LibreOffice only forked from OpenOffice.org six weeks ago. Already, however, news about its future directions is starting to trickle out.
The details are sometimes sketchy, but they suggest that LibreOffice and OpenOffice.org could diverge more quickly than most observers imagined.
Initially, The Document Foundation (TDF), which oversees LibreOffice development, announced a general set of principles in The Next Decade Manifesto, a name that refers to the fact that OpenOffice.org recently celebrated its tenth anniversary.
This manifesto promises that TDF will support free software and open standards and work methods, as well as other causes, such as the preservation of endangered languages.
To make matters clearer, the manifesto also states what TDF rejects, including monopolies and proprietary formats. The manifesto also promises open, peer-reviewed development.
These declarations are promising, so far as they go. They try to distinguish TDF not only from Microsoft Office, its main proprietary format, but also from Oracle Corporation, the current owners of OpenOffice.org, which many people believe lacks dedication to free software practices.
However, the declarations are too high-level to give much indication of what directions LibreOffice might be heading.
Recently, though, more concrete information has been released. This information appears in LibreOffice blog entries, notably those of Charles Schultz for November 10 and October 28, and in TDF announcements, one in German and the other in English.
Although the announced priorities are still vague in one or two places, and no one of them gives a complete list of changes, they are still the most detailed information about LibreOffice’s directions that has been released.
Fortunately, Italo Vignoli, the member of the TDF steering committee who wrote the English release, has provided some clarifications to me privately.
Reconsidering The Office Suite
The vaguest reference in the public statements is the declaration that, “After 20 years of feature oriented software, it is now the right time to bring back content at the centre of user focus.”
The 20 years is a reference to the age of the code, and its mention suggests a criticism of how Oracle, Sun Microsystems, and the original StarDivision decided its development priorities, but the exact criticism is uncertain.
Asked to explain this declaration, Vignoli writes, “So far software has been focused more on features than on contents, and a good user is considered [one] who is able to use features and not [one] who is able to develop good content.”
As a result, modern office suites include many features that users either do not need or do not use. “Of course, this does not mean that software should have less features,” he adds.
According to Vignoli, one thing that needs to be taken into consideration is the proliferation of hardware platforms. “Editing and reading on a large screen is not like reading on a small screen,” he notes.
“In addition, being mobile adds another layer of complexity, because the relationship with contents is different when you are on the road: your attention is lower and your time pressure is higher.”
To judge from these comments, TDF is apparently using the break with OpenOffice.org to reconsider priorities.
My speculation is that something like OOO4Kids, with its different interfaces for different levels of users might be an answer to unwanted features, while the mention of multiple hardware platforms suggests that TDF may be considering the frequent requests for a version of the code suitable for mobile devices.
The general nature of the responses suggests that TDF is still developing the details, but would prefer to pay greater attention to usability than OpenOffice.org did in the past.
New Development Paths
The significance of other details in the official statements is more obvious, although once or twice I asked Vignoli to clarify:
Each application in LibreOffice will undergo what the English new release calls “an extensive rewrite”: This is by far the most ambitious direction, and undoubtedly the most needed.
OpenOffice.org source code is infamously cryptic, and is probably a major reason why the project has had trouble attracting new developers.
Sun Microsystems tried to evade this problem by making it easier to write extensions in a variety of programming language, but that is only a temporary solution, which ultimately adds to the confusion.
The fact that TDF is willing to tackle such a huge problem at all is promising, although no timeline for the rewrite is giving.
But, as Charles Schulz writes, “Our code base is getting old. Worse, the whole frigging software looks and feels like weâ€™re stuck in the Bush area. Many things were not fixed, some others need a complete rewrite.”
The Rewrite Will Begin With Calc
To some, this starting point may seem a surprise. The word processor being the most heavily used application in any office suite, you might assume that the rewrite would begin with Writer.
But Writer is already far in advance of MS Word, so starting with the LibreOffice spreadsheet — the second most widely used application in an office suite — seems only natural.
The Calc rewrite is to feature a new engine called Ixion, and to include improved interaction with databases and support for VBA macros to improve compatibility with MS Excel.
To anyone who knows their mythology, the engine’s name will sound like an indirect comment on the difficulty of the task; according to Greek mythology, Ixion is a prince who is bound to a fiery wheel for eternity down in Hades.
A Reduction Of The Dependency On Java
Since Sun owned Java’s code, the corporation encouraged its use in OpenOffice.org. Java is not required for basic functionality in OpenOffice.org, but some templates, wizards, and extensions do require it.
Now, with Oracle suing Google over Java copyrights and patents, avoiding Java is a sensible decision, especially for a group like TDF that has probably drawn Oracle’s ire already.
Java has become a potential liability to LibreOffice, just as it was in some observers’ estimation before Sun released its source code.
Better First Draft Functions For Writer
This change is mentioned only in the German news release, and no more details are given. However, it may refer to improved outlining tools, which have often been requested.
Improved “Layout Fidelity” In Writer And “Slideshow Fidelity” In Impress
These features are mentioned only in the English release. Vignoli explains, “Many people use Writer and Impress with sophisticated layouts, which sometimes create problems when printing or displaying, especially when there are multiple fonts with different metrics….
In addition, documents are moved between platforms (Windows to Mac to Linux, and vice versa) and this increase fidelity problems.”
If the gap between on-screen display and output, or between layouts on different versions of the software are too great, then users have to print regularly to check their results.
Similarly, if different software versions fail to give the same results, users have to spend extra time tweaking their documents.
At times, they may have to limit layouts to design elements that can cross platforms without creating problems.
Improved Format Conversion
Asked if this referred specifically to compatibility with Microsoft Office and its newer OOXML format, Vignoli replies, “Definitely.
Users are discovering interoperability, and the future should be one where we all use our preferred format (although at TDF we all prefer the same one” — that is, Open Document Format).
The difficulty with this goal is that given the constant changes in Microsoft formats, efforts at compatibility have historically always been at least one step behind what users need. However, simply by including the support for OOXML in Go-OO, TDF could improve conversion for many users.
Missing And Unmentioned, And The Future
One omission that might surprise some observers is the lack of any mention of Mono support and features.
Given that TDF’s supporters include Novell, the foremost Mono development, and that Mono featured in Go-OO, which many TDF supporters worked upon, many assumed that Mono would also be a presence in LibreOffice. Yet, so far, it has not been mentioned.
Another point that gets little attention is major changes to the interface. Charles Schultz does mention the need, but not, apparently, as a declared goal.
However, not only could the interface do with some rationalization, but LibreOffice also has to decide if it will include any support for a ribbon interface like the one in the current releases of Microsoft Office.
Still, as vague and as ambitious as some of the proposed changes sound, at least they are a start.
No doubt, with TDF still deciding on its governance as well as its development plans, expecting more so early in the organization’s existence is unrealistic. Still, the directions mentioned do suggest that TDF’s members have a reasonably solid sense of priorities.
Another uncertainty is how quickly these changes will be implemented, and which take priority.
Although LibreOffice has shown a flurry of activity in its first six weeks, it is still be written mainly by 20 coders, and the proposed changes are likely more than they could easily handle in the next few months.
In fact, in announcing these directions, TDF might run the risk of promising more than it can deliver.
All the same, if these goals are only partly realized, then LibreOffice’s compatibility with OpenOffice.org may diminish as early as the next release.
Files could still be shared between the two office suites, but some features in one might become unavailable in the other — unless, of course, OpenOffice.org borrows LibreOffice’s code for itself.
That would be a humiliating exercise for Oracle, but it might well become a necessary one.
Meanwhile, despite some vagueness, TDF sounds like it has more direction than Oracle.
If nothing else, it sounds more willing to make a fresh start and to question long-established conventions.