Novell's promotion of Go-OO has earned it a lot of attention recently because Novell ridicules OpenOffice.org and harms the brand [1, 2, 3, 4, 5, 6]. There is a lot more to OpenOffice.org than just the software; there are translations, ISVs, support firms and so on.
Quite a few people were unhappy with what Novell had done. And yes, Michael Meeks cannot magically disassociate himself from Novell and use the "I'm just a hacker" defence (ironically enough, Meeks has obtained software patents, which he filed with Novell). As critics of the cannibalistic approach taken by Novell, we decided to approach other people who are affected, merely reaching out for their opinion.
“...independent contributors outnumber Sun engineers by 10 to 1 inside the QA project.”Going more into details, Sun has the technical leadership in the OpenOffice.org project. I personally don't have a problem with that. What this means is that sometimes, patches are refused on purely technical merit. Whether those decisions are technically debatable might perhaps be the case sometimes. But generally speaking there is no problem. It is -- I believe -- quite easy to find both corporate and independent contributors who submitted patches, code or anything you can find in the way of contributions who were able to do so without any difficulty, provided they were following the guidelines and that their contributions were technically acceptable. That being said, OpenOffice.org has a very, very complex code base. This in turn causes a problem that is often overlooked: you need to study the code and the architecture, and thus devote a significant amount of your time doing so before efficiently contributing to OpenOffice.org. That's why we always find it hard to recruit engineering resources: you don't contribute code with your left foot when you're patching OpenOffice.org. But I agree that everything should be done in order to lower the barriers of participation to our project.
BN: What role does QA play in the lifecycle of OOo development?
CS: Since we're developing an end-user software suite we cannot tolerate leaving our software at a low level of quality. Of course, there are always bugs and we have ramped up our QA teams and resources significantly over time. QA gets to register the builds, test them at various levels according to the development, localization and QA processes. It also approves and decides whether the builds should be released or not. So to answer your question directly: QA and the QA project play a central role in our development and release process. By the way, it should perhaps be noted that independent contributors outnumber Sun engineers by 10 to 1 inside the QA project.
BN: Would you classify Go-OO as a branch or a fork?
CS: Both. I would have rather liked to answer: a branch, mostly, but some recent developments about Go-OO have obviously changed this situation. What should perhaps be reminded is that Go-OO is a Web site that hosts a concurrent build system to the one existing on the OpenOffice.org web site, called “ooo-build”. This build system has been around for ages. In fact, it's been used by many Linux distributions that found it more convenient for various reasons (basically, the builds were optimized for Linux).
“That furiously looks like someone is ready to fork by diverting and duplicating development resources from the original project.”At the same time, this build system was also used (even by Sun) to test new patches. The common conception here is that while the OpenOffice.org -Sun- build system (simply called “vanilla” for convenience purposes) is sometimes more conservative in that it does not integrate all the patches that fast. The reason for that is simple: QA. The ooo-build does not really test the patches it integrates, while the vanilla build system does. In short, the ooo-build is faster and easier to use, but produces builds that crash more often and have more bugs. You can experience that if you use any *Suse distribution or Ubuntu. Most of the other distributions have gradually stopped using it, precisely because of a certain lack of reliability that was experienced. The OpenOffice.org project now provides OpenOffice.org packages in .rpm, .deb and .tgz formats. We are also looking to improve our packaging on Linux: While straightforward anywhere else, the OpenOffice.org installation is still complex for an inexperienced end-user on Linux.
But the ooo-build has its own relevance and its own use. In this sense, it was a branch for a long time, and there was a widely-held view among the OpenOffice.org community that its existence was actually helpful.
The way you transition from a Web site with a separate build system to a fork is in fact quite easy. And what is only needed is the will for those Web site owners to decide to create a fork. At this stage, we can still keep a status quo, make sure we work out on any technical issues we can to have the two kinds of builds produced compatibly (that means mostly directly upgradeable from one another) and there, there will not be a fork, mainly a branch. Unfortunately go-oo has turned from an “annex” web site where several specific resources were available to a development platform parallel to what exists on OpenOffice.org: mailing lists, patches, builds, etc. That furiously looks like someone is ready to fork by diverting and duplicating development resources from the original project.
BN: Would you feel more comfortable if it was a project like Debian that deviated and managed a derivative of OOo?
