User: Password:
Subscribe / Log in / New account

IBM to contribute Symphony to

From:  Rob Weir <apache-LO3lX/>
Subject:  Symphony contribution
Date:  Thu, 14 Jul 2011 00:00:57 +0200
Message-ID:  <>
Archive-link:  Article

I'm sending this with my IBM "hat".

I'm going to be speaking at the ODF Plugfest in Berlin Friday
afternoon, and will be making some announcements.  I wanted you to
hear this first, before anyone else knows.

You know about IBM Lotus Symphony, our free (as in beer) product which
is based on OOo.  We're doing well with it, I think.  Along with
various numerous interop, performance enhancements and functional/bug
fixes, we've done some significant work in the accessibility and user
interface in general.  If you saw recently, PC Magazine gave Symphony
3.0 its "Editors' Choice Award" [1].  In the review they praised the
"interface that's been tweaked by IBM to make it by far the
user-friendliest no-cost productivity suite, and one's that's friendly
enough to rival the spacious and informative interface that Microsoft
created for Office 2010 and that Apple created for iWork '09."  So the
UI enhancements are getting some notice.

However, we at IBM have not been exemplary community members when it
came to  This wasn't necessarily by design, but for
various reasons, that was the effect.  Yes, we participated in various
community councils, and sponsored conferences and worked together on
standards.  But when it came down to the code, we maintained Symphony
essentially as a fork, and although we occasionally contributed code
back, we did not do this well, or often.

We'd like to make some changes in how we do things, and the fresh
start at Apache is a good opportunity for this.

We will be doing the following:

First, we're going to contribute the standalone version of Lotus
Symphony to the Apache project, under the Apache 2.0
license.   We'll also work with project members to prioritize which
pieces make sense to integrate into OpenOffice.  For example, we've
already done a lot of work with replacing GPL/LPGL dependencies.
Using the Symphony code could help accelerate that work and get us to
an AOOo release faster.  We've already converted the help files to
DITA, which could help accelerate that work, if we chose to go in that

Aside from the work that would help accelerate getting AOOo to our
first release, we've also added other features that I think we should
consider merging in.  For example, the IAccessible2 work which helps
Symphony work better with assistive technology.    I know an older
version of this work sits in an OOo CWS someplace, but it will be
easier to integrate that work if we start with our latest code.  We've
also added VBA macro support, which is great for MS Office interop.

Also, as the PC Magazine review notes, we've done some really good UI
work.  I invite you to download Symphony [2] and take a closer look at
this.  Yes, it is different from what OOo has today.  And a move of
that magnitude has an impact on documentation and translations as
well.   But the feedback we've received from customers and reviewers
is very positive.  Do we integrate parts of the Symphony UI?  That is
something for the project to discuss and decide on.

Finally, we will be proposing [3] a new incubation project at Apache,
for the ODF Toolkit.  These Java libraries enable new kinds of
lightweight document processing applications.  We think this would
work well as an Apache project, and we look forward to moving that
into incubation and developing that complementary project forward.

So that's essentially what I'll be announcing on Friday.  The above
contributions will occur over the next couple of months, starting with
the ODF Toolkit.  I hope you see the exciting possibilities as much as
I do.






(Log in to post comments)

It is on!

Posted Jul 14, 2011 16:33 UTC (Thu) by kragilkragil2 (guest, #76172) [Link]

OK, now it seems like the Apache/IBM and the TDF are going to duke it out.
If OO.o really integrates most of what IBM is offering then it will be getting harder for LibrOffice to keep up with integrating changes.

IMO Apache should just integrate most of what IBM has to offer, sure it would make OO.o essentially Synphony, but if you want OO.o just use LibreOffice. Having 2 Office suites that are more or less identical serves no purpose.

It is on!

Posted Jul 14, 2011 18:01 UTC (Thu) by mjw (subscriber, #16740) [Link]

Yes, it will be interesting to see who will catch up to who. On the one hand OpenOffice/Apache Office has basically not seen any development for the last 2 months, while TDF/LibO has seen 3 releases in that same timeframe. On the other hand maybe IBM will finally kickstart Apache into action.

Does anybody understand what exactly is offered though?
It seems it isn't the UI parts, since those are java and would need a special eclipse framework and a jvm, and they are only offering C++ code. And this post makes it sound like they will not have any community input since they will do the integration in their own ClearCase (!?!) repository:

Wondering about the name

Posted Jul 14, 2011 18:05 UTC (Thu) by dmarti (subscriber, #11625) [Link]

"Apache Symphony" has a certain ring to it. has always been a clunky name, since there are too many things called "Open.*" and the ".org" (required to avoid a trademark conflict with some other "Open.*" thing) is confusing. Will Apache be able to use the "Symphony" name?

