Hacker News new | past | comments | ask | show | jobs | submit login
What's next for Apache OpenOffice (lwn.net)
74 points by cesarb on Sept 8, 2016 | hide | past | favorite | 70 comments



From what I can see, if Apache OpenOffice wants to be a viable project with Libreoffice already out there and well ahead from a user perspective, it needs to focus on things that are attractive about it that differentiate it from Libreoffice. Right now, to the best of my admittedly limited knowledge, that's primarily their license. My thought on this is that they need to pick some (fairly radical) directions to take the project in that will provide clear alternatives to LibreOffice. Maybe they could review the ideas here: https://wiki.documentfoundation.org/Development/Crazy_Ideas and pick some of the more ambitious ones that Libreoffice is unlikely to attempt.


The best thing that could happen to the community at this point is to hand over the Openoffice trademarks to The Document Foundation and for them to retire the Libreoffice branding, shift everything over to the more recognizable Openoffice branding.

That way, it would actually bring people on board who are not technical, since pretty much no one who isn't into Linux knows what Libreoffice is.


This! Some people can even save face by calling it a merger, much like when io.js "merged with" node.js.


I think the answer to the headline is "sadness." but the situation is important to understand on two levels, both from technical level and from a governance level.

In my experience, understanding governance is not something that comes easily to engineers. This project however provides a textbook example of the forces involved. Governance, or the right to set direction, can be imposed if the governed are prevented from expressing an opinion on it or choosing an alternative. Dictators who simply kill anyone who stands against them, small groups where a superior assigns governance like a VP telling your group that some person is your new manager. But actual, durable, governance only occurs with the implicit assent of the governed.

Libreoffice (LO) and Apache Open Office (AOO) both existed at a point in time of the fork, when a portion of the governed (in this case the developers) withheld their assent to the impositions by AOO and started LO. From that point forward there were two options for working on open software for business tools.

Given the choice, contributors chose the LO governance over the AOO governance, and as a result what was the premier and leading open office software project is now contemplating its own demise. No one person was responsible, everyone was responsible. This sort of thing happens all the time in all sorts of situations, whether it is countries that lose faith in their leaders, or startups that are suddenly involuntarily losing all their staff, or open source projects that stop being relevant.

From a technical level this is a tremendous validation and condemnation of the power of open source. The ability to fork the software and produce a better, more secure, more maintainable product is huge. I have had the opposite experience of watching a proprietary program I really liked die because the "owner" of the software lost interest in maintaining it. It was frustrating to be unable to contribute to its survival.

But it is also a condemnation in the form of how open source is so loosely governed, there was no effective process of keeping it from getting to the fork or die state. There doesn't exist any sort of neutral arbitration board that would allow troublesome developers who were out of sync with the governed to be removed from the project. I believe that is part of the reason we have a dozen different window systems for Linux.

So bottom line is that it is a sad day for the once dominant Open Office platform, but a very good learning experience for the community.


> Libreoffice (LO) and Apache Open Office (AOO) both existed at a point in time of the fork, when a portion of the governed (in this case the developers) withheld their assent to the impositions by AOO and started LO.

I'm sorry, but this is wrong. LibreOffice was started almost a year before AOO. Take a look at https://en.wikipedia.org/wiki/LibreOffice#History for details.


That's basically quibbling, though the OP did err in his history a bit. AOO existed before that point too, it just wasn't called AOO and it wasn't part of Apache; it was run by Oracle, and before that, Sun. It was still the same project before Oracle dropped it on Apache. The portion of the governed were mad at the impositions by Oracle, not Apache, and started LO. A year later, with almost no developers left and all the Linux distros having switched to LO, Oracle finally threw in the towel and gave the dying OO.o project to Apache, where it became AOO, and the few people still hanging in there yell at clouds.


> That's basically quibbling

It's absolutely not quibbling, the OP framed his whole post around the governance of Apache OpenOffice vs LibreOffice:

>> Libreoffice (LO) and Apache Open Office (AOO) both existed at a point in time of the fork, when a portion of the governed (in this case the developers) withheld their assent to the impositions by AOO and started LO.

>> Given the choice, contributors chose the LO governance over the AOO governance

>> No one person was responsible, everyone was responsible.

No, Oracle was DIRECTLY responsible (with some blame falling on Sun). Oracle dumped OpenOffice on the Apache project when it became obvious that their continued involvement would not be profitable.

>> But it is also a condemnation in the form of how open source is so loosely governed, there was no effective process of keeping it from getting to the fork or die state.

No, this was a CLEAR condemnation of Oracle as a steward of open-source projects, and a good example of the community pulling together when the governance structure stopped working.


For what it is worth, my comment is, in essence, that the "power" of governance flows from the governed and that the Open Office project provides us with an excellent example of how the governed can express their disagreement with the governance. The example works because the "official" lineage of the project is dying, while the "forked" lineage of the project is not. It also keeps a lot of variables such as opportunity cost, market size, Etc., the same which helps it as tool for learning.