CS: Anyone has the right to fork. It's Free and Open Source Software anyway. But I don't think a fork is a solution as it does all but adding up resources. Rather, it divides them, duplicates efforts and confuses users. There is worse stuff: in our case, I don't think that the forker would have the necessary resources to maintain the development efforts and have a coherent roadmap. At this stage, I would even be curious to know how bug squashing and issue management would be properly handled. As an example, I wonder how some of the large deployments of this particular flavour of OpenOffice.org would react if they were told that their own feedback was going to a fork of OpenOffice.org.
There is another couple of things that are of importance to me. Go-OO, if we are to believe its credentials, belongs to Novell. Now it is worth pointing out that at no point in the history of OpenOffice.org we ever got anything in the way of an official statement about Novell. That means that this is a silent fork. There is, if that is the intent of this company, no word, no declaration, nothing that basically says: “we feel we're doing a better job than you do” or “ we feel we're being unfairly treated”. That is something I find odd. The second element of importance is that we should get some sorts of governance structure and charter by Novell. You don't send your code in the wild and not asking yourself some questions. I know that OpenOffice.org was fiercely criticized by some people employed by Novell for having a copyright assignment, something Novell often demands in its own sponsored projects. But this legal vagueness of sorts is a bit odd: whom does your builds belong to? What happens in case of a legal problem? Is there a code steward? You don't need to be a consultant to ask those questions. And so far we have no answer.
BN: What role has the Novell-implemented OOXML translator played in allowing Microsoft's plot against ODF to carry on?
CS: Common work on OOXML and a translator was part of the Novell and MS agreement, as far as we know. Having played a role in the OOXML standardization “adventure”, Novell was being constantly taken as an example of “another open source implementation” of OOXML. Sometimes, as it was the case in Mexico we had Novell employees, such as Miguel de Icaza, sitting on the Mexican standards organization and strongly advocating for OOXML to be standardized. To me it looks like Novell has been vassalized and under the influence of Microsoft to the point where they had to defend the indefensible. Now, I was not born yesterday, and I know that in theory as well as in practice, corporations' primary role is to generate revenue. Hence you will find several corporations out there who will help FOSS with the right hand and promote the exact opposite with the left. Novell strikes me as different: it blurs the lines, puts a little bit of this in a little bit of that, calls a cat a dog and delivers software that is open source with conditions.
BN: What role, if any, do you believe Novell/Microsoft patents play here? What about Sun?
CS: It's very hard to tell. My personal view is that Microsoft does not have many patents and that most them are low quality assets. In short, when Microsoft makes claims about owning some significant amount of IP inside Linux for instance, it spreads FUD, and does just this. Anything further directly coming from Redmond would be very unlikely, because they have nothing. In short, it's “all hat no cattle” as they say in Texas. But they keep on applying pressure and make extravagant claims about their supposed ownership of every bit of open source code out there. I am in favour of full disclosure. Open Source code is, well, open source. It's out there. Anyone can grab it, freely modify and redistribute it. Proprietary code? I'm sure we would find some code blurbs that could turn out to be funnier than Easter eggs.
I have read, reread, and read again the Novell/Microsoft agreement. I think it's not clear whether this is an outright violation of the GPL in spirit or a legal flaw that has been exploited in it. But it surely changed the strategy of Novell in a way that poses a certain number of threats to FOSS users. It is also easy to notice that Novell's behaviour changed inside the OpenOffice.org community right after that agreement.
BN: Going forward, how do you suggest that the projects target their main competitor, Microsoft Office, rather than one another?
CS: First, remember that Novell acquired both Suse and Ximian. The Ximian team is still working inside Novell, and it looks like the Ximian business model got ultimately translated inside Novell's own strategy. Basically, when it comes to its open source offerings, Novell implements the Ximian strategy of taking the code, branching it, repackaging it and generating revenue from it. The way Ximian was doing it was a bit problematic, as it was not really beneficial to the communities it was deriving the code from and the value proposition to their customers wasn't clear either. I guess it's not my business, but such a mindset has partly led us to where we are today.
“We want to take OpenOffice.org to the next level, because we don't use office suites the same way were using them five years ago.”At this stage, I don't see any plans -nor any relevance- for the OpenOffice.org project to target go-oo. It just doesn't make any sense: what would be talking about? Different patches? I don't think the market even cares about that, I don't think it's even an audible message. I know that some people send messages out there, “my build is better than yours, I don't like your community”, but these same people should think: does it really benefit customers?
