OpenOffice is now Apache Office

StarOffice invented StarOffice. Sun purchased StarOffice and renamed it OpenOffice. Oracle purchased Sun to get control of Java and everyone expected OpenOffice to bloat out into JavaOffice. Instead Oracle gave OpenOffice to the Apache Foundation to maintain along side Apache and a whole lot of other open source projects. Why? How will the change affect you?

Libre Office

The Oracle treatment of OpenOffice forced a bunch of OpenOffice developers to split away and form LibreOffice. Libre Office is rapidly replacing OpenOffice in Linux distributions. LibreOffice is not yet a full replacement for OpenOffice but it is so close that few people notice the change. By 2012 LibreOffice should be a complete replacement for OpenOffice and all Linux distributions will deliver LibreOffice. OpenOffice will be obsolete.

Oracle

Oracle had no plans to fix the major problems with OpenOffice and no plans to invest in improvements. From the development view, the major problem is the software bloat due to building everything as huge chunks of code. Abiword and other programs easily beat OpenOffice at basic tasks and for frequent use. OpenOffice needs a major rewrite to achieve that level of efficiency. Oracle were not making money from OpenOffice and decided to dump OpenOffice instead of pouring more money into it.

Apache Foundation

Oracle gave OpenOffice to the Apache Foundation instead of the LibreOffice project. Why did Oracle choose the Apache Foundation to maintain OpenOffice? Nobody knows a definitive answer except perhaps Larry Ellison, the Oracle CEO, and he occasionally makes decisions that look illogical.

The Apache products use an Apache license that is accepted as open by the open source Debian project. Some people think the Apache license is not as open as other open source licenses. Acceptance by Debian is the critical bit because Debian is the base for Ubuntu, the world's most popular version of Linux, and several other distributions of Linux.

The Apache team know a lot about development and should be able to handle the OpenOffice project. There will be time lost due to the transition. I do not see any advantage to the Apache Foundation or any reason for them to focus on improving OpenOffice. They might gain some corporate sponsorship. They might make a profit and channel the profit into improving Apache to compete against Lighttpd and Nginx.

The problem with OpenOffice, compared to Abiword, is almost exactly the same as Apache versus Lighttpd. If Apache are so slow at trimming Apache down to compete with Lighttpd, how are they going to magically perform the same liposuction on OpenOffice?

Possible improvements

LibreOffice currently aims at functionality. When they have a full function compatibility, they can start rewriting for efficiency. File level compatibility will be maintained because the document file format standard is an open standard. LibreOffice will start loading code for only what you use. The startup time on your netbook, or smartphone, should be as quick as Abiword.

Think of the difference in cooking terms. It is the start of the day. All you want is a quick hit of coffee. You ask OpenOffice for an espresso. OpenOffice switches on the oven, the microwave, the toaster, the barbeque, the television, and the espresso machine. When LibreOffice is reworked, it will switch on only the espresso machine.

Expect the LibreOffice rewrite to take two years. They should, in the same process, remove all Java and adopt standard libraries for some areas. If LibreOffice can use some of the same libraries as other open source projects, it will free up LibreOffice developers for more important work.

Qt is the leading graphical user interface on smartphones and competing with GTK for the desktop. Gimp and some other applications are switching to a standard set of evolving libraries. If the libraries can be shared at the operating system level, all that code would be loaded once and shared by multiple applications, reducing the total load on your computer.

LibreOffice 3.4 is announced and 3.4 contains some changes to common code. 3.4 on Linux will use the same Cairo font rendition as Linux so the LibreOffice font presentation will look exactly the same as the same font used elsewhere in Linux.

The LibreOffice project is under the control of the people who produce the OpenOffice file format standards. Enhancements to the standards will probably occur in LibreOffice before OpenOffice. There could be a code race between the two organisations. I suspect at this stage there are not enough developers to support the two separate projects and real progress will only occur the developers can reduce their workload by replacing base code with libraries from other projects.

Abiword replacement?

Abiword has several advantages over the OpenOffice Writer. Abiword loads fast. Abiword uses less memory. Abiword has real time cooperative editing.

The Libre Office developers say the OpenOffice code is too difficult to change. A rewrite is needed to make Libre Office flexible. You can assume LibreOffice will not gain the good features of Abiword until LibreOffice is redeveloped. LibreOffice 3.4 does have some changes to reduce memory usage and to reduce duplication.

The move to netbooks and keyboardless netbooks (smartphones, slates, and the zillion other keyboardless tablet computers including the iPad, iPad 2, and the soon to be announced iPad 3) requires small efficient fast loading and fast acting applications. Both OpenOffice and LibreOffice are too big. Both OpenOffice and LibreOffice share a legacy architecture preventing them changing fast enough to keep up with modern requirements.

Conclusion

LibreOffice has replaced OpenOffice everywhere I used OpenOffice. LibreOffice 3.4 already shows rapid improvement. I am looking forward to LibreOffice after the rewrite so I can replace Abiword with a single line of office applications.