I did NOT mean to imply that it was the Apache foundation that killed (or at least mortally wounded) Open Office. What I was trying to state was that a series of governance choices killed it, the breaking point can be identified in time as the moment LO forked itself. Then from that point forward, you can compare the governance choices made by the OO lineage, and the governance choices made by the LO lineage, and see what worked and what did not. And that is what makes it such a great example (in my opinion).

For people reading and following along, and who might be considering being the steward of an open source project, they might take a moment and look at both paths thoughtfully to understand the limitations of being "the boss" on a project and how those limitations are manifested.

I get where you are coming from Cam, I screwed up by not pointing out it was Oracle in charge of Open Office when LO forked, but as that screw up changes nothing about how governance is perceived by the governed or rejected by the governed, I expect that is why your point might be considered a quibble.


I still must emphasize that you picked the incorrect governance system to include in the conflict with the developers in the open source community. The governance choices of Apache OpenOffice were largely irrelevant. Sun owned StarOffice, and continued to pay developers to work on the codebase after releasing it as OpenOffice.org. This justified their strong control over the governance of OOo, despite the friction this caused for the community. Had the community forked, it would have been uncertain which codebase would have won (gotten the most development and usage). The governance system in this case was not that of a pure open source community, but rather heavily influenced by a single corporate entity.

When Oracle acquired Sun, they ceased to govern and pulled their developers off of the project. At that point it became impossible for the community to continue working within the existing project, so the remaining developers forked to create LibreOffice. It is not adequate to say that "governance choices" killed the project, Oracle killed it but held on to the trademark.

I can think of no reason why Oracle would decide to donate the OOo project to the Apache foundation other than to generate confusion in the marketplace and deflect criticism.

At this point, I don't think the AOO has done anything wrong governance-wise, but there's just very little reason for LO to work with them. AOO development is mostly stagnant, and going through another governance restructuring and licensing change probably doesn't seem appealing to LO. I doubt the Apache project will donate the trademark to LO because they probably don't want to give up on an Apache-licensed project.

I do completely agree that governance requires assent by the governed, and the project owning the OOo trademark does not have that, but your post mis-attributed the who and the how.


>I can think of no reason why Oracle would decide to donate the OOo project to the Apache foundation other than to generate confusion in the marketplace and deflect criticism.

For a tax break? No, those assets have already been written off as R&D. Maybe it's just for the PR. I cannot conceive of Larry Ellison doing anything out of altruism.


Can you write off IPR as R&D? You could probably massively over-inflate the value of the OO.org trademarks (and the copyrights). Donating then might be a financial mechanism, it dumps the trademarks but also prevents others from using them without having to use them yourself???

Oracle and Apache's accounts would show what's happening there.


This is a conspiracy theory, but perhaps MS paid them off to do it this way just to generate confusion.


> That's basically quibbling, though the OP did err in his history a bit. AOO existed before that point too, it just wasn't called AOO and it wasn't part of Apache; it was run by Oracle, and before that, Sun.

So OOo had: 1/ a different governance 2/ different copyright owners 3/ a different license 4/ different sponsors 5/ different contributors 6/ a different name than AOO, but other than that it was totally the same?

"Other than that" being "own the trademark" here and only that. Just for fun, AOO didnt even use the trademarked term itself.


The way I remember it, mostly from following discussions on LWN, Oracle wanted to wash its hands of it, and IBM who built software on top of this platform wanted an Apache style licence. I feel IBM isn't getting enough of the blame for it coming to this.


True, IBM has built some stuff on top of OOo and I'm certain they prefer a license that doesn't require contributing their changes back. From what I understand they probably have the strongest interest in AOO's continued existence, but I didn't know if they were involved with Oracle's decision to donate OOo to the Apache project.


They were deeply involved. This is all heavily referenced in the history on Wikipedia.


> "There doesn't exist any sort of neutral arbitration board that would allow troublesome developers who were out of sync with the governed to be removed from the project."

Correct me if I'm wrong, but I thought the problems with OO.org started/accelerated when Oracle bought Sun. If that's the case, how are you meant to kick out the owners of the project? Forks are an effective defense mechanism against problems at the core of a project.


You're exactly right. The developers were mad at the project governance, and it hit a peak when Oracle bought out Sun. They created Go-OO, where they had more freedom to do their work without being stifled by the project's management, and this became LO.

And yes, forks are an effective defense mechanism against governance problems in FOSS projects. We saw it here, and we saw it years before with XFree86. Forks are an expression of democracy, where the people who actually do most of the work decide their leaders are ineffective or bad, and leave and make their own project, but thanks to FOSS licensing, they don't have to start from scratch, and are able to "fork" the code and continue their work largely without interruption.


This is another timeline mistake.

Go-oo was created in 2003, not in 2010! Even that early in the history of OpenOffice.org, there were problems with the development of the office suite. Source: https://en.wikipedia.org/wiki/Go-oo

The timeline goes like this:

1. Sun Microsystems released the first version of OpenOffice.org in 2000.

2. There were issues with building the software on Linux, therefore go-oo was created in 2003. The mismanagement of the community was brewing so long ago, as the OpenOffice.org core developers were not community-savvy.

3. Sun Microsystems were bought by Oracle and there was some hope that things would change to the better. Things did not get better, Oracle practically gave up on OpenOffice.org and the developers forked the source into LibreOffice. LibreOffice started in January 2011.