In regards to Microsoft Office, which is the true competitor to OpenOffice.org, our value proposition is clear: we are a full-featured office suite that brings its users the benefits of true open standards, quality, stability and Free Software. We want to take OpenOffice.org to the next level, because we don't use office suites the same way were using them five years ago. So we will increasingly interact with the Internet and on an online level, becoming the hub for creative writing, design and office work for everyone. That's what we stand for, and we will remain true to our mission and to our soul. ⬆
Comments
AlexH
2009-01-09 20:22:11
Baby In The Bath Water
2009-01-09 21:52:21
That makes all the difference in the world.
Note: this comment was posted from Novell's headquarters.
Roy Schestowitz
2009-01-09 22:00:14
Baby In The Bath Water
2009-01-09 22:11:33
And that's exactly what you did.
Note: this comment was posted from Novell's headquarters.
Jose_X
2009-01-09 22:28:22
Would Roy be the best person to draw up the questions? Why doesn't a go-ooo supporter give it a shot?
Perhaps it would be a good idea if BN would carry out that interview, but this may open up to some criticism if the questions are viewed as unfair.
BTW, haven't various go-ooo devs (or at least one) done interviews already which have been featured here? What is wrong with those?
***
On the meaning of "fork":
[blog piece] >> Unfortunately go-oo has turned from an “annex” web site where several specific resources were available to a development platform parallel to what exists on OpenOffice.org: mailing lists, patches, builds, etc. That furiously looks like someone is ready to fork by diverting and duplicating development resources from the original project.
basic wikipedia def: >> In software engineering, a project fork happens when developers take a copy of source code from one software package and start independent development on it, creating a distinct piece of software.
Wikipedia further defines a "branch" as a common term for *a type of* fork that is related to the main code base in various ways (eg, branches are forks that can be used for versions of main).
CHS appears to *differentiate* between branches and forks: you either have one or the other but not both. The rough criteria he uses to determine whether something that looks like a branch might instead be a fork is based on the degree of duplication of developer resources for this branch/fork wrt the original code base.
In sum, wikipedia defines "branch" as a flavor of "fork", and defines a "fork" as an independent forking from a given code base. Meanwhile, CHS and many others appear to define "fork" as a forking from a given code base, and where developer resources have been duplicated, and where the forked code base has diverged enough from the main code that the original community stops worrying about coding in a way that is consistent with the new code base. CHS defines a branch as a forking that otherwise remains "near" the main code base.
I prefer the wikipedia definition because it seems clearer to refer to a forking as a "fork" and then use more description to describe the fork further if desired. I do recognize the usage "that is a fork" to mean what CHS means, but I find it a vague expression/notion and do prefer to still have a name for the clean concept of any forking (call any "forking", naturally enough, a "fork").
Roy Schestowitz
2009-01-09 22:33:40
RLillySr
2009-01-09 22:34:52
dont know who you are, other than reading these threads, you can get annoying and with statements like that, pathetic
Miguel de Icaza
2009-01-09 22:40:39
Charles is misrepresenting my participation in the Mexican standards group. I was an invited expert on the subject on the deliberations for OOXML as part of Linux AC (the civil association that promotes Linux in Mexico) just like a bunch of other people were.
I do not sit in the standards group, I do not have a vote, and did not have a vote. I participated in one phone call where I explained my position to the people in the committee.
Miguel.
Roy Schestowitz
2009-01-09 22:45:49
Why do you defend Microsoft against The European Commission?
What do you think of Microsoft's anti-Linux tactics, as last published this afternoon?
Jose_X
2009-01-09 22:51:42
Were you replying to me? I presume then that you are using your definition of "fork".
I'm not interested in changing how you use that word, but understand that what you call "a fork, no more no less" ..I call.. "a fork that is being maintained independently from the original code base, as opposed to a fork that isn't, for example, like a branch."
So wrt go-ooo, I would say that it seems maybe go-ooo is being taken in a direction guided by a community different than the oo.o community. Call it a rose or a fork or a foo, I don't care.
I don't follow go-ooo except through what I hear on BN/LT/etc, so I don't really know. What I know is that I don't want dotnet hooks near me if possible because I don't want to facilitate or encourage the growth of dotnet. If I cared that much about similar technology, I'd petition to have Java (or something else) changed to be more like X or like Y [and there likely are ways to improve Java, perhaps more in line with dotnet.. and vice-versa perhaps as well].
Roy Schestowitz
2009-01-09 22:54:56
I was referring to no definitions. I said what I was told (based on another person's definition).
saulgoode
2009-01-09 23:10:46
I would have liked to seen the copyright assignment issue more thoroughly addressed -- OO.o is likely foregoing some significant contributions with their stance -- even if just to the point of future plans for resolving the problem. Not that Mr Schultz is necessarily the one to address it, but he does seem somewhat aware of the administrative structure of the project.
I particularly appreciated Mr Schultz's distinction when commenting about "corporations out there who will help FOSS with the right hand and promote the exact opposite with the left". Free Software is a complex web of alternative business approaches and varying degrees of participation. Novell makes some significant contributions to Free Software; however, their efforts to leverage patent indemnities and to undermine the GPL are nonetheless deserving of the criticism they have received.
@Jose_X, I disagree with regard to branching versus forking. Mr Schultz's interpretation is much more consistent with the employment of those terms whereby a particular project may offer multiple branches within the same basic infrastructure (e.g., CVS repository, bug report mechanism, language translation, etc).
Jose_X
2009-01-09 23:12:01
Then, if nothing more, my last comment served as extra clarification to the earlier longer reply.
Some call go-ooo a fork, and some call it a foo I could imagine.
Roy Schestowitz
2009-01-09 23:14:13
saulgoode
2009-01-09 23:15:50
AlexH
2009-01-09 23:24:13
Jose_X
2009-01-09 23:26:51
His branch and fork definitions may be more consistent with and/or shared as definitions by a majority of the people, at least under certain context. I am not disputing that.
I am adding that Wikipedia has a set of definitions that is different but which is itself consistent and which I prefer to use many times because I think it makes sense to have a term for referring to an arbitrary (eg, private) copy/change done to any code base.
If I copy your project and start playing with it on my PC, would you consider it a fork.. a branch.. ?
I'd consider it a "fork" quite simply ..no matter what my future intentions were; no matter what opinion you had formed about my "derived" project; no matter how large of a diff I had developed to date; no matter how I was managing the code "repository"; no matter how many people were helping me and how we were communicating with each other as developers on this new code base; etc.
In any case, we can use many different words. The important point I'm trying to bring up is that some people use "fork" for what I am stating (and which coincides with the Wikipedia definition) and that is the simple notion that someone has copied code and has begun to make changes.
Roy Schestowitz
2009-01-09 23:32:50
Jose_X
2009-01-09 23:52:14
Jose_X
2009-01-09 23:55:33
Maybe gather/archive the answers from anyone anywhere interested in replying.
Roy Schestowitz
2009-01-09 23:57:50
Victor Soliz
2009-01-10 00:53:18
You know, You Novell guys sometimes do the sins you blame Roy of. Noticed how you are blaming Roy of claiming that he represents the majority? I don't see such claim, in fact, your comment was the first appearance of the word majority. Time to panic?
Victor Soliz
2009-01-10 00:55:41
Roy Schestowitz
2009-01-10 01:01:45
Needs Sunlight
2009-01-10 17:57:42
I have a lot of respect for Charles' work and independence. He was a good choice for a neutral, outside view. I'm disappointed in the whining and attacks subsequently coming from DeIcaza and the other microsofters.
Roy Schestowitz
2009-01-10 20:17:06
AlexH
2009-01-10 20:22:43
Roy Schestowitz
2009-01-10 20:23:45
Roy Schestowitz
2009-01-10 20:25:28
http://blogs.sun.com/webmink/entry/thursday_s_links_january_8th
Dan O'Brian
2009-01-10 21:04:18
Roy Schestowitz
2009-01-10 21:07:00
Dan O'Brian
2009-01-10 21:16:54
He criticizes both Novell and Sun in his article, concluding that both have good and bad points.
Andre
2009-01-13 11:41:25
The third question is: How to make OpenOffice.org independent from dominant vendors?
Maybe a bailout plan is worth to get considered. A proposal would be to ask for a one billion EUR European FLOSS Fund which aims to lower dependencies of European enterprises and citizens on Microsoft Office and strategically funds alternatives as Openoffice.org which promote competition. This would be a minor lobbying effort and feasible in the current environment. If there is political agreement in Europe to become software-independent the means would be found, e.g. a contract with the French military etc.
I mean, we accepted oil and gas to be strategic resources and supprt our military to deal with these dependencies. Why overlook the software related dependencies of our economy which enable a vendor to milk our economy, and undermine and endanger European democratic decision making procedures?
You just need a sponsor for an anti-lockin and Eurofloss campaign.
Alberto Barrionuevo
2009-01-13 16:11:45
in the Mexican standardization committee you represented the interests of Novell and Microsoft: in other words, the approval of the DIS 29500 (OOXML) as in the precarious status that it was at the time. It was, and still is, totally biased and discriminatory to the interests of the FLOSS community and Microsoft/Novell competitors, so not fulfilling by far indeed the neutrality that you must expect from an international standard.
Apart from that, you can dress your participation in the committee as you like. The fact that you had voting rights or not has little value, since everybody could vote if s/he would wanted (actually, it was the way used to hijack the final result with several hundred votes coming up and down from everywhere with many ballot recounts and recounts always getting out different "final" results... until extremely late in the night).
So at the end, only your words were the fact to take in account to judge your participation. And, so sorry, but they spoke enough about who was your master voice at that time.
Un saludo, //Alberto Barrionuevo FFII Vice President at the time NoOOXML.org campaign leader for LatAm
Roy Schestowitz
2009-01-13 16:39:05
"Get a load of this, will you? Participants at the BRM were told that only technical issues can be addressed. Bill Gates, however, reportedly contacted the president of Mexico to try to influence Mexico to accept OOXML. Unless, of course, they were talking about representing dates in Excel or other technical issues. Hardy har. Like, totally. I’m sure."
http://www.groklaw.net/article.php?story=20080327104739103
Scott
2009-01-14 00:54:23
Andre
2009-01-14 17:16:47
It is well documented that Novell is actually contractually obliged to support Open XML. The company became a drone. Roy is very emotional and attached about this but it also makes way for Microsoft technology on other platforms. Nothing wrong to buy companies or make business deals as such. My personal opinion with a grain of salt.
So it is two levels of questionable standard committee behaviour: 1. Objections are technically valid and you vote they are not and cover them (unprofessional attitude) 2. You serve an agenda but fake a community guy and accuse others
My assumption is that there are also objective and universal criteria that determine standard quality. You can even argue that Open XML is good enough as a standard because there are worse standards, compare the Open XML specification with other undocumented parts of IT infrastructure, bug level of productive software or websites and you might get another perspective. We both had the controversy before we started with the disapproval campaign.
It is also acceptable for others to take commercial avantage for unprofessional actions. It is a matter of compensation for you as you risk your reputation. When your company is contracted to do what you do. Roy will disagree and has every right to think Novell does 'evil'. His opinion.
It is the same with an army. You may personally politically disagree with the invasion of Iraq but as a commander of troops you have to carry out your mission and likewise order your troops to carry it out. Roy would then so to speak argue for conscientious objection to the mission. No one should expect that from a soldier who made the decision for this job. Of course your rank makes a difference in terms of responsibility. And as a soldier don't take it personal when you get shot. And mercanaries soldiers who talk about their personal convictions deserve medical treatment.
*But* people should not make their defense against standard war critics or opponents let them sacrifice personal dignity and respect:
"Charles is misrepresenting my participation in the Mexican standards group. I was an invited expert on the subject on the deliberations for OOXML as part of Linux AC (the civil association that promotes Linux in Mexico) just like a bunch of other people were."
We know that the situation was different in Mexico and there is no benefit to play games. Probably Charles misrepresented the Ximian business model, it is a bit polemic and sticks, but there should not be any desire to see others proof the "just an invited Linux expert" story wrong.
I don't really see an advantage to let Novell become a next Caldera. Fair enough it does not seem to be the objective behind the Novell investment.
-- André
When I did lobbying in Parliament I often used to get "invited" by MPs. That is the process to enter the building if you don't have a permanent badge (which makes MPs believe you are a professional lobby gun, has also sometimes disadvantages).
Roy Schestowitz
2009-01-14 17:24:32
The "above my pay grade" remark springs to mind.