The Document Foundation Planet (en)

Friday
14 November, 2025

Official TDF Blog: The role of ODF in digital sovereignty (digital freedom)

@ 12:25 UTC

Digital sovereignty, or the ability of nations, organisations and individuals to control their own digital destiny, is a fundamental issue of the 21st century. At the heart of this challenge lies a seemingly trivial question: who controls the format of the documents that contain our intellectual property or personal information?

In this context, the standard and open Open Document Format (ODF) – the native format of LibreOffice documents, also supported by other suites – is the fundamental technology for those seeking true digital independence.

Digital sovereignty includes the ability to control access to one’s own information without depending on third parties, to make independent technological choices based on one’s own needs, to ensure independent access to strategic data without depending on the commercial interests of Big Tech, and to maintain this technological self-determination in the face of market consolidation.

When government agencies, businesses, or citizens store their documents in proprietary formats controlled by Big Tech, they surrender part of their sovereignty and depend on these external entities to access their own information.

Why document formats are important for sovereignty

Document formats are infrastructure, which—like roads, power grids, or telecommunications networks—are fundamental to the functioning of modern societies. Consider what happens when strategic documents exist only in formats controlled by a single vendor:

Vendor Lock-In: Organisations find themselves trapped, unable to switch to alternative software without costly conversion processes and potential data loss.
Loss of Control: formats can and do change without notice and beyond the control of users, increasing the effect of vendor lock-in.
Fragility of Access: if the vendor controlling the format changes the format or discontinues support, as was the case with Windows 10, access to documents becomes problematic or impossible.
Economic Dependency: The cost of the licence required to perform software updates creates a relationship of economic dependency in order to access one’s own data, and effectively transfers ownership of the data to the vendor that controls the format.

Why ODF is the only tool for digital sovereignty

ODF is governed by OASIS, an international standardisation organisation that protects its transparent development, and is published as ISO/IEC 26300-2015 (and soon ISO/IEC 26300-2025). Unlike proprietary formats, ODF specifications are public and can be freely implemented, are developed through a transparent, multi-stakeholder process, are not controlled by a single government or company, and are subject to international standardisation bodies.

This means that governments and companies can participate in defining the format specifications, rather than being forced to passively accept changes imposed by a single vendor based on its commercial strategies.

Thus, ODF specifications allow anyone to create an office suite that natively supports the format and promotes digital sovereignty, without any authorisation, licence fees or fear of legal action, while supporting the local software industry.

ODF enables true interoperability, not only between different software packages, but also between countries, languages and political systems. A document created in Brazil can be opened and edited in India, Germany or Japan using locally developed software. This breaks …

Thursday
13 November, 2025

Official TDF Blog: Announcement of LibreOffice 25.8.3

@ 13:21 UTC

Berlin, 13 November 2025 – LibreOffice 25.8.3, the third minor release of the free, volunteer-supported office suite for personal productivity in office environments, for Windows, MacOS and Linux, is now available at www.libreoffice.org/download. The new version fixes 70 issues compared to the previous release, which came out in October [1].

LibreOffice 25.8.3 is based on the LibreOffice Technology, which enables the development of desktop, mobile and cloud versions – either from TDF or from the ecosystem – that fully supports the two document format standards: the open ODF or Open Document Format (ODT, ODS and ODP), and the closed and proprietary Microsoft OOXML (DOCX, XLSX and PPTX). Products based on the LibreOffice Technology are available for all major desktop operating systems (Windows, macOS, Linux and ChromeOS), mobile platforms (Android and iOS) and the cloud.

For enterprise-class deployments, TDF recommends the LibreOffice Enterprise optimized versions from ecosystem companies, with dedicated value-added features and other benefits such as SLAs and security patch backports for three to five years (www.libreoffice.org/download/libreoffice-in-business/).

English manuals for the LibreOffice 25.8 family are available for download at https://books.libreoffice.org/en/. End users can get first-level technical support from volunteers on user mailing lists and Ask LibreOffice website: ask.libreoffice.org.

Downloading LibreOffice

All available versions of LibreOffice for the desktop can be downloaded from the same website: www.libreoffice.org/download/. To improve the interoperability with Microsoft 365, we suggest installing the Microsoft Aptos font from this web page: learn.microsoft.com/en-us/typography/font-list/aptos.

LibreOffice enterprise and individual users can support The Document Foundation and the LibreOffice project by making a donation: www.libreoffice.org/donate.

[1] Fixes in RC1: wiki.documentfoundation.org/Releases/25.8.3/RC1. Fixes in RC2: wiki.documentfoundation.org/Releases/25.8.3/RC2.

Thursday
06 November, 2025

LibreOffice QA Blog: QA/Dev Report: October 2025

@ 12:00 UTC

General Activities

LibreOffice 25.8.2 was announced on October 9
LibreOffice 25.2.7 was announced on October 30
Olivier Hallot (TDF) added help pages for R1C1 Calc formula syntax and DOI citation recognition and improved and updated help on dimension lines, form properties, master documents, command line operations, online update, text boundaries and VBA constants. He also adapted the helper script used for patch submission to a version that works with Help
Gábor Kelemen (Collabora) improved the script for finding unneeded includes in the code and did many code cleanups
Tomaž Vajngerl (Collabora) continued working on sheet view functionality in Calc
Pranam Lashkari, Dennis Francis, Szymon Kłos, Jaume Pujantell and Gülşah Köse (Collabora) worked on LOKit/jsdialog used by Collabora Online
Rashesh Padia (Collabora) made the revamped Impress transition list more robust
Michael Meeks (Collabora) did code cleanups and optimisations in PPTX export code
Miklós Vajna (Collabora) improved image handling in Markdown import and export, continued improving the handling of tracked changes that depend on each other and fixed issues with handling of bulleted lists in PPTX files
Xisco Faulí (TDF) added sqlite3, dbm and pythonw.exe to the internal Python, fixed an Impress printing crash, added a few new automated tests and upgraded many dependencies
Michael Stahl (Collabora) implemented per-line paragraph properties for Word Compatibility Mode, fixed a PDF export issue involving variable fields and hidden text and fixed automatic captioning of images in Writer in the case of a single pasted image
Mike Kaganski (Collabora) improved help for inserting page numbers and accessing remote files, made Google Drive two-factor authentication work on Windows, made date and time arithmetic more robust, made the loading of macro class modules happen in the correct order, preventing name conflicts, made it so the automatic updater does not run in headless mode, made the display of empty hidden paragraphs between tables match that of Microsoft Word, improved text property handling in PPTX files, reduced console noise in debug builds by marking menu items without icons, fixed an issue with Writer column separator colour sometimes not being saved, made shape identification in ODF export more robust and improved the behaviour of conditional hiding of sections. He also did many code cleanups and optimisations
Caolán McNamara (Collabora) added Adobe Font Development Kit for OpenType (AFDKO) library for converting and merging Type 1 fonts to OTF when importing PDFs via pdfium and fixed an (unreleased) issue with scrolling the Calc formula input box. He also fixed crashes and many issues found by static analysers and did code cleanups and optimisations
Stephan Bergmann (Collabora) did build fixes, adapted the code to compiler changes and did code cleanups
Noel Grandin (Collabora) made it faster to reject tracked changes in Calc, export EPUB files and render SVGs with pattern fills. He also did many code cleanups and optimisations, especially in the area of transparency handling
Justin Luth (Collabora) made it so undoing an autocorrection triggered by a newline does not …

Tuesday
04 November, 2025

Miklos Vajna: Interdependent tracked changes improvements in Writer, part 4: direct accept/reject

@ 07:11 UTC

Writer has some support for interdependent (or hierarchical) tracked changes: e.g. the case when you have a delete on top of an insert. See the third post for background.

This work is primarily for Collabora Online, but the feature is available in desktop Writer as well.

Motivation¶

Interdependent changes mean that the UI shows one type of change on top of another change, e.g. formatting on top of insert. Writer knows the priority of each type, so in case you have an insert or delete change and on top of that you have a formatting, then the UI will look "through" the formatting and work on the underlying insert or delete when you navigate with your cursor to a position with multiple changes and you press Accept on the Review tab of the notebookbar.

Usually this is what you mean, but what if you want to work on the formatting at the top, directly? You can now open the Manage Changes dialog using the Manage button on the Review tab of the notebookbar and if you go to the formatting change row of the dialog, then pressing Accept there will accept the formatting change, not the insert or delete change. This is possible, because the dialog gives you a way to precisely select which tracked change you want to work with, even if a specific cursor position has multiple tracked changes.

Results so far¶

Here is a sample ins-then-format.docx document from the core.git testcases, the baseline has an insertion, and part of that is covered by an additional formatting change on top:

If you just go in the middle of the document and press Accept, that will work with the more important insert change, so the result looks like this:

But now you can also open the Manage Changes dialog, to be more specific by directly selecting the formatting change:

And when you accept the formatting change directly, the result will be just the insert change:

You can save and load the results in both DOCX and ODT, as usual.

How is this implemented?¶

If you would like to know a bit more about how this works, continue reading... :-)

As usual, the high-level problem was addressed by a series of small changes. Core side:

Want to start using this?¶

You can get a development edition of Collabora Online 25.04 and try it out yourself right now: try the development edition. Collabora intends to continue supporting and contributing to LibreOffice, the code is merged so we expect all of this work will be available in TDF's next release too (26.2).

Wednesday
22 October, 2025

LibreOffice Dev Blog: enumarray for better data arrays – EasyHack

@ 19:30 UTC

In LibreOffice C++ code, there are many cases where developers have string literals that should be used in the code. If these are messages in the graphical user interface (GUI), they should be added to the translatable messages. But, there are many cases where the string literals has nothing to do with other languages, and they should be used as they are. In this case, enumarray is helpful. Although they are not limited to string literals, and can be used for any data.

Using Symbolic Constants

In old C code, using #define was the preferred way one could give a name to a string literal or other kinds of data. For example, consider this code:

const char[] FRAME_PROPNAME_ASCII_DISPATCHRECORDERSUPPLIER = "DispatchRecorderSupplier";
const char[] FRAME_PROPNAME_ASCII_ISHIDDEN = "IsHidden";
inline constexpr OUString FRAME_PROPNAME_ASCII_LAYOUTMANAGER = "LayoutManager";
const char[] FRAME_PROPNAME_ASCII_TITLE = "Title"_ustr;
const char[] FRAME_PROPNAME_ASCII_INDICATORINTERCEPTION = "IndicatorInterception";
const char[] FRAME_PROPNAME_ASCII_URL = "URL";

And also, the relevant states:

#define FRAME_PROPHANDLE_DISPATCHRECORDERSUPPLIER 0
#define FRAME_PROPHANDLE_ISHIDDEN 1
#define FRAME_PROPHANDLE_LAYOUTMANAGER 2
#define FRAME_PROPHANDLE_TITLE 3
#define FRAME_PROPHANDLE_INDICATORINTERCEPTION 4
#define FRAME_PROPHANDLE_URL 5

Although this C code still works in C++, it is not the desired approach in modern C++.

Using enumarrays

In modern C++ code, you can use enumarry, which is defined in o3tl in LibreOffice. The above code becomes:

enum class FramePropNameASCII
{
    DispatcherRecorderSupplier,
    IsHidden,
    LayoutManager,
    Title,
    IndicatorInterception,
    URL,
    LAST=URL
};

And also, the string literal definitions:

constexpr o3tl::enumarray<FramePropNameASCII, OUString> FramePropName = {
    u"DispatchRecorderSupplier"_ustr,
    u"IsHidden"_ustr,
    u"LayoutManager"_ustr,
    u"Title"_ustr,
    u"IndicatorInterception"_ustr,
    u"URL"_ustr
};

Why an enumarray?

The names are much more readable, as they do not have to be ALL_CAPPS, as per convention for symbolic constants in C. Their usage is also quite easy. For example, one can use [] to access the relevant string literal:

- xPropSet->getPropertyValue( FRAME_PROPNAME_ASCII_LAYOUTMANAGER ) >>= xLayoutManager;
+ xPropSet->getPropertyValue( FramePropName[FramePropNameASCII::LayoutManager] ) >>= xLayoutManager;

Final Notes

In LibreOffice, enumarrays are not limited to string literals, and they can be used with other data. This task is filed as tdf#169155, and if you like, you may try finding some instances in the code and modernize it using enumarrays.

To learn more about LibreOffice development, you can refer to TDF Wiki. You may follow this blog to read about EasyHacks, tutorials and announcements related to LibreOffice development.

Tuesday
21 October, 2025

LibreOffice QA Blog: QA/Dev Report: September 2025

@ 11:20 UTC

General Activities

LibreOffice 25.2.6 was announced on September 8
Olivier Hallot (TDF) improved the help for Select Function in Calc’s formula bar, expanded help for the selection of chart data sources, added AutoFilter and Pivot Table/Chart to the help page on sheet protection, added information about summary above/below to the help of Calc’s SUBTOTAL() function, documented sensitivity in the help page for Solver and added the description meta element to the Help page templates
Gábor Kelemen (Collabora) improved the script for finding unneeded includes in the code and did many code cleanups
Tomaž Vajngerl (Collabora) implemented a sheet view functionality in Calc allowing to manipulate AutoFilters in one view without affecting other views
Pranam Lashkari, Maya Stephens, Gökay Şatır, Rashesh Padia and Mohit Marathe (Collabora) worked on LOKit used by Collabora Online. Maya also fixed an issue with Ctrl+clicking hyperlinked objects
Miklós Vajna (Collabora) expanded Writer Markdown export support, worked on change tracking of formatting, continued improving the handling of tracked changes that depend on each other and fixed bullet list style going missing with PPTX export
Xisco Faulí (TDF) implemented right-to-left text direction and bidirectional text support in SVG import, added venv to the internal Python, fixed crashes, added several new automated tests, upgraded many dependencies and did many code cleanups and optimisations
Michael Stahl (Collabora) improved the DOCX compatibility of floating tables, helped Miklós with interdependent change tracking in Writer and fixed an issue with document compression
Mike Kaganski (Collabora) fixed Help for BASIC’s Str function, adapted Writer’s change tracking to DOCX’s inability to track the deletion of comments, made date handling code more robust, improved the Visual Studio IDE integration, made the handling of double variables in BASIC more robust, replaced dtoa library with fast_float for faster string-to-float parsing and fixed an RTF import issue with some paragraphs appearing outside of their frames. He also did many code cleanups and optimisations
Caolán McNamara (Collabora) fixed many issues found by static analysers and did code cleanups and optimisations
Stephan Bergmann (Collabora) worked on the WASM build. He also adapted the code to compiler changes and did code cleanups
Noel Grandin (Collabora) fixed an (unreleased) issue with image fills disappearing at lower zoom levels, improved the performance of spreadsheets with thousands of shapes, improved the performance of 3D charts, fixed a performance issue with importing SVGs with small scaling, improved the performance of Calc’s Remove Duplicates, implemented Skia native rendering of bitmap tiling to improve the performance of filled polygons and fixed an issue with Insert Special Character dialog not being available with mergedlibs builds (Windows releases). He also did many code cleanups and optimisations, especially in the area of transparency handling
Justin Luth (Collabora) fixed an issue with the alignment setting in Page Number Wizard, fixed localised accelerator shortcut conflicts, fixed many issues with page styles, page breaks and spell checking, improved the Bullets and Numbering dialog by showing the currently …

Thursday
16 October, 2025

LibreOffice Dev Blog: enum class instead of unscoped enum – EasyHack

@ 14:05 UTC

Since C++11 when enum class (also named scoped enum) is introduced, it is preferred to plain enum which is inherited from C programming languages. The task here is to convert the old enum instances to enum class.

Rationale

enum class has many benefits when compared to plain enum, as it provides better type safety among other things. Implicit conversion to integers, lack of ability to define the underlying data type and compatibility issues were some of the problems with plain enum that enum class solved in C++11. Although since then enum has improved and one can specify underlying type in the scoped enumerations.

Plain enums pollute namespace, and you have to pick names that are too long, and have to carry the context inside their names. For example: INETMSG_RFC822_BEGIN inside enum _ImplINetRFC822MessageHeaderState. With an enum class, it is simply written as HeaderState::BEGIN. When placed inside a file/class/namespace that makes it relevant, it is much easier to use: it is more readable, and causes no issues for other identifiers with possible similar names.,

See this change:

593f08303 convert enum _ImplINetRFC822MessageHeaderState to enum class

You can read more about that in:

Why is enum class considered safer to use than plain enum?

Finding Instances

You may find some of the instances with:

$ git grep -w enum *.cxx *.hxx|grep -v "enum class"

When you count it with wc -l, it shows something more than 2k instances.

Examples Commits

You can see some of the previous conversions here, which is around 1k changes:

$ git log --oneline -i -E --grep="convert enum|scoped enum"

This is a good, but lengthy example of such a conversion:

9072c5c855 convert SbxFlagsBits to scoped enum

Implementation

First of all, please choose good names for the new enum class and values. For example, you may convert APPLICATION_WINDOW_TITLE into Application::WindowTitle. Therefore, do not use the old names as they were.

Converting enum to enum class is not always straightforward. You should try to understand the code using the enum, and then try to replace it with enum class. You may need to add extra state/values for situations where 0 or -1 or some default value was used. There are cases where a numerical value is used for different conflicting purposes, and then you have to do some sort of conflict resolution to separate those cases.

You may end up modifying more and more files, and a few static_casts where they are absolutely necessary because you are interpreting some integer value read from input. These are the places where you should check the values yourself in the code. You have to make sure that the numerical value is appropriate before casting it to the enum class.

If you want to do bitwise operations, you should use o3tl::typed_flags, for example:

enum class FileViewFlags
{
    None = 0x00,
    MultiSelection = 0x02,
    ShowType = 0x04,
    ShowNone = 0x20,
};

template<> struct o3tl::typed_flags : o3tl::is_typed_flags<FileViewFlags, 0x26> {}

Then, you may use it like this:

    if (nFlags & FileViewFlags::MULTISELECTION)
        mxTreeView->set_selection_mode(SelectionMode::Multiple …

Tuesday
07 October, 2025

Miklos Vajna: Markdown import/export in Writer

@ 06:13 UTC

Writer recently got a Markdown import & export filter and there were a number of improvements to that.

This work is primarily for Collabora Online, but the feature is available in desktop Writer as well.

Motivation¶

Ujjawal Kumar contributed a markdown import to Writer, as part of Google Summer of Code (GSoC) this summer. Mike Kaganski of Collabora also created a minimal markdown export in Writer. I looked at the feature differences between the two, and filled in various gaps in the markdown export. I also added a few general markdown import/export improvements relevant for normal Writer documents, like embedded image support.

Results so far¶

Here is a sample case of a document using inline code spans:

Exporting this to markdown & loading back to Writer, the code span was lost:

And now it's preserved:

This also works with code blocks.

Second, here is a document with lists:

Exporting this to markdown & loading back to Writer, the lists were lost:

And now they are preserved:

This also works with nested lists.

Third, here is a document with an image:

Exporting this to markdown & loading back to Writer, the image was lost:

And now it's preserved:

This also works with embedded and anchored images.

Fourth, here is a document with a table:

Exporting this to markdown & loading back to Writer, the table was lost:

And now it's preserved:

This also works with table alignments and nested tables (to the extent the markdown markup allows that).

Fifth, here is a document with a quote block:

Exporting this to markdown & loading back to Writer, the quote's paragraph indentation was lost:

And now it's preserved:

How is this implemented?¶

If you would like to know a bit more about how this works, continue reading... :-)

As usual, the high-level problem was addressed by a series of small changes. Core side:

Want to start …

Thursday
18 September, 2025

LibreOffice Dev Blog: Debugging tips for LibreOffice

@ 14:11 UTC

If you are working with LibreOffice code, trying to understand the code, fix bugs, or implement new features, you will need to debug the code at some point. Here are some general tips for a good debugging experience. Let’s start from the platform

Choose the Right Debug Platform

Choosing a platform to debug usually depends on the nature of the problem. If the problem is Windows-only, you need a Windows environment to build and debug the problem. But, if the problems can be reproduced everywhere, then you can choose the platform of your choice with the debugging tools that you prefer to debug the problem.

On Linux, it matters if you are running X11 or Wayland. Also, as there are multiple graphical back-ends available for LibreOffice, it matters if you are using X11, GTK3/4, or Qt5/6 back-end for your debugging. Some bugs are specific to GTK, then you should use GTK3 UI for testing. In 2025, GTK4 UI of LibreOffice is still experimental, so it is better to work with GTK3. For making the debugging easier, many developers work on X11 (gen) UI for debugging.

Debugging Tools

Various debugging tools can be used to debug the soffice.bin/soffice.exe LibreOffice binary that you have built. For the common debuggers, you can use GDB on Linux, lldb on macOS, and WinDbg or Visual Studio on Windows.

For using the above debuggers, you can use the IDE or front-end that support them. Various IDEs are usable with LibreOffice code. For a detailed explanation, refer to this Wiki article:

TDF Wiki: Development/IDE

Make sure that you can build and debug a simple program before trying to build and debug LibreOffice.

Environment Variables

To have a better debugging experience, or to avoid problems you may have to customize the debugging session with environment variables. A complete article of the TDF Wiki is dedicated to discuss the environment variables that can be used with LibreOffice:

TDF Wiki: Development/Environment variables

Here is some of the most important ones:

1) Using the X11 user interface:

If you want to use the X11 back-end that is simpler, and usually easier to work with on debug sessions, you have to set SAL_USE_VCLPLUGIN environment variable:

export SAL_USE_VCLPLUGIN=gen
That is specially useful when you are debugging graphical problems. But in some cases, you may need to avoid it or at least customize it. For example, while debugging mouse-related problems you may need to tell LibreOffice to avoid mouse grabbing this way:

export SAL_NO_MOUSEGRABS=1

2) Using GTK user interface

If you are using GTK user interface, then you may use GTK inspector to interactively debug LibreOffice GUI. You can use it this way:

export GTK_DEBUG=interactive

Pretty Printers

In solenv/gdb/ inside LibreOffice source code, you may find pretty printers for GDB. This is helpful when debugging LibreOffice with GDB, to be able to see data in a more readable way.

Dumping Data

Sometimes when you debug a LibreOffice application, it is easier to …

Tuesday
16 September, 2025

LibreOffice QA Blog: QA/Dev Report: August 2025

@ 09:42 UTC

General Activities

LibreOffice 25.8.0 and LibreOffice 25.8.1 were announced on August 20 and August 29 respectively
Olivier Hallot (TDF) updated help for the option to load printer settings with document, sorting blocks of cells in Calc, hyphenation, statistical functions, number of lines in charts, exponentiation operator in Calc, remote files, Edit menu in Calc, object rotation, Math options and MATCH function in Calc
Celia Palacios added help for the new Intersect() method in ScriptForge
Gábor Kelemen (Collabora) did many code cleanups
Tomaž Vajngerl (Collabora) did many code cleanups and added OOXML test documents for text fitting / scaling
Pranam Lashkari and Marco Cecchetti (Collabora) worked on LOKit used by Collabora Online. Marco also made it so hovering with the mouse over Chart data range colour palette entries in the Sidebar shows a live preview in the active chart
Miklós Vajna (Collabora) added list and inline code block support for Markdown export and continued improving the handling of tracked changes that depend on each other
Xisco Faulí (TDF) fixed crashes, added over a dozen new automated tests, upgraded many dependencies and did many code cleanups and optimisations
Michael Stahl (Collabora) made it so pasted anchored objects are no longer selected by default while adding an expert configuration option for the behaviour, added overline support to XHTML export and worked around a dbus bug affecting the build process on some Linux systems
Mike Kaganski (Collabora) fixed an issue with embedded fonts getting dropped from opened files in certain scenarios on Windows, made it so the user can choose to either discard license-restricted embedded fonts in an opened document or switch to read-only mode, improved PPTX compatibility with trailing empty lines in automatically shrinking text boxes, fixed long links getting truncated when exporting to XLSX, fixed issues with inserting hyperlinks in Calc via the API, made Calc text insertion API methods more robust, fixed inserting PDFs into spreadsheets, fixed a string handling issue in Basic’s Format function, fixed a VBA macro issue with dates and fixed processing of escaped backslashes in RTF files. He also did many code cleanups and optimisations
Caolán McNamara (Collabora) fixed many issues found by static analysers and did code cleanups and optimisations
Stephan Bergmann (Collabora) worked on the WASM build. He also adapted the code to compiler changes and did code cleanups
Noel Grandin (Collabora) improved the scrolling speed in Writer documents with lots of comments. He also did many code cleanups and optimisations, especially in the area of transparency handling
Justin Luth (Collabora) improved DOCX compatibility with margins of aligned floating objects, fixed right/left only page breaks going missing with DOC/DOCX export, fixed a DOCX indentation issue, fixed column breaks going missing in certain DOCX files, fixed an issue with numbered lists created by AutoCorrect, made it so justified text with section breaks in saved DOC files no longer triggers an MS Word bug and fixed numbering or bullets getting lost when applying a paragraph style
Michael Weghorn …

Monday
15 September, 2025

Mike Kaganski: A fairy tale about poor UX enforcing vendor lock-in

@ 08:25 UTC

Once upon a time, there was a girl, who used WhatsApp in her iPhone. She was rather active there, and collected quite some important data in the app over time. But some things in her iPhone were inconvenient; and the phone was slowly aging. So she wanted to change her phone some day.

For her birthday, a fairy, who learned somehow about the girl’s wish, presented her a new Android phone. That was a nice new phone, and the girl was so happy! She decided to move everything from the old phone to the new one immediately.

She was worrying about how to move the precious data between the devices; but she felt a huge relief, when the phone spoke: “The fairy told me how important your data is to you; and I have magic powers to handle it all. Just connect the old phone to me with a cord”. So she did.

The new phone started its work; and the girl could see how the progress bar was gradually moving to completion; but suddenly it stopped; minutes passed, but the bar was motionless. The girl was impatient to start using her new shiny device, but she knew that she needs to wait. And she waited; and waited; but after an hour passed, she noticed something horrible: the old phone was sucking the life out of the new device through the cable!

The scared girl could only hope that the process would resume, and finish before the new phone is out of power. She searched and learned, that iPhones are known for their insatiable hunger, and whenever they are connected to anything with energy, they start sucking it. She couldn’t even ask the new phone to shine less brightly to save the energy – because it wasn’t ready for such things yet. She used her wireless charger, but its powers were fewer than the hunger of iPhone, combined with the hard work done by Android. The energy level still decreased too fast.

In the end, when the hope almost vanished, the progress resumed moving! But immediately, the new phone said: “When I collected your data from your old phone, something bad happened, and I failed to collect something. I will continue, but please check later, what’s missing!”.

Only a couple of energy drops were remaining in the new phone, when it finished its task, and could be disconnected from the vampire. But the girl was terrified, when she opened WhatsApp, connected to it (using a magic SMS confirmation), only to see that all her data is lost! She tried to open WhatsApp on the old phone to check if something is still there, and saw that the app had disconnected her. So she used the SMS magic again, and – to her great relief – everything was there!

She askes WhatsApp, how to move the data; and it answered, that if she moved from iPhone to iPhone, or from Android to Android, she could use a backup; but from …

Thursday
11 September, 2025

LibreOffice Dev Blog: Using C++ STL functions instead of loops – EasyHack

@ 14:24 UTC

C++ Standard library, which resides in std:: namespace provides common classes and functions which can be used by developers. Among them, Standard Template Library (STL) provides classes and functions to better manage data through data structures named containers. Here I discuss how to use STL functions for better processing of data, and avoid loops.

Checking Conditions

To iterate over a container to see if some specific condition is valid for all, any, or none of the elements in that container, C/C++ developers traditionally used loops.

On the other hand, since C++11, there are functions that can handle such cases: all_of, any_of and none_of. These functions process STL containers, and can replace loops. If you want to know if a function returns true for all, any, or none of the items of the container, then you can simply use these functions. This is the EasyHack dedicated to such a change:

153109 – Use all_of, any_of or none_of from STD instead of looping to check a condition over a range

Here is an example patch which uses any_of instead of a loop:

-    bool bFound = false;
     // convert ASCII apostrophe to the typographic one
     const OUString aText( rOrig.indexOf( '\'' ) > -1 ? rOrig.replace('\'', u'’') : rOrig );
-    size_t nCnt = aVec.size();
-    for (size_t i = 0;  !bFound && i < nCnt;  ++i)
-    {
-        if (aVec[i] == aText)
-            bFound = true;
-    }
+    const bool bFound = std::any_of(aVec.begin(), aVec.end(),
+        [&aText](const OUString& n){ return n == aText; });

As you can see, the new code is more concise, and avoids using loops.

Conditional Copying, Removing and Finding

If you want to copy, remove or simply find a value in a container which conforms to a specific functions, you may use copy_if, remove_if or find_if.

153294 – Use copy_if, find_if or remove_if to simplify loops

Again, this is an example patch:

-  for ( size_t i = 0; i < SAL_N_ELEMENTS( arrOEMCP ); ++i )
-        if ( arrOEMCP[i] == codepage )
-            return true;
-
-    return false;
+    return std::find(std::begin(arrOEMCP), std::end(arrOEMCP), codepage) != std::end(arrOEMCP);

Final Words

Refactoring code is a good way to improve knowledge on LibreOffice development. The above EasyHacks are among EasyHacks that everyone can try.

More information about EasyHacks, and how to start working on them can be found on TDF Wiki:

TDF Wiki – EasyHacks

Tuesday
09 September, 2025

Miklos Vajna: Interdependent tracked changes improvements in Writer, part 3

@ 06:23 UTC

Writer has some support for interdependent (or hierarchical) tracked changes: e.g. the case when you have a delete on top of an insert. See the second post for background.

This work is primarily for Collabora Online, but the feature is available in desktop Writer as well.

Motivation¶

With the already mentioned improvements in place, the area of format redlines with character style or direct formatting changes were still lacking: Writer's original model here was just marking a text range as "formatted" and then either accept the format redline as-is, or reject reverting back to the paragraph style (default formatting), losing the old character style or old direct formatting.

Results so far¶

Here is a sample case of a document where the old character style is Strong (~bold) and the font size is 24pt, while the new character style is Quote (~italic) and the font size is 36pt. The rest of the document uses no specific character styles and has the font size of 12pt:

Rejecting that format redline resulted in just the defaults, i.e. no character style and 12pt font size:

But now we track the old character style & direct format:

This required changes in the DOCX import, ODF import and ODF export, too.

How is this implemented?¶

If you would like to know a bit more about how this works, continue reading... :-)

As usual, the high-level problem was addressed by a series of small changes. Core side:

Want to start using this?¶

You can get a development edition of Collabora Online 25.04 and try it out yourself right now: try the development edition. Collabora intends to continue supporting and contributing to LibreOffice, the code is merged so we expect all of this work will be available in TDF's next release too (26.2).

Thursday
28 August, 2025

LibreOffice Dev Blog: Unit conversion in LibreOffice code – EasyHack

@ 14:06 UTC

LibreOffice handles different input and output formats, and also displays text and graphics alongside inside the GUI on computer displays. This requires LibreOffice to understand various different measurement units, and convert values from one to another.

Unit selection

The unit conversion can be done by writing extra code, where one should know the units, and calculate factor to convert them to each other.

For example, consider that we want to convert width from points into 1/100 mm, which is used in page setup.

We know that:

1 point = 1/72 inch
1 inch = 25.4 mm = 25400 microns
factor = 25400/(72*10) ≈ 35.27777778

Then, it is possible to write the conversion as:

static int PtTo10Mu( int nPoints )
{
return static_cast<int>((static_cast<double>(nPoints)*35.27777778)+0.5);
}

A separate function that casts integer nPoints to double, then multiplies it by the factor which has 8 decimal points, and then rounds the result by adding 0.5 and then truncates it and stores it in an integer. This approach is not always desirable. It is error-prone, and lacks enough accuracy. For big values, it can calculates values off by one.

Another approach is to use o3tl (OpenOffice.org template library) convert function. It is as simple as writing:

int nResult = o3tl::convert(nPoint, o3tl::Length::pt, o3tl::Length::mm100)

As you can see, it is much cleaner, and gives the output, properly rounded as an integer!

You need a double? No problem! You can use appropriate template to achieve that:

double fResult = o3tl::convert<double>(nPoint, o3tl::Length::pt, o3tl::Length::mm100)

These are the supported units, defined in the header include/o3tl/unit_conversion.hxx:

mm100 – 1/100th mm = 1 micron

mm10 – 1/10 mm

mm – millimeter

cm – centimeter

m – meter

km – kilometer

emu – English Metric Unit (1/360000 cm)

twip – Twentieth of a point (1/20 pt)

pt – Point (1/72 in)

pc – Pica (1/6 in)

in1000 – 1/1000 in

in100 – 1/100 in

in10 – 1/10 in

in – inch

ft – foot

mi – mile

master – PPT Master Unit (1/576 in)

px – Pixel unit (15 twip, 96 ppi)

ch – Char unit (210 twip, 14 px)

line – Line unit (312 twip)

Handling Overflows

If you are doing a conversion, it is possible that the result overflows. With o3tl::convert() you can handle it this way:

sal_Int64 width = o3tl::convert(nPoint, o3tl::Length::pt, o3tl::Length::mm100, overflow, 0);
if (overflow)
{
...
}

Code Pointers

To to find instances to change, one can try finding some magic numbers listed here. For example, consider measuring a line based on twips:

line – Line unit (312 twip)

If you search for 312, you may find some examples:

$ git grep -w 312 *.cxx

Final Words

The task described here is filed as tdf#168226:

Use o3tl::convert() for unit conversion instead of custom formula

EasyHacks are well-defined small tasks that are designed to help newcomers begin LibreOffice programming. If you like it, you can start working on it!

Using o3tl::convert() not only simplifies the …

Friday
15 August, 2025

Miklos Vajna: Multi-page floating tables in Writer: keep together or not

@ 06:14 UTC

This post is part of a series to describe how Writer now gets a feature to handle tables that are both floating and span over multiple pages.

This work is primarily for Collabora Online, but is useful on the desktop as well. See the 11th post for the previous part.

Motivation¶

Previous posts described the hardest part of multi-page floating tables: making sure that text can wrap around them and they can split across pages. In this part, we'll look at a conflicting requirement. On one hand, headings want their text to not split across pages (and shapes anchored into paragraphs are considered part of the paragraph, too). On the other hand, it should be OK to have a floating table at the bottom of a page and the following heading to go to the next page.

It turns out, Writer gave "keep together" a priority, while Word gave "floating tables are OK to split to a previous page" a priority.

Note that if you have a shape (e.g. a triangle) and not a floating table, then both Word and Writer prevents the move of that shape to a previous page (if the shape is anchored in a heading); this difference was there just for floating tables.

Results so far¶

Here is how the tdf#167222 bugdoc looks like now in Writer:

And here is how it used to look like:

And here is the reference rendering:

This means that we leave layout for shapes unchanged in general: shapes anchored in headings are still considered to be part of headings and don't split. But for floating tables, we now allow them to split and use space at a previous page if they fit there.

How is this implemented?¶

If you would like to know a bit more about how this works, continue reading... :-)

As usual, the high-level problem was addressed by a series of small changes:

Want to start using this?¶

You can get a development edition of Collabora Online 25.04 and try it out yourself right now: try the development edition. Collabora intends to continue supporting and contributing to LibreOffice, the code is merged so we expect all of this work will be available in TDF's next release too (26.2).

Monday
11 August, 2025

LibreOffice QA Blog: QA/Dev Report: July 2025

@ 17:04 UTC

General Activities

LibreOffice 25.2.5 was announced on July 17
Olivier Hallot (TDF) updated help for CSV import, explained Property Mapping in help for Charts and improved help for Calc’s FILTERXML function and AutoFilter
Gábor Kelemen (Collabora) did many code cleanups
Tomaž Vajngerl (Collabora) made internal hyperlinks in a table of contents accessible when exported to PDF/UA
Pranam Lashkari, Szymon Kłos and Hubert Figuière (Collabora) worked on LOKit used by Collabora Online
Parth Raiyani (Collabora) did reorganisations in some dialogs

Miklós Vajna (Collabora) polished the support for floating tables in Writer, fixed some crashes and continued improving the handling of tracked changes that depend on each other
Xisco Faulí (TDF) added Albanian and Moldovan locale, fixed short weekdays in Romanian locale, improved the translation checker script, added some new automated tests, upgraded many dependencies and did many code cleanups and optimisations
Michael Stahl (Collabora) fixed an issue with expansion of list level numbering formats with repeated levels and fixed a column width issue in RTF tables
Mike Kaganski (Collabora) implemented Markdown export, fixed not being able to apply colour to Chart walls via Sidebar, fixed an issue with paragraph numbering in RTF files, helped Miklós with floating table polishing, fixed an issue with date conversion in Base, made URL handling more robust in Extension updating code, fixed and issue with spacing in lists in RTF files, fixed RTF export issues causing loss of bullet fonts and “No character border” explicit formatting and fixed some crashes. He also did many code cleanups and optimisations
Caolán McNamara (Collabora) fixed many issues found by static analysers and did code cleanups and optimisations
Stephan Bergmann (Collabora) worked on the WASM build. He also adapted the code to compiler changes and did code cleanups
Noel Grandin (Collabora) made Skia rendering backend mandatory on Windows and greatly improved the import time of CSV data with trailing newline characters. He also did many code cleanups and optimisations, especially in the area of transparency handling
Justin Luth (Collabora) made it so failed command line operations return exit status 1, allowing for automated bisecting of command line issues among other things, fixed an issue with spellchecking and the option “Check uppercase words” and fixed a style continuity issue with page breaks in DOCX files
Michael Weghorn (TDF) continued cleaning up and reorganising accessibility-related code, made the orientation radio buttons in Envelope dialog accessible, fixed an issue with unwanted focus accessibility events being fired in Borders tab page of Writer’s Paragraph dialog, made the border preset selection be clearly indicated when focused, implemented support for native colour pickers in GTK and Qt UIs and did cleanups and reorganisations in Android, vcl and report design code. He also worked on using native widgets in Qt UIs
Balázs Varga (Collabora) implemented support for Microsoft Media Foundation APIs on Windows for playback of common codecs, fixed Calc’s MATCH function returning an incorrect result with inline arrays and fixed an issue with …

Tuesday
29 July, 2025

LibreOffice QA Blog: LibreOffice 25.8 RC2 is available for testing

@ 11:55 UTC

LibreOffice 25.8 will be released as final on August, 20, 2025 (check the Release Plan). LibreOffice 25.8 Release Candidate 2 (RC2) brings us closer to the final version, which will be preceded by Release Candidate 3 (RC3). Since the previous release, LibreOffice 25.8 RC1, 70 commits have been submitted to the code repository and 34 issues got fixed. Check the release notes to find the new features included in this version of LibreOffice.

LibreOffice 25.8 RC2 can be downloaded for Linux, macOS and Windows, and it will replace the standard installation.

In case you find any problem in this pre-release, please report it in Bugzilla (you just need a legit email account in order to create a new account).

For help, you can contact the QA Team directly in the QA IRC channel or via Matrix.

LibreOffice is a volunteer-driven community project, so please help us to test, we appreciate your contribution! Happy testing!!!

…

Friday
25 July, 2025

Mike Kaganski: Microsoft, anybody home?

@ 06:39 UTC

You know what: Microsoft became miserably incompetent in IT.

I develop open-source code. But that never made me one of the “I hate proprietary software or IT giant corporations” types. I always saw the nice things that Microsoft offered to its users; I saw not only downsides in its products. And I also used (and continue to use) things created by it: Windows to start with (and I develop there, being able to debug and address issues specific to the platform that most of our users use); but also its email service for personal mail.

This Monday, I decided to send something to LibreOffice dev mailing list. Something I do from time to time, you know. Not too fascinating, right?

Well, this time, it turned out, Microsoft decided to teach me to fear them. Thunderbird shown me a message, that the mail couldn’t be sent (well, not a problem: will re-try again…), but then I found myself logged off, with “Your account has been blocked” message. They decided, that I violated their service agreement!

FTR: here is the mail. I was able to send it using another tech giant’s mail service. You may see that it’s full of links. Yes, that’s true; I prefer to provide references to my words. But tell me where was it violating anything in MS agreement?

OK, they have a stupid AI that is worse than good old filters. OK, they made it react immediately, as an undoubted authority. But that’s not a big problem, right? They provide a way to appeal! Let me do that.

And of course, they ask for the phone, and I provide it, just to get a nice reply:

And guess what: there is no other method!

OK! Let’s ask their support. (I am approaching to the point that fascinated me most.) I found a link to “Contact Microsoft Support” on the “Troubleshooting verification code issues” page; and after some automatic answers there, which didn’t answer my problem, I finally got a button telling me … tada …

Yes, you got it right. “Here is a page where we discuss problems signing in. You attempted our FAQ suggestions? You still can’t sign in? No problem! Contact our Support team, and we will solve your problem is a minute! But first, please sign in to continue.”

Heh. I used my wife’s account to contact support. And then I was given a very secret link to an appeal form, where I could file a support ticket. And the next morning, I got a message! Yay! It told me to do something! Let me try! What is that they tell me to do? Reading… hmm… go to sign-in page, and when they tell me that my account is blocked, provide a phone number? Wasn’t it exactly the thing I attempted and failed, and told them about that? But hey, they obviously fixed that problem overnight, they couldn’t just send me the useless instructions, right …

Thursday
10 July, 2025

LibreOffice QA Blog: LibreOffice 25.8 RC1 is available for testing

@ 11:00 UTC

LibreOffice 25.8 will be released as final at the end of August, 2025 ( Check the Release Plan ) being LibreOffice 25.8 Release Candidate 1 (RC1) the third pre-release since the development of version 25.8 started at the beginning of December, 2024. Since the previous release, LibreOffice 25.8 Beta1, 178 commits have been submitted to the code repository and 101 issues got fixed. Check the release notes to find the new features included in this version of LibreOffice.

LibreOffice 25.8 RC1 can be downloaded for Linux, macOS and Windows, and it will replace the standard installation.

In case you find any problem in this pre-release, please report it in Bugzilla ( You just need a legit email account in order to create a new account ).

For help, you can contact the QA Team directly in the QA IRC channel or via Matrix.

LibreOffice is a volunteer-driven community project, so please help us to test – we appreciate it!

Happy testing!!

…

Tuesday
08 July, 2025

Miklos Vajna: Interdependent tracked changes improvements in Writer, part 2

@ 06:43 UTC

Writer has some support for interdependent (or hierarchical) tracked changes: e.g. the case when you have a delete on top of an insert. See the first post for background.

This work is primarily for Collabora Online, but the feature is available in desktop Writer as well.

Motivation¶

With the already mentioned improvements in place, a few areas were still lacking: we didn't have UI for all cases where the DOCX import was possible already; combining tracked changes (redlines) were not complete (so you don't have to reject all parts of a logical redline one by one) and some of the undo/redo code didn't work as expected.

Results so far¶

Here is a sample case where the UI was missing to create something that was possible to import from DOCX: a format redline on top of an insert redline.

If you had a document with an insert:

And you selected BBB to mark those characters as bold, we just updated the existing insert redline to be bold:

But now we track a format change on top of the insert separately:

This is also visible if you open the track changes dialog, which explains that now you have part of the insert redline covered by a format redline:

How is this implemented?¶

If you would like to know a bit more about how this works, continue reading... :-)

As usual, the high-level problem was addressed by a series of small changes. Core side:

Want to start using this?¶

You can get a development edition of Collabora Online 25.04 and try it out yourself right now: try the development edition. Collabora intends to continue supporting and contributing to LibreOffice, the code is merged so we expect all of this work will be available in TDF's next release too (25.8).

Monday
07 July, 2025

LibreOffice QA Blog: QA/Dev Report: June 2025

@ 11:33 UTC

General Activities

LibreOffice 25.2.4 was announced on June 6
Olivier Hallot (TDF) added help for compact layout Pivot Tables, Writer table formula MOD and improved the help for Writer’s Send menu commands, font colour, text attributes for drawing objects, Calc shortcut keys, Calc’s LOOKUP function, Of-Pie charts and file conversion filters
Pierre F. added help for Writer table formula INT
Gábor Kelemen (Collabora) simplified code for VCL settings and did many other code cleanups
Tomaž Vajngerl (Collabora) continued polishing support for embedded fonts in PowerPoint files
Marco Cecchetti (Collabora) worked on LOKit used by Collabora Online
Jaume Pujantell (Collabora) improved the import of ref fields in DOCX files
Parth Raiyani (Collabora) makde it so the Layouts panel in Impress Sidebar now uses a native IconView widget instead of the ValueSet widget
Miklós Vajna (Collabora) fixed a Writer list indent removal issue, implemented RTF export of section breaks right sections, fixed an issue with images inside shapes being sized incorrectly in RTF files and continued improving the handling of tracked changes that depend on each other
Xisco Faulí (TDF) implemented Writer table formula MOD, added support for transparent fill colour in SVGs, removed the Euro converter wizard, added several new automated tests, upgraded many dependencies and did many code cleanups and optimisations
Michael Stahl (Collabora) continued working on multi-user editing based on a conflict-free replicated data type (CRDT) and improved the stability of handling Writer comments within the Navigator
Mike Kaganski (Collabora) fixed invisible tree view expansion triangles in Python script organizer, fixed an issue with storing selected encodings in the Text Import dialog, fixed a document read error, fixed an issue with accessing VBScript objects in macros, made menu command code more robust, fixed display of Cyrillic text in RTF files, fixed issues with pasting shapes between LibreOffice applications, fixed Manage Changes dialog not enabling Accept / Reject buttons initially, improved the display of Calc’s Number format dialog in the case of a selection containing different number locales, fixed an RTF table width issue, made bracketing of selected text more robust, fixed a Skia/Vulkan rendering issue affecting line numbers in Basic IDE, fixed an issue with Basic IIf function when used with array indexes, reduced console message noise when running LibreOffice from the command line and fixed several crashes. He also did many code cleanups and optimisations
Caolán McNamara (Collabora) helped Heiko with vertical tabs, fixed crashes and many issues found by static analysers and did code cleanups and optimisations
Stephan Bergmann (Collabora) worked on the WASM build
László Németh added an indicator for justified lines with overly large word spacing

Noel Grandin (Collabora) improved rendering speed of transparent shape fills dramatically, fixed a Windows GDI backend resource use issue seen in documents with lots of styles when the style preview is visible, made Skia rendering backend mandatory on macOS and improved the performance of style handling in Calc. He also did many code cleanups and optimisations
Justin Luth (Collabora …

Friday
13 June, 2025

LibreOffice QA Blog: LibreOffice 25.8 Beta1 is available for testing

@ 15:47 UTC

LibreOffice 25.8 will be released as final at the end of August, 2025 ( Check the Release Plan ) being LibreOffice 25.8 Beta1 the second pre-release since the development of version 25.8 started at the beginning of December, 2024. Since the previous release, LibreOffice 25.8 Alpha1, 782 commits have been submitted to the code repository and 154 issues got fixed. Check the release notes to find the new features included in this version of LibreOffice.

LibreOffice 25.8 Beta1 can be downloaded for Linux, macOS and Windows, and it can be installed alongside the standard version.

In case you find any problem in this pre-release, please report it in Bugzilla ( You just need a legit email account in order to create a new account ).

For help, you can contact the QA Team directly in the QA IRC channel or via Matrix.

LibreOffice is a volunteer-driven community project, so please help us to test – we appreciate it!

Happy testing!!

…

Tuesday
10 June, 2025

LibreOffice QA Blog: QA/Dev Report: May 2025

@ 11:29 UTC

General Activities

LibreOffice 24.8.7 was announced on May 8
Olivier Hallot (TDF) added a help page for Page Layout, expanded help for paragraph justification, updated menu paths in Help, added help pages for newly-added Calc functions and of-pie charts, updated help for Business cards and Labels, improved extended tooltips and error messages for Manage Names dialog, corrected an example spreadsheet used for Calc Data Statistics help and improved help for IsNull BASIC function among many other Help cleanups and updates
Gábor Kelemen (allotropia) did many code cleanups
Tomaž Vajngerl (Collabora) continued polishing support for embedded fonts in PowerPoint files and fixed unexpected changing of background images upon saving and reloading in Draw
Darshan Upadhyay, Szymon Kłos, Michael Meeks and Jaume Pujantell (Collabora) worked on LOKit used by Collabora Online. Szymon also implemented saving checkbox state to XLSX files
Gökay Şatır (Collabora) fixed an issue in Draw with connector text disappearing when “Adjust to contour” option was active
Marco Cecchetti (Collabora) added a feature to select colour palettes for chart data series

Pranam Lashkari (Collabora) fixed an issue with unwanted expansion of reference mark fields after insertion and typing
Miklós Vajna (Collabora) continued improving the handling of tracked changes that depend on each other
Xisco Faulí (TDF) implemented new Calc functions TEXTSPLIT, TEXTBEFORE and TEXTAFTER, made it so glue points in PowerPoint shapes are imported, fixed an issue with connectors in PPTX files becoming misaligned due to negative rotation, added some new automated tests, upgraded many dependencies, fixed crashes and did many code cleanups and optimisations
Michael Stahl (allotropia) worked on multi-user editing based on a conflict-free replicated data type (CRDT) leveraging yrs, a Rust port of Yjs
Mike Kaganski (Collabora) fixed an error when accessing cells via BASIC methods after deleting cells with the RemoveRange method, made it so BASIC’s Time() function returns a Date type, greatly improved the loading time of Writer documents with lots of bookmarks and lots of tables, made it so empty Writer paragraphs correctly follow proportional line spacing smaller than 100%, fixed an issue preventing the export of Draw / Impress documents to SVG from Basic IDE context, improved BASIC error messages, fixed incorrect width in SVG text with “fit-to-size” attribute, fixed an issue with macros not pausing for the duration of executing dialogs and helped Heiko with the new Welcome dialog
Caolán McNamara (Collabora) fixed an issue with expanding the list of events in the Events tab of Customize dialog, fixed crashes and many issues found by static analysers and did code cleanups and optimisations
Stephan Bergmann (allotropia) worked on the WASM build. He also adapted the code to compiler changes and did code cleanups
Noel Grandin (Collabora) improved the loading speed of XLSX files with lots of customFormat attributes in rows, dramatically improved the rendering speed of documents with large page fills when hardware acceleration is used, greatly improved the loading speed of XLSX files with lots of formulas, conditional formatting and comments and improved …

Monday
02 June, 2025

Miklos Vajna: Interdependent tracked changes improvements in Writer

@ 11:54 UTC

Writer has some support for interdependent (or hierarchical) tracked changes: e.g. the case when you have a delete on top of an insert. While there were some working cases, handling of many combinations were missing. I started to make systematic improvements in this area in the recent past, this post gives you an overview what's done so far.

This work is primarily for Collabora Online, but the feature is available in desktop Writer as well.

Motivation¶

DOCX files in Word can often have overlapping tracked changes: Writer tries to split these up to make sure there is only one tracked change under the cursor at the same time. Still, it's possible that you have a tracked change with multiple types: e.g. a delete on top of an insert.

The focus in on 3 combinations which appear in DOCX files a lot: "insert, then delete", "insert, then format" and "delete, then format".

This mostly affects the UI and import/export filters of ODT and DOCX.

Results so far¶

Given an insert, then delete:

Most operations worked nicely here, but in case your cursor was in the middle of AAA and you did a reject, followed by an undo, proper handling of that was missing, now implemented.

But then given an insert, then a format:

Then a handling of more actions were missing:

DOCX import is now implemented.
ODT import is now implemented.
Accepting when you're inside AAA is now implemented: the insert is accepted for BBB but the format stays unchanged.
Rejecting when you're inside AAA is now implemented: the insert is rejected and BBB is also removed, together with the format on top of it.
Accepting the BBB now correctly operates on the insert type, so the format type remains after accept.
If you accept BBB, now the surrounding AAA and CCC also get accepted as well, as expected.
Now if you reject BBB, then it gets removed from the document, since you rejected an insert.
When you reject BBB, the surrounding AAA and CCC also get rejected.

The combined implementation of these should give you a smooth feeling in case you're used to how Word works: if there is a format redline combined with an insert, then the operations act on the insert type, and format is only accepted/rejected when there is no insert "under" the format.

Similarly: it's a bit of an implementation detail that Writer splits redlines on DOCX import: so if you e.g. accept AAA then we combine that with BBB and CCC when it makes sense, so you need to click a lot less.

Finally, given a delete, then a format:

Then again handling of some actions were missing:

DOCX import is now implemented.
ODT import is now implemented.
ODT export is now implemented.
Accepting AAA now correctly operates on the delete type of BBB.
Rejecting AAA …

Wednesday
28 May, 2025

= allotropia: Collabora and allotropia merge

@ 10:20 UTC

This deal unites the largest team of corporate Office engineers to deliver on Collabora Productivity’s mission to restore Digital Sovereignty to its users, while making Open Source Office Rock. It supercharges Collabora’s Online Office products and services portfolio with rich German language capability, deeper experience of vertical applications, new Web Assembly skills, and a wider unified partner ecosystem. Through improved product richness this sharpens the competitive edge of FLOSS Office productivity against mass-market proprietary alternatives.

CAMBRIDGE, UK – May 28th 12:00 CEST – 2025

Collabora Productivity, the world’s leading provider of collaborative Open Source Office editors have completed a merger with allotropia. Collabora has invested heavily in building Collabora Online (COOL) – a market leading, on-premise, secure, interoperable, open-source solution for document editing and collaboration deployed to any modern browser. This is complemented by desktop and mobile apps across Linux, Windows, Mac, Android, iOS and Chrome-OS. Collabora provides support subscriptions to enterprise customers worldwide via a network of hundreds of trusted partners. This is now augmented by allotropia’s partner and customer base. Together with our partners we deliver document and productivity excellence integrated with our partners product and service offerings.

allotropia’s expertise around Web Assembly combined with Collabora Online will we expect, in time, enable customer use-cases such as well as office-as-component embedding scenarios in vertical applications as well as off-line and end-to-end encrypted editing, and. This work builds on some visionary prototype funding from the Bundesministerium des Inneren (BMI) for a collaboration between the companies to enable the use of Collabora Online off-line in the browser.

Further details of product investment, and direction will be announced and decided in workshops with our key customers and partners at our annual COOL Days conference in Budapest next week where staff, community and our customer and partner-ecosystem meet, swap ideas, and hear about the latest work in our upcoming major release featuring improved performance, usability, interoperability and much more.

“Collabora is excited to welcome each member of the allotropia team today!” said Michael Meeks, CEO, Collabora Productivity, “We are excited to work together to accelerate our product development, enjoy our first COOL Days together, and plan the next features and possibilities to delight our customers.”

Collabora has invested in building a network of hundreds of partners and is approaching one hundred million docker image downloads of its document editing server software, with millions of paying users of its products, all of whom will start to benefit from this merger from today.We expect to bring the experience that allotropia has from it’s relationship with CIB around vertical desktop applications (Fachverfahren) to help partners and customers migrate their Windows & Microsoft Office based business process to easy to deploy multi-platform web applications.

“With our awesome team of engineers, and our WebAssembly know how, we can add significantly to Collabora’s powerhouse of Office engineering prowess & their product offerings”, says Thorsten Behrens, CEO of allotropia, “we’ve worked with them as partners for many years, and align perfectly in our goals …

Monday
26 May, 2025

Mike Kaganski: How could QA catch this in advance?

@ 09:59 UTC

Yesterday I merged a fix for Writer’s tdf#165094. Not that it was something exceptional; something that often happens when we change the huge code: a regression. Something that we try to do for them: a fix. Why mention it here?

It happens to show something, that people underestimate. The complexity of what they call “proper testing” – you know, that “I found a bug! Do you even try to test your software???” rant you often see in discussions. Let’s look at this case.

The problem was, that in some specific document, where there was a manually inserted page break, that page break, defined in a hidden paragraph, disappeared after an upgrade. Sounds easy? Should be caught immediately in the release testing? But other page breaks weren’t lost.

Debugging showed, that the bug would only occur when all of the following happened:

The page break was defined in a hidden paragraph (something already known from the reporter – thank you Gabor!), and
There were at least 26 paragraphs before that hidden paragraph, all on the same page, and
The page break defined a paragraph style, and
That page break defined a page number, and
That assigned new page number happened to be the same “oddity” as the current one (i.e., either the number of that page with 26+ paragraphs was odd, and the new page number was odd; or the number of that page with 26+ paragraphs was even, and the new page number was even), and
After the hidden paragraph (which defined the page break), a table immediately followed.

I suppose, that’s a combination of factors, that any QA engineer would naturally test first, don’t you agree? (Disclaimer: no I don’t think so.)

Note that the complexity of this constellation of causing factors is, again, not uncommon in our codebase. In fact, it only needed less than ten features to take their specific forms, from thousands of features and options that the suite offers.

But it is completely unsurprising, that the bug, that requires such a constellation of factors, actually appeared in our bug tracker. Given the tens of millions of users, who work with who knows how many documents, every low-probability event will happen, sooner or later. This is good; and we are thankful to everyone who files bugs.

And let me say, that we at Collabora Productivity are glad to do many good things to make the office suite better for everyone.

Thursday
08 May, 2025

Miklos Vajna: Reinstate for tracked changes in Writer

@ 06:44 UTC

Writer has the concept of rejecting tracked changes: if a proposed insertion or deletion is not wanted, then one can reject it to push back on the proposal. So far such an action left no trace in the document, which is sometimes not wanted. Calling reinstate on a change behaves like reject, but with history: it reinstates the original state, with the rejected change preserved in the document.

This work is primarily for Collabora Online, but the feature is available in desktop Writer as well.

Motivation¶

When Alice works on a document to insert e.g. new conditions for a contract, then perhaps Bob is not happy with the proposal. But just rejecting the change "silently" would not be polite: the tracked change then disappears, so possibly Alice thinks it was accepted and Bob didn't communicate the pushback explicitly in the resulting document, either.

Reinstate is meant to improve this interaction: if an insert is reinstated, then an explicit delete is created on top of the insert, so Alice can see that Bob was not happy with the proposal. Or in case Alice proposed a delete, Bob can reinstate that by adding the same content again to the document, without typing the text manually after the delete.

This is a UI feature: the resulting model still only contains inserts and deletes, so it works even with DOCX files.

Results so far¶

Given an insert:

Now you can easily create a delete on top of the insert:

And given a delete:

Now you can easily create an insert right after the delete, preserving complex content:

As you can see, this creates the opposite of the original change as a new tracked change, so it will in the end still reject the change, but without deleting the original change.

How is this implemented?¶

If you would like to know a bit more about how this works, continue reading... :-)

As usual, the high-level problem was addressed by a series of small changes. Core side:

Online side:

cool#11357 sw redline reinstate: add UI for this

Want to start using this?¶

You can get a development edition of Collabora Online 25 …

Wednesday
30 April, 2025

Marius Popa Adrian: FirebirdSQL introduces support for Windows ARM64 builds

@ 11:36 UTC

This FirebirdSQL pull request introduces support for Windows ARM64 builds to the Firebird project. The changes cover updates to build scripts, configuration files, and Visual Studio solution/project files to accommodate ARM64 architecture, ensuring compatibility and enabling compilation and functionality on Windows ARM64 platforms.

Marius Popa Adrian: SQL-compliant aliases GREATEST and LEAST for the existing MAXVALUE and MINVALUE functions.

@ 07:24 UTC

This FirebirdSQL pull request introduces SQL-compliant aliases GREATEST and LEAST for the existing MAXVALUE and MINVALUE functions. These aliases align with the SQL:2023 standard and provide a more intuitive and widely recognized syntax. The changes include updates to documentation, keywords, parser tokens, and system function definitions to support these new aliases.

Friday
18 April, 2025

LibreOffice Dev Blog: Splash screen with VCL weld – difficultyInteresting EasyHack

@ 13:37 UTC

As a LibreOffice user, you have certainly seen the LibreOffice splash screen. It is displayed when you open LibreOffice, it has a progress bar, and when loading the application is finished it goes away. Here we discuss a suggested improvement for this splash screen.

Current Implementation Approach

Currently, the splash screen is implemented by creating a custom widget with a custom painting mechanism that draws the splash image and also the progress bar and moves the progress indicator.

This has some drawbacks:

1. The splash screen does not always scale to the same size as the main LibreOffice Window.

2. The style of the progress bar is somehow different from other UI elements, looks mostly like gen interface.

3. It needs and uses a custom paint code.

4. It does not conform to the dark/light theme.

5. It is not easily localize-able. In fact, the only text is from the displayed image, in English. When you build from sources, the image file is instdir/program/intro.png.

LibreOffice splash screen bitmap

6. It is a separate binary (oosplash). You may run it with:

$ ./instdir/program/oosplash

LibreOffice dev splash screen

VCL Weld Mechanism

I have previously written about VCL weld mechanism, which is based on creating user interface files (.ui) and loading them inside the application.

The weld mechanism greatly reduces the complexity of creating user interfaces, and also improves other aspects of the user interface, including the consistency.

Code Pointers

Most of the code for the current implementation resides in:
desktop/source/splash/splash.cxx.

The SplashScreenWindow class has an custom paint method, SplashScreenWindow::Paint(), which draws the bitmap, and also the progress. A new UI file is needed for this purpose, which should use GtkProgressBar, which will be considered a weld::ProgressBar. VCL then uses appropriate progress bar widget in different graphical plugins of VCL.

You may look into some dialogs like tip of the day to get some insight:

Easyhacking: How to create a new “Tip-Of-The-Day” dialog

It would be interesting to avoid a separate binary, but it is fine to keep things as is, and just change to use .ui file.

Final Words

The above issue is tdf#166128. If you would like to work on fixing it, you can just follow the Bugzilla link to see more information.

You may also use ideas from a minimal weld application here:

VCL weld: create LibreOffice GUI from design files

Overview

Welcome to The Document Foundation Planet

Languages:

Actions:

Metadata:

Keyboard shortcuts:

Friday14 November, 2025

Official TDF Blog: The role of ODF in digital sovereignty (digital freedom)

@ 12:25 UTC

Thursday13 November, 2025

Official TDF Blog: Announcement of LibreOffice 25.8.3

@ 13:21 UTC

Thursday06 November, 2025

LibreOffice QA Blog: QA/Dev Report: October 2025

@ 12:00 UTC

General Activities

Tuesday04 November, 2025

Miklos Vajna: Interdependent tracked changes improvements in Writer, part 4: direct accept/reject

@ 07:11 UTC

Motivation¶

Results so far¶

How is this implemented?¶

Want to start using this?¶

Wednesday22 October, 2025

LibreOffice Dev Blog: enumarray for better data arrays – EasyHack

@ 19:30 UTC

Using Symbolic Constants

Using enumarrays

Why an enumarray?

Final Notes

Tuesday21 October, 2025

LibreOffice QA Blog: QA/Dev Report: September 2025

@ 11:20 UTC

General Activities

Thursday16 October, 2025

LibreOffice Dev Blog: enum class instead of unscoped enum – EasyHack

@ 14:05 UTC

Rationale

Finding Instances

Examples Commits

Implementation

Tuesday07 October, 2025

Miklos Vajna: Markdown import/export in Writer

@ 06:13 UTC

Motivation¶

Results so far¶

How is this implemented?¶

Want to start …

Thursday18 September, 2025

LibreOffice Dev Blog: Debugging tips for LibreOffice

@ 14:11 UTC

Choose the Right Debug Platform

Debugging Tools

Environment Variables

Pretty Printers

Dumping Data

Tuesday16 September, 2025

LibreOffice QA Blog: QA/Dev Report: August 2025

@ 09:42 UTC

General Activities

Monday15 September, 2025

Mike Kaganski: A fairy tale about poor UX enforcing vendor lock-in

@ 08:25 UTC

Thursday11 September, 2025

LibreOffice Dev Blog: Using C++ STL functions instead of loops – EasyHack

@ 14:24 UTC

Checking Conditions

Conditional Copying, Removing and Finding

Final Words

Tuesday09 September, 2025

Miklos Vajna: Interdependent tracked changes improvements in Writer, part 3

@ 06:23 UTC

Motivation¶

Results so far¶

How is this implemented?¶

Want to start using this?¶

Thursday28 August, 2025

LibreOffice Dev Blog: Unit conversion in LibreOffice code – EasyHack

@ 14:06 UTC

Handling Overflows

Friday
14 November, 2025

Thursday
13 November, 2025

Thursday
06 November, 2025

Tuesday
04 November, 2025

Wednesday
22 October, 2025

Tuesday
21 October, 2025

Thursday
16 October, 2025

Tuesday
07 October, 2025

Thursday
18 September, 2025

Tuesday
16 September, 2025

Monday
15 September, 2025

Thursday
11 September, 2025

Tuesday
09 September, 2025

Thursday
28 August, 2025

Friday
15 August, 2025

Monday
11 August, 2025

Tuesday
29 July, 2025

Friday
25 July, 2025

Thursday
10 July, 2025

Tuesday
08 July, 2025

Monday
07 July, 2025

Friday
13 June, 2025

Tuesday
10 June, 2025

Monday
02 June, 2025

Wednesday
28 May, 2025

Monday
26 May, 2025

Thursday
08 May, 2025

Wednesday
30 April, 2025

Friday
18 April, 2025