4. Oracle had enough with OpenOffice.org and wanted to get rid of it. Sadly, they asked IBM and Rob Weir to help them. While LibreOffice was already getting established, they decided to donate all of OpenOffice.org to the Apache Software Foundation.

The Apache Software Foundation does not have any experience with software intended for end-users such as office suite software. They mismanaged the whole thing. Among the mistakes, they would release new versions infrequently, they spent two years before they made any efforts to attract translators, documentation, QA team.

What is happening now is painful. The majority of Windows users end up finding openoffice.org and download the obsolete version of Apache OpenOffice. OpenOffice.org existed for a decade and the non-technical users still remember "openoffice". If these users were shifted to LibreOffice, it would be tremendous help to advance LibreOffice even further.


In the interest of full disclosure, I am a contributor to AOO, albeit one who has not been as active as I should have been in the past. Which makes me part of the problem. But that aside...

Rumors of the demise of AOO are very premature. Some internal speculation aside, there's no particular reason to think the project will be shut down. Yes, the ASF board wants to see certain things happen, but none of them are things that all of us don't also want to see. Security fixes being released in a timely fashion? Of course we want that. Frequent releases? Yep. Awesome new features? Hell yeah!

So the question is, can we get back to a place where we can deliver those things? I think so. And here's some reason why:

1. This recent "retirement" discussion has served as something of a "call to arms" or rallying cry for some AOO contributors. Count me among their number, just to show an example. I'd been pretty slack about contributing, but this has motivated me to get off my arse and get busy doing stuff. And I don't think I'm the only one.

2. We've had a number of new contributors, or at least potential new contributors, show up in the past week or so.

3. We're starting a new, more focused recruiting effort to recruit developers to the cause. Perhaps you yourself have given some thought to working on an OSS office suite? Here's a great chance!

4. We have had a release manager volunteer step forward to manage the 4.1.3 release.

5. A lot has happened just in the last week in terms of cleaning up documentation around the build process and clarifying issues there. I think it's safe to say we want to make it as easy as possible for new (or existing) contributors to build and run AOO.

Now the realist in me demands that I acknowledge that we still face a tough slog ahead. But nobody said this stuff was easy. And I feel confident that we can get this project back on track and re-establish it as the world's most awesome office suite.

Anyway, again, please consider getting involved now if you have any interest in office suites. And this is especially true for Mac users, as we could especially use more help from people who can build and test on the Mac platform.


AOO IS NOT MAINTAINED. It can't do builds, it can't make releases. It can't apply a security bug fix and build a release within 9 months. For an office suite, that's not OK.

LO is where all the active OpenOffice developers went to, LO is the continuation of OO.o.

AOO is an old and unmaintained snapshot of OpenOffice, with an apparently long-broken release build process.

I don't think there should be just one office suite, or even just one open source office suite, and I don't think everyone should just use the one with the most features. Not at all. But the AOO situation is just tricking people unfamiliar with it, to use the OO.o from 3 years ago instead of the OO.o with security and other fixes, and updates for compatibility with new operating systems (and new features which may or may not be desired).

If you hate LO developers or project, maybe you have some good reasons, and you should be able to work on a competitor based on whatever open-source code base you want. But the right thing to do at this point is abandon the "OpenOffice" name. Random people looking for "OpenOffice" need to be confronted with the fact that there are now two forks, LibreOffice and something else. And, at least for today, they should definitely choose LibreOffice.


The real, non-political competitor to the OpenOffice lineage has always been Calligra[1]. It sees dramatically less development, but it is also one of the KDE related projects that has a knack for using its toolkit to dramatically reduce its own code bloat - what takes hundreds of thousands of lines of reinventing the wheel in OO/LO is several hundred lines of extremely well maintained Qt data structures.

Even when the zombie AOO dies it does not mean the end to competition in the free office suite space. Even beyond Calligra there are some Gnome based efforts and third party projects (including web based ones) that provide similar tools. There is no shortage of competition, and if there were ever a slowdown in LO we would just see another fork akin to the Oracle escape.

https://www.calligra.org


> Calligra

That webpage though. If I told my aunt to download Calligra she would have no idea what to do once she got there. You have to click the tiny "Get Calligra" in the upper right, then scroll down and click "For MS Windows" in the middle of a list, then read through a list to find the download that's actually Calligra, then download the right msi for your architecture. Compare that to openoffice.org, and then to libreoffice.org, and watch the page get better and better for aunts everywhere.

> several hundred lines of extremely well maintained Qt data structures

Sounds amazing to me, I may try it out too. But people who aren't tech literate and are looking for a free alternative to MS Office know OO.org, and that's the real crux of the issue - OO.org is effectively dead. If OO.org instead redirected to libreoffice.org with a little disclaimer, more people would be using better free software.

The OO.org codebase isn't going anywhere. But Apache has a responsibility to its users and its reputation, and IMHO they should redirect the OO.org domain to LO.org, and let the community decide if the original OO.org codebase is worth resurrecting.


it can't make releases.

We're working on a 4.1.3 release right now.

Yes, the project has had its issues, yes there's been adversity. But we'll fight through it and come out better in the end.


> > it can't make releases.

> We're working on a 4.1.3 release right now.

Sorry, but I saw that before. The last time people said AOO couldn't make releases, they made a herculean effort and managed to make one release (4.1.2 IIRC). And here they are again, struggling to make a release.

I'll believe it not when they say "look, we made a release", but when they show the ability to do so consistently.


>We're starting a new, more focused recruiting effort to recruit developers to the cause. Perhaps you yourself have given some thought to working on an OSS office suite?

Why would I want to waste my time working on a has-been office suite when I can work on the leading FOSS office suite, which is LibreOffice?

Why would I want to commit my time to a project that seems to only seems to exist because of a licensing dispute?

Why would I want to join a chronically understaffed project when I could instead join its competitor, which has far more development support?

Why would I want to join a project which all major Linux distros have abandoned, when I could instead join its competitor, which all major Linux distros have switched to?

I'm sorry, I haven't seen any good answers to these questions at all. Working on a FOSS project that is popular and stands a chance of becoming even more popular among non-techies is laudable. Working to increase FOSS fragmentation, in my opinion, is not. I haven't seen one valid argument yet about what AOO brings to the table the LibreOffice does not. The "diversity" argument isn't valid here; this isn't like, for instance, The GIMP vs. Krita, where the two products are very, very different underneath, have different UIs, even different feature sets, different strengths and weaknesses, and different tasks at which each one is better-suited for. Instead, you have one product that's a fork of the other and has all the developers, and another one that is a dying husk. This resembles the X.org vs. XFree86 situation more than anything, and I would hope everyone here remembers how that one went. What you're doing here is basically like Dawes begging for more help for XFree86 even though everyone else is perfectly happy working on and using X.org, all the distros use it, and everyone's forgotten XFree86.


Ultimately, if you don't want to work on AOO, there's probably not much I can say to convince you. And that's fine. No hard feelings. So the rest of this really isn't even for you, it's for other people who are reading and who don't already have their minds made up:

Why would I want to waste my time working on a has-been office suite when I can work on the leading FOSS office suite, which is LibreOffice?

Because you like competition and you want to knock that "leading FOSS office suite" off its perch.

Why would I want to commit my time to a project that seems to only seems to exist because of a licensing dispute?

1. The license thing does matter, because it's easier for other projects to incorporate patches that are licensed under the ALv2. And even where that's not actually true, it is the perception. In either case, I'd posit that contributions to AOO stand to benefit everybody, whether they're using AOO, LO or something else.

2. The Apache Way. Some people will just find that they enjoy the style of collaboration that results from the way Apache projects operate. It won't be for everybody, and it's probably a minor point, but it matters to some people.

3. Everybody loves an underdog and here's your chance to be a hero. Contribute to other projects and you're "Just another person who contributed some patches". Help AOO (especially right now) and you get a chance to feel like you're really making a difference.

Why would I want to join a chronically understaffed project when I could instead join its competitor, which has far more development support?

You can probably have more impact working on AOO. There's more room to influence major technical decisions exactly because there aren't as many active developers. Want to radically re-invent how office suites work and create something truly novel? There's probably more opportunity to to that within the AOO project than others.

Why would I want to join a project which all major Linux distros have abandoned, when I could instead join its competitor, which all major Linux distros have switched to?

Meh... I don't see how that matters for anything. Why should someone else's decisions influence what you decide to work on?

I'm sorry, I haven't seen any good answers to these questions at all.

And truth be told, that's at least in part because we haven't spent a lot of time trying to come up with answers to them. We've been content to just accept that some people will want to work on AOO for their own reasons, and others will prefer to work on other projects for their reasons. We'll see if that changes in the future. I just started a thread on the AOO dev list trying to provoke some discussion around this very topic.

Working on a FOSS project that is popular and stands a chance of becoming even more popular among non-techies is laudable.

Yes, that's why we do it. :-)


>Because you like competition and you want to knock that "leading FOSS office suite" off its perch.

So, basically spite.

The real competitor is Microsoft Office; that's what all FOSS software is competing against in trying to get people comfortable with using FOSS software and stop being locked-into proprietary software and file formats. You're actually harming the FOSS cause by promoting factionalism and fragmentation over silly issues.

>3. Everybody loves an underdog and here's your chance to be a hero.

OO isn't the underdog, LibreOffice is. Microsoft Office is the reigning champion and bully. OO is just muddying the waters and sowing confusion.

I do appreciate the candor in your response though, but it completely validates all my suspicions about the motivations of the few people still hanging on, much like Mr. Dawes keeps hanging onto XFree86 and refusing to throw in the towel even though everyone has abandoned him.


So, basically spite.

Not at all. Where did you get that? It's just a basic human desire to compete and win. The same reason people play football, play chess, arm wrestle, compete in spelling bees, etc. And still, all of that said, for me personally the "competition" factor is one of the least important.

The real competitor is Microsoft Office; that's what all FOSS software is competing against in trying to get people comfortable with using FOSS software and stop being locked-into proprietary software and file formats.