Wondering about the name

Posted Jul 14, 2011 18:11 UTC (Thu) by corbet (editor, #1) [Link]

Dunno about Symphony, but an ongoing conversation suggests that ".org", at least, will go. "Apache OpenOffice" is apparently good enough to be a distinct name.

Wondering about the name

Posted Jul 18, 2011 18:02 UTC (Mon) by xorbe (guest, #3165) [Link]

I thought .org was tacked on because another pre-existing commercial Open Office got upset.

It is on!

Posted Jul 15, 2011 0:32 UTC (Fri) by AndreE (guest, #60148) [Link]

I might be wrong, but I was under the impression that Lotus Symphony was actually quite rubbish. At least those were the reviews that went out when it was launched.

This may have changed, but for OpenOffice to benefit from this, then Lotus Symphony must actually include some features and functionality that is considered quite desirable.

It is on!

Posted Jul 15, 2011 5:28 UTC (Fri) by rahvin (subscriber, #16953) [Link]

Even if there is the apache license is compatible with the LO licenses. That means LO can simply take the code they like and plunk it into LO. This has always been the problem with the fork for the OO side, the code can go only one way and that's into LO, LO's improvements can't go back to Apache OO.

Either way I doubt there is much to worry about. I just don't see the momentum in OO. IBM is going to throw over the fence some code but they aren't committing to keep contributing.

It is on!

Posted Jul 16, 2011 18:31 UTC (Sat) by Wol (guest, #4433) [Link]

Which Symphony?

First of all, *most* office suites were rubbished in the late 90s / early 2000s because they weren't Microsoft Office. And if you're talking Symphony of the early 90s or the 80s, surely it evolved a lot after that?

Secondly, IBM Lotus Symphony is not Lotus Symphony. I don't know the details, but from what I can make out, IBM killed the original Lotus Symphony, and replaced it with their own version of OOwriter called Lotus Symphony.


It is on!

Posted Jul 15, 2011 9:51 UTC (Fri) by njwhite (guest, #51848) [Link]

Indeed, IBM powered OpenOffice can continue to build the codebase in the direction it has a reputation for; large, java-y, slow, etc. And that may well work for their customers.

Meanwhile LibreOffice can work towards creating an office suite which is usable and faster.

The two projects diverging is to my mind a good thing; I want a free, community governed office suite (LibreOffice), and the less "brand confusion" the better.

It is on!

Posted Jul 16, 2011 14:39 UTC (Sat) by Wol (guest, #4433) [Link]

Don't forget. LibreOffice will probably have a much easier time of it integrating Symphony than will OOo!

MOST of the effort in the first few releases of LO has been directed towards removing dead code, refactoring redundant code, translating the comments into English, and a whole host of *programmer* *friendly* changes.

OTOH, OOo was (still is?) a pretty ugly spaghetti monster. Okay, they could probably take and use a whole host of patches that are basically just "delete this useless code, delete that useless code", but it's probably almost more effort to review it than just redo it from scratch.

That's a good reason why LO is likely to start leaving OOo in the dust - it's far more developer-friendly, and once you get a commanding lead, it's hard for the other people to catch up ...


IBM to contribute Symphony to

Posted Jul 14, 2011 17:54 UTC (Thu) by jonabbey (guest, #2736) [Link]

I'd be happy to take a look at Symphony on my linux box, but it's completely opaque as to how to get the damn thing to run.

I finally found something called '/opt/ibm/lotus/Symphony/framework/rcp/rcplauncher' and tried running that, but it complained about my home directory being NFS mounted, told me that I should set a 'nonpersistent' setting some place to avoid it cacheing, and threw an exception trace.

I have no idea how to configure settings in the thing, and IBM's Symphony page is itself opaque next to almost any open source product. No 'Documentation' link, e.g.

The only reason that I even knew to look for a 'nonpersistent' setting is that I ran it from the command line. If I tried using the Freedesktop launcher icon, it didn't visibly do anything at all.

I used OS/2 back in the day, and this kind of user unfriendly software experience is very familiar to me. It's disappointing that IBM still can't
adopt to outside convention after all these years.

By having a 'bin' directory, say.

It's pretty obvious why Symphony has gotten wholly overlooked in the Linux world, guys.

IBM to contribute Symphony to

Posted Jul 14, 2011 18:12 UTC (Thu) by nix (subscriber, #2304) [Link]

By having a 'bin' directory, say.
Or even, radical idea, by just working if your home directory is NFS-mounted, like every other piece of Linux and Unix software out there. I can understand not wanting to run over NFS if you're a massive relational database, but for anything else there is just no excuse.

IBM to contribute Symphony to

Posted Jul 14, 2011 18:40 UTC (Thu) by dlang (subscriber, #313) [Link]

actually, massive relational databases (like Oracle) don't complain about running over NFS, in fact, I've had cases where that has been the vendor recommended configuration (the performance will depend heavily on your NFS server, and you don't want to have multiple systems accessing the data over NFS at the same time... but it can work very well under the right conditions)

IBM to contribute Symphony to

Posted Jul 14, 2011 23:08 UTC (Thu) by nix (subscriber, #2304) [Link]

Oh yes, of course. Sorry, the last time I installed a big Oracle DB was so long ago (2002? something like that) that direct I/O was still the preferred-verging-on-mandatory I/O method.

IBM to contribute Symphony to

Posted Jul 14, 2011 18:00 UTC (Thu) by jonabbey (guest, #2736) [Link]

Also, it appears that Symphony incorporates Eclipse, and is implemented as a hierarchy of Eclipse plugins? Did you guys replace the entire O.o interface with an SWT/Eclipse one? Why?

IBM to contribute Symphony to

Posted Jul 14, 2011 18:09 UTC (Thu) by jonabbey (guest, #2736) [Link]

Ah, yes, it is built as a set of Eclipse RCP plugins.

IBM to contribute Symphony to

Posted Jul 14, 2011 19:16 UTC (Thu) by job (guest, #670) [Link]

My guess is that it's because Symphony is about embedding Openoffice in their Lotus suite (large parts of which is written in Java).

IBM to contribute Symphony to

Posted Jul 15, 2011 2:16 UTC (Fri) by Hausvib6 (guest, #70606) [Link]

I can't get those grinding harddisk noise off my head after reading your comment, I accidentally started imagining the splash screens of "Eclipse" and "" at once.

Anyway, it looks like that IBM wants to steer the development of AOOo (and I'm sure Oracle know this would happen) which may push further dependencies on Java regardless of whether it's a good or bad. It would be great if TDF can eradicate the (currently optional) dependencies of Java in LibreOffice since they have liberated theirselves from the evil kingdom of Oracle in the first place. Bonus points for having a lean and mean office suite that starts as fast as nano.

IBM to contribute Symphony to

Posted Jul 15, 2011 11:29 UTC (Fri) by kragilkragil2 (guest, #76172) [Link]

If you think that the DF can make LO as fast as nano, or just fast in general you probably also believe in aliens, easter bunnies, Santa Clause, some $Deity, miracles and tooth fairies.

LO might get a lot faster (than it currently is or faster than Symphony), but it will never be fast. Just like a 16-Wheeler will never win a Formula 1 race. A thing called reality will make sure of that.

IBM to contribute Symphony to

Posted Jul 15, 2011 12:37 UTC (Fri) by oblio (guest, #33465) [Link]

It's not about performance, it's about perceived performance. And it's 95% doable.

Think Chrome > galculator.

IBM to contribute Symphony to

Posted Jul 15, 2011 13:42 UTC (Fri) by Hausvib6 (guest, #70606) [Link]

Relax, it's just a hyperbole.


Posted Jul 16, 2011 14:44 UTC (Sat) by jensend (guest, #1385) [Link]

One of the things which makes LWN a better forum than Slashdot etc for real tech discussion is that not every single topic devolves into adolescent religion-bashing.

Quite a lot of the most important contributors to computer science, open source, etc are deeply religious people (Larry Wall and Donald Knuth are the first two that come to mind, if you're looking for concrete examples) and quite definitely believe in God and in miracles. These people deserve your respect.

If you want to discuss the rationality of believing in God and in miracles, you're quite welcome to: 1) take a good philosophy of religion course so you understand the theist responses to the silly arguments atheists are used to trolling everybody with and then 2) find a forum more appropriate for the topic. But presuming that everybody thinks religion is idiotic and then using it as a way to insult people is a puerile way to drag the level of discussion at this forum into the mud.

