES_c8cb7e943
ICE03   ERROR   Invalid identifier; Table: Directory, Column: Directory, 
Key(s): LO_progr_resou_kmr@l_LCMES_c75f6f5fa
[... and hundreds others ...]

These look very much derived from translation resources language tags,
where ca@va seems to stem from ca-valencia or kmr@l from kmr-Latn.
Whyever that is crippled that way..

   Eike



I now know where these "@" signs come from. We use this script 
C:\24.2.3.1\solenv\bin\localestr: #!/bin/sh case "$1" in ca-valencia) 
echo ca@valencia ;; kmr-Latn) echo kmr@latin ;; sr-Latn) echo sr@latin 
;; hu-hung) echo hu@hung ;; *) echo $1 | tr '-' '_' ;; esac in these 
places: C:\24.2.3.1\scp2\macros\macro.pl 
C:\24.2.3.1\solenv\gbuild\AllLangMoTarget.mk 
C:\24.2.3.1\solenv\gbuild\extensions\post_PackageInfo.mk and now out of 
curiosity I have 2 questions: 1. What was/is the reason that for 4 
languages we replace "-" with "@" instead of underscore? 2. Is it 
obsolete or is it still needed?


Jakub Kościelak

Re: Windows installer errors and warnings

[Eike Rathke]

Hi Jakub,

On Tuesday, 2024-04-23 14:14:30 +0200, Jakub Kościelak wrote:

> 1.
> 
> ICE03 ERROR Invalid Language Id;
> 
> Example:
> 
> sd87.mo
> gid_file_res_sd_lang__libreoffice_program_resource_sat_lc_messages_sat sd.mo
> 55167 1606 16384 15468
> 
> We use a Languge Id that is not in the documentation ;-) e.g. 1553 or 1606.

Yes. As MS doesn't support all languages, we define LCIDs in the
user-defined ranges, see include/i18nlangtag/lang.h for documentation.
Here
1553 = 0x0611 = {eo} = EsperantoLANGUAGE_USER_ESPERANTO
1606 = 0x0646 = {sat-IN} = Santali (India)  LANGUAGE_USER_SANTALI_INDIA

> https://learn.microsoft.com/en-us/openspecs/office_standards/ms-oe376/6c085406-a698-4e12-9d4d-c3b0ee3dbc4a

Oh great, yet another place they simultaneously update, where so far
I assumed that finally this evolved to the canonical place:
https://learn.microsoft.com/en-us/openspecs/windows_protocols/ms-lcid/70feba9f-294e-491e-b6eb-56532684c37f

Someone might want to check for differences..


> Here they also wrote that language identifiers are outdated and locale names
> should be used:
> 
> https://learn.microsoft.com/en-us/windows/win32/intl/language-identifier-constants-and-strings
> 
> However, when I used en-US instead of 1033, it reported an error. Maybe
> someone knows how it should be done?

Maybe that's not even applicable to the installer. Or if, then to be
used with some different mechanism than LCIDs.


> ICE03 ERROR   Invalid identifier; Table: Directory, Column: Directory, 
> Key(s): LO_progr_resou_ca@va_LCMES_c8cb7e943
> ICE03 ERROR   Invalid identifier; Table: Directory, Column: Directory, 
> Key(s): LO_progr_resou_kmr@l_LCMES_c75f6f5fa
> [... and hundreds others ...]

These look very much derived from translation resources language tags,
where ca@va seems to stem from ca-valencia or kmr@l from kmr-Latn.
Whyever that is crippled that way..

  Eike

-- 
GPG key 0x6A6CD5B765632D3A - 2265 D7F3 A7B0 95CC 3918  630B 6A6C D5B7 6563 2D3A


signature.asc
Description: PGP signature

Re: Windows installer errors and warnings

[Julien Nabet]

Hello

I can't tell about your comment but since you seem to have some 
knowledge on this part and you even attached a patch, I propose you to 
follow https://wiki.documentfoundation.org/Development/GetInvolved.


You'll be able to submit the patch directly on the gerrit so it'll be 
easier for review.


Julien

Re: Build failed for ooxmlexport21

[Kira Tubo]

Thanks for looking into this for me, Tomaž . Appreciate the advice!

On Mon, Apr 22, 2024 at 6:52 PM Tomaž Vajngerl  wrote:

> Hi,
>
> On Tue, Apr 23, 2024 at 5:28 AM Kira Tubo  wrote:
> >
> > Hello,
> >
> > My build seems to be failing in Gerrit:
> https://gerrit.libreoffice.org/c/core/+/166423/1..2
> >
> > Errors suggest that there is a core file missing for ooxmlexport21 in
> Linux. I already rebased, but still see the error. Please advise.
> >
> > https://ci.libreoffice.org/job/gerrit_master_ml/16581/
>
> No, you have to scroll up and look at the stack.. A crash happened in:
> #11
> /home/tdf/lode/jenkins/workspace/lo_gerrit/Config/linux_gcc_release_64/instdir/program/libmergedlo.so(_ZNK12OutputDevice11ImplNewFontEv+0x106)
> [0x7f9c57e14ce6]
> #12
> /home/tdf/lode/jenkins/workspace/lo_gerrit/Config/linux_gcc_release_64/instdir/program/libmergedlo.so(_ZNK12OutputDevice13GetFontMetricEv+0x34)
> [0x7f9c57e17034]
>
> Or if you look before stack dump you see: "terminating test due to
> missing font: Rockwell" - which is indeed the font used in the
> document.
>
> I think you have just taken the document from the bug document. I
> strongly recommend not to do this. You know what specifically is
> wrong, so the best is to reproduce the bug in a clean document and you
> will end up with a test document that is smaller, is focused to show
> precisely what the bug is without anything unneeded and will avoid
> issues like this.
>
> Also I recommend to name the bug with what is tested in addition to
> the bug number or put the bug number in the comments inside the test.
>
> > Thanks in advance!
> >
> > Regards,
> > Kira
>
> Regards, Tomaž
>

Re: Build failed for ooxmlexport21

[Tomaž Vajngerl]

Hi,

On Tue, Apr 23, 2024 at 5:28 AM Kira Tubo  wrote:
>
> Hello,
>
> My build seems to be failing in Gerrit: 
> https://gerrit.libreoffice.org/c/core/+/166423/1..2
>
> Errors suggest that there is a core file missing for ooxmlexport21 in Linux. 
> I already rebased, but still see the error. Please advise.
>
> https://ci.libreoffice.org/job/gerrit_master_ml/16581/

No, you have to scroll up and look at the stack.. A crash happened in:
#11 
/home/tdf/lode/jenkins/workspace/lo_gerrit/Config/linux_gcc_release_64/instdir/program/libmergedlo.so(_ZNK12OutputDevice11ImplNewFontEv+0x106)
[0x7f9c57e14ce6]
#12 
/home/tdf/lode/jenkins/workspace/lo_gerrit/Config/linux_gcc_release_64/instdir/program/libmergedlo.so(_ZNK12OutputDevice13GetFontMetricEv+0x34)
[0x7f9c57e17034]

Or if you look before stack dump you see: "terminating test due to
missing font: Rockwell" - which is indeed the font used in the
document.

I think you have just taken the document from the bug document. I
strongly recommend not to do this. You know what specifically is
wrong, so the best is to reproduce the bug in a clean document and you
will end up with a test document that is smaller, is focused to show
precisely what the bug is without anything unneeded and will avoid
issues like this.

Also I recommend to name the bug with what is tested in addition to
the bug number or put the bug number in the comments inside the test.

> Thanks in advance!
>
> Regards,
> Kira

Regards, Tomaž

Re: Is there an Impress Remote for PC?

[Miklos Vajna]

Hi,

On Thu, Apr 18, 2024 at 08:27:44PM +0300, Владислав Тараканов 
 wrote:
> Yes, we want to try to implement this on our own. For now we plan to
> base it on the implementation in
> https://git.libreoffice.org/impress_remote. If I understand correctly,
> we can first look at this part:
> android/sdremote/mobile/src/main/java/org/libreoffice/impressremote/communication.
> At first glance, this package implements data exchange with Impress
> over the network.

Yes, that's the client code. The other side is mostly under
sd/source/ui/remotecontrol/ in core.git, I believe.

Regards,

Miklos

[Bug 91282] Fullscreen toolbar disappears after docking it, leaving fullscreen mode and re-entering

[bugzilla-daemon]

https://bugs.documentfoundation.org/show_bug.cgi?id=91282

--- Comment #20 from QA Administrators  ---
Dear p10,

To make sure we're focusing on the bugs that affect our users today,
LibreOffice QA is asking bug reporters and confirmers to retest open, confirmed
bugs which have not been touched for over a year.

There have been thousands of bug fixes and commits since anyone checked on this
bug report. During that time, it's possible that the bug has been fixed, or the
details of the problem have changed. We'd really appreciate your help in
getting confirmation that the bug is still present.

If you have time, please do the following:

Test to see if the bug is still present with the latest version of LibreOffice
from https://www.libreoffice.org/download/

If the bug is present, please leave a comment that includes the information
from Help - About LibreOffice.

If the bug is NOT present, please set the bug's Status field to
RESOLVED-WORKSFORME and leave a comment that includes the information from Help
- About LibreOffice.

Please DO NOT

Update the version field
Reply via email (please reply directly on the bug tracker)
Set the bug's Status field to RESOLVED - FIXED (this status has a particular
meaning that is not 
appropriate in this case)


If you want to do more to help you can test to see if your issue is a
REGRESSION. To do so:
1. Download and install oldest version of LibreOffice (usually 3.3 unless your
bug pertains to a feature added after 3.3) from
https://downloadarchive.documentfoundation.org/libreoffice/old/

2. Test your bug
3. Leave a comment with your results.
4a. If the bug was present with 3.3 - set version to 'inherited from OOo';
4b. If the bug was not present in 3.3 - add 'regression' to keyword


Feel free to come ask questions or to say hello in our QA chat:
https://web.libera.chat/?settings=#libreoffice-qa

Thank you for helping us make LibreOffice even better for everyone!

Warm Regards,
QA Team

MassPing-UntouchedBug

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

Re: Is there an Impress Remote for PC?

[Владислав Тараканов]

Yes, we want to try to implement this on our own. For now we plan to base it on 
the implementation in https://git.libreoffice.org/impress_remote. If I 
understand correctly, we can first look at this part: 
android/sdremote/mobile/src/main/java/org/libreoffice/impressremote/communication.
 At first glance, this package implements data exchange with Impress over the 
network.

  
>Четверг, 18 апреля 2024, 10:49 +04:00 от Miklos Vajna :
> 
>Hi,
>
>On Wed, Apr 17, 2024 at 12:12:52PM +0300, Амин Иргалиев < irgalie...@mail.ru > 
>wrote:
>> Some users report that they would like the LO Impress Remote functionality 
>> for computers to make it comfortable to hold conferences. Provide remote 
>> access to the device doesn’t solve this problem, since it is not very safe 
>> to give access to the device.
>>
>> No one knows if it is possible to achieve similar functionality with 
>> existing LO features? If not, is there a chance that such an application 
>> will be in demand enough that it can be made not as a local solution, but as 
>> a whole for LO?
>The Impress remote is currently for Android/iOS, but I don't think
>anything would stop you from writing a desktop app that uses the same
>protocol.
>
>If you are interested in hacking on this yourself? Do you need some code
>pointers for the protocol?
>
>Regards,
>
>Miklos 
 
 
--
Vladislav Tarakanov