Meh... I could care less about competing with MS Office. I don't even think about them at all these days. As long as free software and file formats are available to those who want them, I'm not all that concerned with trying to change the world and make it so that everybody is using F/OSS. That's a challenge I'm happy to leave to rms or whoever.

You're actually harming the FOSS cause by promoting factionalism and fragmentation over silly issues.

Yeah, yeah... same thing they said 16 years about about Gnome vs KDE. But yet there are still Gnome and KDE, and even more. There's always somebody crying "fragmentation" and pleading for the end to something: AOO, BSD, KDE, whatever it happens to be.

I do appreciate the candor in your response though, but it completely validates all my suspicions about the motivations of the few people still hanging on

Please don't make the mistake of assuming that my motives are the same as anyone else's. I certainly don't claim to speak for anybody else on the AOO team besides myself.


> The real competitor is Microsoft Office

Is it? I haven't used any desktop office suite in more than five years, at least. I would think the real competitor is Google Docs, and I don't see any current credible path forward that brings OO or Libre Office to the web.


> "I would think the real competitor is Google Docs"

If you're happy with GD, that's cool, but I personally don't see much reason to switch.

As for bringing open-source office suites online, it is happening:

https://wiki.documentfoundation.org/Development/LibreOffice_...


LO on the web is in progress. It'll have collaborative editing, the one killer feature of GDocs.

GDocs is a terrible word processor (with absolutely shocking OOXML conversion abilities; I don't believe this as a serious objection to LO any more) and a terrible spreadsheet. The one feature it has is live collaborative editing.

I am enormously pleased that LO 5.2 has working two-factor auth to GDrive, which means I can use a proper word processor or spreadsheet with it transparently. Even MS Office still doesn't talk transparently to GDrive out of the box.


I'd love to see LibreOffice or Apache OpenOffice compiled to the Web (eg. asm.js, WebAssembly)


> Because you like competition and you want to knock that "leading FOSS office suite" off its perch.

If I prefer competition among OSS office suites, and am unsatisfied with the leader, why would the competitor I choose be the one that's basically a trailing, less maintained for several years, version of the same codebase as the leader, rather than something that's both more genuinely different and has been more actively maintained, like Calligra?


Maybe you wouldn't. Maybe you just feel some comfort / familiarity with the ASF, who knows? I'll freely admit, all of this is pretty subjective and pretty personal.

My take is, as long as you're contributing to something OSS, you're doing something laudable. I don't care if you contribute to AOO, LO, Calligra, a new fork of KOffice, GNumeric, or something else.

Of course I'd like to see people chip in with AOO, but I can't sit here and tell you there are a lot of objective reasons to decide to do so. I just hope there's some non-zero number of people who do make that choice, regardless of their reasoning.


> Because you like competition and you want to knock that "leading FOSS office suite" off its perch.

There are so many huge game breaking features missing from the free software world - great 2d animation, great 3d (and 2d) cad, great sound editing, great? video editing (I personally love kdenlive, but know the complaints are endless), great Gimp? (I think Krita, again, is great in this space, but everyone wants Gimp to be something amazing vs photoshop). Even beyond just applicaiton suites, we have none of the cool "future" features proprietary platforms are pushing like voice recognition (Mycroft? I think it went semi-proprietary?) or good VR support to show off either.

It is a major problem in free software how people start out thinking that we need another text editor or file manager or music player to come show all these 10 year old projects how its done. We end up with dozens of lackluster showings in all these fields rather than one or two great options because few want to work on someone elses code and would rather "compete" instead.

But until the leadership of a project proves itself juvenile / dysfunctional enough not to work with new contributors like yourself (note, most new contributors are the problem though, when they run into issues, you should look at what you're doing wrong first), it seems like just a waste of everyones time to work on competing rather than just adding whatever cool feature you want to the market leader. We're supposed to be better than the mainstream market's preschool hide your toys from the other kids and never share mentality. We are supposed to be working together to improve software, and forking / competition should only be a last resort when leadership in the leading project is not inclusive enough of the community and needs to be overthrown.

So, on topic - what has the Document Foundation done to warrant toppling? I'd be happy to support the AOO crusade to take back the throne if they are actually doing something wrong.


> Because you like competition and you want to knock that "leading FOSS office suite" off its perch.

It is really sad but this looks to be what brings together those that remain on the dev mailing on Apache OpenOffice.

Is there anything else there other than hatred towards LibreOffice?


> Why should someone else's decisions influence what you decide to work on?

Because you want to do something impactful that others will use and get benefit from.


If you, as an AOO contributor, prefer the Apache license and you want to contribute to that project is fine, that's your free choice :)

What I'm worried about is the end user, the one who only knows "openoffice is the alternative to ms office", and is stuck with an unmaintained product.

A person I know told me a few months ago "Hey, openoffice don't render square roots in an acceptable way, should I go back to ms office?", and I had to explain he was using a mostly-abandoned software and he should switch to libreoffice because that bug was fixed months ago in there.

You're maintaining a software for end users, not for the HN population: they don't care about the Apache license or the GPL, but they care to have a working piece of software :)

I really hope the OOo trademark is handed to TDF, not because of license wars, but because that's the better thing for the users of AOO/LO.


Why do you contribute to AOO and not LO? On what basis should one decide to contribute to one or the other?