Stop yelling

Posted Jul 17, 2011 1:56 UTC (Sun) by kragilkragil2 (guest, #76172) [Link]

I am well aware that most people think relegion is a good thing. But facts and reality don't matter to people in that regard, they have their believes (probably tought by their parents) and I am not going to change that.
And I didn't say those people are stupid, they are just irrational and believe in imaginary things. They can believe in God and miracles all day long, that is fine by me, but don't call it facts or rational. You can pray all day for the miracle that will make LO as fast as nano. It won't happen. I am sure.
But I also think there are a lot of bright people in FOSS and science who don't believe in superstion, for sure more than in the general population. Go read Linus' blog and his comments in there and find out where he stands for yourself.
I know that the net result of all religions on this planet is utterly detrimental to mankind (how many wars and genocides have/had strong religous aspects?). So the world would be a much better place without them. Human beings in general are hard wired for compassion, love and help etc. We need money- and power-grabbing religions for nothing. Being guided by the wisdom of your fellow man, facts, reality, by science and by lessons learned from our history would be a much better thing for mankind than hypocritical false religions.

Over and out (for real) I know you won't agree and tell me how wrong I am and that is fine. I won't come back here to these comments. This discussion is off topic anyways. There will be no LO miracles.

Why are you so sure?

Posted Jul 17, 2011 9:16 UTC (Sun) by khim (subscriber, #9252) [Link]

I know that the net result of all religions on this planet is utterly detrimental to mankind (how many wars and genocides have/had strong religous aspects?).

Sorry, but I don't see the logic: how second part is related to the first one? Most technical wonders (both in ancient times and today) are created for the war (the latest, most well-known example is the Internet), so I'm not all that sure religions and wars are somehow "detrimental". And even if you ignore that aspect you must remember that religions do not create or start wars - they are used as justification, sure, but actual reasons are usually deeper. Think Crusades (the original ones: 11th, 12th, and 13th centuries in Europe). They were side-effect of the solution to quite real and acute problem: overpopulation. Rules which governed inheritance were actually quite positive WRT to overall progress of the society, but they created lots of "excess people" who had titles but no land and little money; Crusades neatly tied up this dangerous mass. When another solution to the problem was found Crusades stopped. Religious people still wanted to continue, but society rejected these pleas.

Thus no, I don't think it can be easily said that "the net result of all religions on this planet is utterly detrimental": they play role of "safety valve" and pacify the population (give good, noble, justification for objectively needed atrocities) and in fact it looks like net sum is quite positive indeed.

Human beings in general are hard wired for compassion, love and help etc.

Where such crazy ideas come from? It's just stupid. Yes, humans are capable of compassion, love, help and so on - but to say that are hardwired to do that... Sorry, but not even close. They are hardwired to survive and propagate (just like all other creatures). When compassion, love and other such qualities are helpful (and they often do) - everything is fine and religion takes a backseat, when they are detrimental - you need a religion to justify you actions.

Being guided by the wisdom of your fellow man, facts, reality, by science and by lessons learned from our history would be a much better thing for mankind than hypocritical false religions.

What this has to do with anything? Religions rarely (if ever) actually guide the mankind. Sure, they are used to justify things a lot, but to actually guide? Nope. You've put them on pedestal they don't deserve - no wonder you've gotten wrong impression about them.

Why are you so sure?

Posted Jul 17, 2011 13:40 UTC (Sun) by corbet (editor, #1) [Link]

OK, folks, can we stop this one here? Religion is getting pretty off-topic for LWN, and I'm not sure that much good will come from discussing it further.

Except, of course, that blasphemous users of the wrong editor are still fair game.

IBM to contribute Symphony to

Posted Jul 15, 2011 7:11 UTC (Fri) by xtifr (subscriber, #143) [Link]

Y'know what? I don't think this is a battle. Symphony is something that IBM cares about, because they have customers to support. And currently, it's based on OOo. OOo and LO have already drifted apart, and it would take some effort to rebase Symphony on LO and guarantee the quality that IBM wants to guarantee. Symphony is already working today.

By liberating Symphony, IBM has maximized their options for minimal effort. It's possible that people will take the effort to re-base a Symphony on top of LO, in which case IBM gets the benefit of LO's improvements. But if that doesn't happen, they're no worse off than they were, and they still get some "many eyes" effect on the Symphony code. It's pretty much a win-win for IBM no matter how this plays out. The worst possible scenario is that they're no worse off than they were.

I'm going to reject the this-is-an-attack-on-LO theories until I see a lot stronger evidence. IBM is generally pro-copyleft, especially when it comes to their own code. I think this is just a good move for IBM and basically a good move for the rest of us. I mean, who can complain about a large amount of working, widely-used code contributed to the community? Even if you don't want to use Symphony.

I might wish this had happened sooner, before the official OOo/LO split. But it didn't, and so, from where we stand today, I think this is, overall, about as good an event as I could have hoped for. Kudos to IBM!

IBM to contribute Symphony to

Posted Jul 15, 2011 8:56 UTC (Fri) by jamesh (guest, #1159) [Link]

People porting the Symphony modifications to LibreOffice is not necessarily a win for IBM. If the result is licensed under the LGPL like the rest of LibreOffice, then they might not consider the result to be usable.

From this email it sounds like they have devoted effort to stripping LGPL code from OpenOffice, so it seems unlikely that they would want to add more.

IBM to contribute Symphony to

Posted Jul 15, 2011 20:53 UTC (Fri) by Wol (guest, #4433) [Link]

But LibreOffice is NOT licenced under the LGPL !!! If you want, you can - !legally! - ignore the (L)GPL entirely!

ALL LibreOffice code (that is, code donated to the LibreOffice project) is licenced MPL. Add the fact that the original code base has been relicenced to Apache by Oracle, and (murkiness aside) you can now legally copy the entire LibreOffice project without using the (L)GPL for as much as one line of code!

The choice of MPL by TDF was apparently deliberate to be friendly to IBM.


IBM to contribute Symphony to

Posted Jul 16, 2011 4:32 UTC (Sat) by am (subscriber, #69042) [Link]

IBM to contribute Symphony to

Posted Jul 16, 2011 14:46 UTC (Sat) by Wol (guest, #4433) [Link]

That page refers to the binary. And you're correct, the binary is LGPL3 (in fact, given the licencing murkiness now Oracle has donated OOo to Apache, maybe it isn't. Anyways, that WAS true before Oracle changed the licence).

As for MPL, look here ...

Let's split the LO code base in two - the code that came from OOo, and the code that's been added whether Go-OO or LO.

The OOo code has been relicenced Apache.
The Go-OO/LO code is available under the MPL.

So, it's a fair claim that you can copy all of LO without going near the (L)GPL. To be safe, yes you probably would want to make sure that all the Oracle code had been relicenced as AL2.


IBM to contribute Symphony to

Posted Jul 16, 2011 21:01 UTC (Sat) by tzafrir (subscriber, #11501) [Link]

For those who did not bother following links: yes, LO *is* licensed under LGPLv3 (and also MPL - dual licensed).

IBM to contribute Symphony to

Posted Jul 15, 2011 22:00 UTC (Fri) by xtifr (subscriber, #143) [Link]

I can't imagine why they'd think the GPL would hinder them in any way. They've shown themselves to be quite intelligent about copyleft in the past, and are fully supportive. As for stripping the LGPL parts--OOo is currently under the Apache license, and they have no guarantee that anyone will modify it to work with LO, so stripping the incompatible-with-Apache parts simply helps keep their options open. This way it can merge with OOo if other options don't play out.

IBM to contribute Symphony to

Posted Jul 18, 2011 11:32 UTC (Mon) by nim-nim (subscriber, #34454) [Link]

Symphony is managed by Java developpers. Most big corp Java people are hopelessly prejudiced against anything GNU-ish, after years of SUN "GNU is evil, do not touch Linux" propaganda (the prejudice is so deep SUN never dared pushing OpenJDK even after choosing to copyleft it, for fear of alienating partners and its own workforce)

IBM to contribute Symphony to

Posted Jul 15, 2011 7:19 UTC (Fri) by slashdot (guest, #22014) [Link]

Apache License 2.0 is GPL-compatible, so whatever good is there can just be integrated into LibreOffice.

IBM to contribute Symphony to

Posted Jul 15, 2011 13:07 UTC (Fri) by kragilkragil2 (guest, #76172) [Link]

Sure, but somebody still has to do it. It is not like TDF has a lot of resources. My guess is that the codebases will diverge if IBM grts its way, which doesn't have to be bad:

Still no code ...

Posted Jul 18, 2011 9:10 UTC (Mon) by mmeeks (subscriber, #56090) [Link]

Nice announcement; but there is still -no- code. Neither code from ASF in a repository under AL2, nor code from IBM. For IBM this is not -that- surprising, some of this code has been announced as to be contributed multiple times before:


And then of course with the ASF OO.o announcement, and now yet again.

And still their code is not published. It makes one wonder if there is a limit to the number of times something can be re-announced before people loose interest. We can anticipate -at-least- one more announcement - as/when the code actually becomes public & we can read/analyse it finally.

Copyright © 2011, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds