By Jake Edge
September 28, 2010
A group of OpenOffice.org developers has announced the creation of an independent
foundation - called the Document Foundation - to guide the further
development of the office suite, which is provisionally named LibreOffice.
At the heart of this effort is longtime OpenOffice.org developer Michael
Meeks. We had the good fortune to discuss the LibreOffice effort with
Michael; read on for his comments on this new initiative
LWN: Probably the first question that will come to mind for most of
our readers is "Why?" — why fork OpenOffice.org? And why now?
Well, it has been ten years since a foundation was promised as
part of the
original OpenOffice.org announcement, and there is now a
confluence of circumstances to realise that goal. We want a vendor
neutral body that lots of companies and non-profits can contribute to
as peers. That foundation is called the Document Foundation, and for
trademark reasons our product will be called LibreOffice.
LWN: What do you see as the advantages of LibreOffice for OpenOffice.org users? developers? distributions?
For developers, we are open for business - we have a realistic
view of the code-base and as such we are interested in including
people's fixes and improvements quickly. When we can get people
working to clean up the code, translate German comments, remove dead
code, fix ergonomic nits, write unit tests and so on - we are
optimistic that we can produce a far better product, and one that (as
developers) we can be proud of.
Linux distributions should find LibreOffice easier to package,
as the development team has a vast amount of Linux distribution
experience.
All of that of course leads to getting a better, more stable,
and featureful office suite into users' hands.
LWN: Do you plan to require copyright assignment or contributor
agreements? If so, what would those entail? And if not, why not?
There are no plans to require copyright assignment, clearly it
is important to determine the origin of all code, so we will use a
clear signing off / attribution trail, and familiar git tooling to
make that easy.
Having to sign formal paperwork before contributing code is
clearly a formidable barrier to entry, even if the rights end up with
a well-governed non-profit. In contrast I believe LibreOffice needs an
"All Contributions Welcome and Valued" sign outside, that says come in
and help, there is a place for you here.
LWN: What are the near-term technical and community goals for the project? What about the longer-term?
In the near term, we expect to clean-up the code; we have a
set of janitors tasks that require (in some cases) no previous
programming experience whatsoever e.g. removing commented out code that
was just left lying around (presumably due to a lack of faith in
revision control). If you want to get the eternal glory of having your
name in the LibreOffice code-base, now is a great time to get involved.
We also want to target tackling many of the problems that have
traditionally made it hard to develop with, such as the arcane and
monolithic build system.
Finally - there are a lot of ergonomic nits in OpenOffice,
that individually are easy to fix but collectively add up to a big
problem. We want to start tackling these in the short term.
Longer term - we are developing a plan, but somehow our press
experts persuaded us to delay announcing it, expect to hear more
around the Linux Plumbers Conference.
LWN: When might we expect the first LibreOffice release? Presumably
it will incorporate the patches that go-oo has been maintaining, but
are there patches from elsewhere that might make their way into the first
release or two? Any exciting features on the horizon that we haven't seen
in go-oo yet?
We have already released
a beta. It
is a distinct piece
of code from go-oo for several reasons, most importantly being that we
don't want to maintain patches anymore. Go-oo was maintained as
patches, such that features could be enabled per-platform or per
distribution simply by not applying them but this brings maintenance,
and development problems of its own.
Instead with LibreOffice we will have several flat git
repositories, such that the git diff output will be your patch, and
committing is as simple as a git push. Of course many of the go-oo
features have been merged, some are still pending review, and going
forward go-oo will be obsoleted by LibreOffice.
LWN: Does LibreOffice plan to track OpenOffice development and
incorporate changes from that code base or does it plan to go completely in
its own direction? Or will there be a gradual shift from one to the other?
Clearly we are going to merge all (suitably licensed) code
into the project from anywhere we can get it. Previously we would work
from whatever Oracle released, but in future we will pick and choose the
best changes and features from wherever they come.
LWN: Are you at all concerned about maintaining such a large body of
code without the resources of a large company like Sun or Oracle behind the
effort?
Clearly Oracle's contribution is real and substantial, and we
would dearly like them to participate in the Document Foundation, a
warm welcome is extended to them. Nevertheless - both Novell and
Red Hat have support capabilities around OpenOffice.org and are
confident that we can fix and improve the code. Clearly, having
dependence on any single company to support or drive the project is a
huge risk factor. There is a perception out there that the code is
terribly tangled and impossible to develop with, but the reality is that
it is just code. Sure you have to read some parts quite carefully, and
empathise deeply with the authors before altering them, but this is true
of all large pieces of code.
LWN: There have been occasional hints that Sun had patents on some
StarOffice/OpenOffice components and we have seen that Oracle is not
terribly shy about patent litigation; does the project have any concerns about patents or patented technology in the codebase?
The OpenOffice.org code-base that LibreOffice is derived from
is licensed under the LGPLv3 - which gives us all a strong explicit
patent license, and a good copyright license, so no. Clearly for new
code we would want a plus ["or any later version"] license, so we are
considering recommending
a LGPLv3+ / MPL combination for entirely new code.
LWN: Who is involved with this new LibreOffice project? Undoubtedly
there were individuals besides yourself, along with companies, and perhaps
other groups, what can you tell us about who they are and what their roles
will be?
Oh certainly, I, and Novell are only a small part of this
effort, a large proportion of the non-Oracle OpenOffice.org community
is of like mind, and are instrumental in helping to create
LibreOffice. I anticipate the Foundation we create ultimately looking
more like the GNOME Foundation than the Mozilla Foundation, i.e. with
only a small staff for co-ordination, rather than for central
development. I hope we will have similar elections of contributors for
representatives and so on.
There is a list of people behind the foundation on the
LibreOffice
web-site, if I start naming them all we will run out of
space pretty quickly. Of course, there are also a good number of heroes
who managed somehow to get their code and fixes into an OpenOffice
product in the past, that should find it a pleasure to contribute in
future.
LWN: Have you had any discussions with Oracle about any of this?
You are inviting them to join forces with the new project, have they
expressed any interest, either formally or informally?
Clearly we have informed Oracle's StarDivision management
ahead of time, as is only polite. As to their reaction - I have many
developer friends in StarDivision whom I respect and have loved
collaborating with in the past. My hope is, that we will work together
again.
[ We would like to thank Michael for taking the time to answer our
questions. ]
(
Log in to post comments)