(I'll admit I am not very likely to contribute to either, but if I do I'll probably contribute to what my Linux distro uses)


Why do you contribute to AOO and not LO? On what basis should one decide to contribute to one or the other?

For me personally, it is a combination of preference for the Apache License, familiarity with the ASF (from being involved in other projects) and a general comfort level with how the ASF works. That and a little bit of wanting to support the underdog, and compete with the perceived front-runner.


> "That and a little bit of wanting to support the underdog, and compete with the perceived front-runner."

Dude, I get it, I really do, and if that's what you want to spend your free time on you should feel good about aiming to do the right thing.

However, I can't help but feel we're missing out the 800-pound gorilla in the corner of the room.

If you want to compete with the perceived front runner, there's only one office suite that is fit to have that title, and it's name is not LibreOffice. You know exactly what I'm talking about.

The gap between the real front runner and the open source alternatives is pretty sizeable. If you're serious about wanting to compete with the perceived front runner, why not look at a AOO/LO merger? Are the project governance approaches really that different? Are the goals really that different? Why not accelerate the growth of open-source office suites rather than having two projects competing for developers? The OpenOffice name still has value, perhaps the ASF approach to project governance will be preferred too, why not bring everyone in under the same banner?

The whole situation reminds me of the Amiga vs Atari wars of the late 80s/early 90s, competing over the same home computing niche whilst the PC slowly ate their lunch. Please don't let history repeat itself. Look beyond the open source world, your main competitor does not live within it.


If there was a viable path forward to a (re)-merger of AOO and LO, I'd be totally fine with it. But I've heard enough rhetoric from both sides that I've essentially quit believing such a thing could ever happen. But hey, miracles do happen now and again.

In the mean-time, MS Office isn't even part of my consciousness. But I'm a pretty rabid FOSS ideologue who has used Linux and FOSS office suites as his primary desktop environment since about 2002. So I'm a little out of the mainstream, you could say...


Thank you for being open to the idea of a merger, even if the possibility seems remote, and thank you for your work on open-source office suites (regardless of the project in question), they're a highly important part of the open source software ecosystem.

I'll leave with one more idea, just in case a full on merger doesn't work out, and that's what I'll call a soft merger. If a goal of modularity is pursued, there may be room for collaboration on shared components across projects. For example, the libraries for reading and writing to proprietary formats would be good candidates to be developed as modules, which could have benefits for more than just office suites. Just something to consider.


I appreciate your comment as a genuine expression of confidence and hope. However, it seems that this generic comment could reference an existing initiative that is already being used by several office suites http://www.documentliberation.org/

So it's not like there aren't already chances for the integration you propose, that span a broader section of the FOSS landscape than just LO/AOO. But ... You'll see that Document Liberation is spearheaded by LO/TDF and ignored by AOO/ASF.


"So it's not like there aren't already chances for the integration you propose, that span a broader section of the FOSS landscape than just LO/AOO. But ... You'll see that Document Liberation is spearheaded by LO/TDF and ignored by AOO/ASF."

Sorry, but that is incorrect. There is the Incubating ODF toolkit at Apache for example. Now it is true, of course, that the TDF is focused on ODF (in other words, that is it's stated goal), but Apache is hardly ignoring it. The mission statements of Apache and TDF are wildly different, but characterizing something as being "ignored" simply shows an ignorance on how the ASF operates.


At least some on both sides of the LibreOffice/OpenOffice split hate each other with the heat of 10,000 suns. There will be no reconciliation.


Why is that? What has fueled this divide?


I guess the people on the LO side, originally hated Oracle (who doesn't?) and what they did to OO. And now the hate for the OO developers is just a leftover feeling that they were supporting Oracle.

And I guess the OO people hate the LO people because they are butthurt that LO has all the developers. LO has a consistent release schedule while OO takes about a year if not longer to crap out a release. Not to mention, any good patch that is added to OO, is taken by the LO developers and added into LO. So no matter what, LO will always be ahead, with the same and more features.

But I am not a developer for either of these projects. This is just my outsider's perspective.


It was a german company, StarDivision, that initial created StarOffice. Sun Microsystems bought StarDivision and open-sourced StarOffice into "OpenOffice.org". All that, around 2000.

It appears that the original developers of StarOffice, who then moved onto OpenOffice.org, did not have a good understanding on how free/open-source development works. They were unwilling to open up the development to new people and often rejected even the smallest contributions. A well-known example was about source code comments in German. There were contributions to write them in English, however the developers were shooting down even such straightforward changes.

In addition, Sun Microsystems would still produce StarOffice, an installation package of OpenOffice.org with some proprietary components. The project managers (Sun Microsystems) at OpenOffice.org would put emphasis to the StarOffice features instead of increasing the community involvement.

Somewhere around 2006, IBM got a license from Sun to produce an office suite based on OpenOffice.org, called IBM Symphony.

After Sun Microsystems was bought by Oracle, the same situation continued for a bit until Oracle could not figure a way to monetize from OOo and pulled the resources.

The lure for an office suite is that Microsoft has MSOffice which is a cash cow. It makes a lot of money for them. Both Sun Microsystems and IBM wanted a piece of that money. The permissive licence would enable them to make money by packaging the open-source OOo with their own proprietary additions so that customers would be compelled to buy it.

What they did not understand was the an office suite is a very complex piece of software. Apart from developers, it requires additional skillsets to produce the final package.

Just like the Linux kernel is stuck to the GPLv2 licence, it makes it easier to get everyone to work together. IBM famously declared around 2000 that IBM loves Linux. They made and make lots of money on Linux, even if Linux does not have a permissive licence.

It is an issue of greed to go for a permissive licence. That is, you benefit from all the volunteer work and then you have the right to add your proprietary bits and pieces that differentiates you from the open-source version.

If IBM had the foresight, they should have accepted the way LibreOffice is taking over. Just like with Linux, IBM is using it to gain big contracts because they pay their developer to develop it and support their customers at the same time. IBM could have done the same with the office suite. But no, they wanted the whole pie and a few additional slices.


Looks like this well end up like X.Org/XFree86 (yes, http://www.xfree86.org/ still exists)

Edit: I should have reloaded before posting, the argument is much better made by Grishnakh above.


"Exists" is a strong word. The CVS link http://www.xfree86.org/cvs/ is a 404 (and, uh, is CVS), the FTP server isn't accepting anonymous logins, and the HTTP releases URL is giving the most amazing 404: http://ftp.xfree86.org/pub/XFree86/4.8.0/binaries/


How is their 404 page Yahoo? Do they use yahoo for hosting??


The discussion from a few days ago on this topic was pretty informative: https://news.ycombinator.com/item?id=12411747


It is a shame that the article is behind a paywall; it sounds as if it is a very well balanced article.

There seems to have been many things that people forgot about when following the previous thread (https://lwn.net/Articles/699047/ and https://news.ycombinator.com/item?id=12411747), which I hope to relate here:

1. The whole issue was an open and honest DISCUSSION. Many people took this as an indication that AOO was dead. I fear that my own response to Dennis' post on-list went a bit too far in reinforcing that (mis)belief but the lack of (perceived) developer energy was the basis for the whole discussion. Dennis did not say "AOO is dead, what should we do" but rather that the AOO community should discuss, as a contingency plan what a retirement would look like.

I may plan or discuss my funeral (or final wishes), but that does not mean I am dead or dying. :-)

But think about this: what other project would be so open and candid? Such openness, and the true appreciation that discussion be done in public is a core part of the Apache Way. It's also a core part of what open source work.

2. Because of this "publicity", the AOO has been overwhelmed by lots and lots of offers of support, which have been graciously and thankfully accepted.

3. People have also forgotten that choices, even in FOSS, are a Good Thing. LibreOffice is very successful, and they should be congratulated for their success. But certainly there is room for other players in this game, and certainly room for one (or more) that are under a permissive license. The thing is is that they don't have to be clones; they can have different audiences, different "missions" so-to-speak.

4. It is sad when we in the FOSS community degrade ourselves to simple, base license-wars. There are good, solid reasons for permissive, weak copyleft and strong copyleft, and I've contributed to them all.

IMO, what's next for Apache OpenOffice is what the Apache OpenOffice community decides; it sounds as if this whole kerflunkle has served as a kick-in-the-arse to the AOO team: they see how important AOO still is to numerous people, and they have loads of new volunteers offering to help. A 4.1.3 release is forthcoming so that is good news and a step in the right direction.


> It is a shame that the article is behind a paywall; it sounds as if it is a very well balanced article.

The link I posted is a "SubscriberLink", which bypasses the paywall; see https://lwn.net/op/FAQ.lwn#slinks for details. Even without it, the article will be outside the paywall in a little more than one week.


> But certainly there is room for other players in this game, and certainly room for one (or more) that are under a permissive license.

Why is "like $X but under a permissive license" something that people champion for? Are we not all in the same boat, trying to replace proprietary software so users can live in freedom? Permissive licenses simply do not help achieve this goal as effectively as copyleft licenses do.

If you want to argue that AOO will be technically superior or whatever, that's fine. But championing "we also allow proprietary forks" is not helping any users of free software.


Permissive licenses simply do not help achieve this goal as effectively as copyleft licenses do.

I use to share that view, but I'm no longer sure if this is true. Do you have evidence you could point to, or is this a instinctual opinion? I don't have evidence going the other way, but I don't think it's a clear cut truth.


> Do you have evidence you could point to, or is this a instinctual opinion?

The Sony PlayStation is based on a fork of FreeBSD. It is entirely proprietary. Lots of different embedded device also use BSD because they are permissively licensed. Anybody who gets such a device is screwed because they cannot modify it or even know what the device is doing. This is true for the majority of IoT devices (which is a serious problem going forward). To be fair, there aren't nearly enough GPL enforcement lawsuits as there should be (which is why I donate to the Software Freedom Conservancy) so it's not like the copyleft side is /much/ better but at least copyleft does help achieve the goal.

In addition, GPLv3's anti-tivoisation clauses are quite unique in the free software world and are specifically written to ensure that users have freedom over devices that try to undermine the spirit of the GPL. You won't find any permissive license that has such protections for your freedom (because it's not enforceable without making the license copyleft).

This argument doesn't hold for standards, generally. Because what matters for a standard is adoption. Sure, you're making it possible to have a proprietary fork of your reference implementation but at least users can be sure that the specification is actually used (and if the spec is Apache licensed then you're safe from patents which is something that can't be said of proprietary specifications). But Apache OpenOffice isn't a standard, so this is not really all that relevant.


> Why is "like $X but under a permissive license" something that people champion for? Are we not all in the same boat, trying to replace proprietary software so users can live in freedom? Permissive licenses simply do not help achieve this goal as effectively as copyleft licenses do.

In general, the people who champion for permissive licenses are not the same people who care at all about replacing proprietary software or "users living in freedom." The purpose of using a permissive license is indeed to aid proprietary software development, not hinder it - and I believe Apache OO was relicensed in this manner for the explicit purpose of being an upstream for proprietary end-user applications, rather than as an end-user application itself.


If the whole issue is "Copyleft Good; Permissive Bad" (as implied by "But championing "we also allow proprietary forks" is not helping any users of free software."), then there is nothing more to discuss.

There are places for permissive FOSS licenses, weak copyleft and strong copyleft. If you don't agree with that (and even the FSF does!), then no matter what, any discussion or debate is moot.


> There are places for permissive FOSS licenses, weak copyleft and strong copyleft. If you don't agree with that (and even the FSF does!), then no matter what, any discussion or debate is moot.

The two examples the FSF agrees with:

1. Base reference implementations of open standards. 2. Small snippets of code (smaller than 300 lines).

AOO is neither of these. Yes, I agree that permissive licenses have a use, but saying that "this $largeproject is under a permissive license!" as a positive thing doesn't sit right with me". It's like saying in a work report "amazing! we haven't had any murders in the office in the past 3 weeks." It's a correct statement, but the implication is that permissively licensing it has a benefit _for users_. This is simply untrue, as AOO is neither a base implementation of a standard nor is it a small piece of software.


> 2. Because of this "publicity", the AOO has been overwhelmed by lots and lots of offers of support, which have been graciously and thankfully accepted.

Is that really happening? I do not see significant traffic on the dev mailing list. In addition, the new "recruitement" mailing list does not appear to be active. Does it have any list archives so that we can establish whether there are loads of actual offers of support?


The archives for the "recruitment" list seem to be at https://mail-archives.apache.org/mod_mbox/openoffice-recruit...


There is no visible evidence for this claim, no.


AOO supporters want to get more developers working on it. But I think whether or not new developers decide to do that is going to depend greatly, for most of them, on whether or not they think AOO is going to succeed in the future.

So let me ask the AOO supporters, where do you predict that AOO will be in a year from now, and two years? And give some specifics, like code clean-up, infrastructure, security fixes, new features, and so on.

And finally, if a year from now, AOO is far behind what you have predicted, will you still keep telling people that AOO is going to be a success? And ditto for two years from now.


So let me ask the AOO supporters, where do you predict that AOO will be in a year from now, and two years? And give some specifics, like code clean-up, infrastructure, security fixes, new features, and so on.

I'll only address the things I have knowledge of / intend to work on.

I am working now on getting a dedicated server setup for running code analysis tools / fuzzing tools /etc. against the AOO code-base. My plan is to make this an automated process that runs frequently (although the exact details remain TBD) and delivers reports that the AOO developers can use to improve code quality and security. It is my hope that this kind of initiative can help make a major improvement in the quality of the releases.

From a feature standpoint, I have some ideas regarding integrating Calc more with the "Big Data" ecosystem, which - coincidentally - largely revolves around the ASF. A lot of the ideas aren't fully fleshed out yet, but I am picturing Calc as a front-end for working with various analytics processing engines. I think that at one time there was an Uno based R integration for Calc, so I'd also like to look at where that stands and see if that could use some work.

Also, I hope that we can dramatically improve the build system and associated documentation to make it markedly easier for new developers to grab the code, produce builds and get going. But I don't know that that's an area I personally will have much time to work on, plus there are other people working on that anyway.

I think we'd all like to see more frequent releases, so definitely put that down as a goal at least.

I have some other random ideas percolating in my head as well, but if I talk about them, somebody will interpret that as a commitment to deliver them, and since I probably can't justify making that kind of commitment right now, I'll stop here.

And finally, if a year from now, AOO is far behind what you have predicted, will you still keep telling people that AOO is going to be a success? And ditto for two years from now.

I don't tell people that AOO is or isn't going to be a "success". It already is a success by many standards, and not so much by others. I don't think looking at it in such binary terms is productive. As long as there is one person using the code, and one person hacking on the code (and the might be one and the same) then I'm all for it.


Even with the "Crazy Ideas" there currently is no reason to assume they would be blocked from being implemented, if only the resources for that were available. So, once the hard part of finding the resources for them is solved, why should one not put them on LO right away -- which is a much better maintained and stable base than the aging AOO?


https://news.ycombinator.com/item?id=2608700

Check those comments from 5 years ago. Some managed to predict the current mess.


Everything that happened was completely predicted, because it had just happened with Apache Harmony when IBM pulled out of that.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: