Weekly Edition
Return to the Development page
| |||||||||||||
Meeks on OpenOffice.org
Worth a read: this weblog
entry by Michael Meeks on Sun's management of OpenOffice.org.
"If OpenOffice was blessed (like other more sensibly structured
projects) with a large, diverse and healthy developer-base, then perhaps we
could expect to go around rejecting big chunks of code, offending
developers and driving away potential contributors. To do this solely in
order for Sun to retain total ownership of the code-base (and even loosely
coupled components) - seems rather a betrayal of it's self-appointed
stewardship role..." Many company-driven projects require transfer
of copyright ownership from contributors, so OOo is not the only place
issues like this will come up.
(See also: this discussion of ooo-build from
2004. ooo-build was not a full fork of OpenOffice.org then and does not appear to be one now.)
(Log in to post comments)
Meeks on OpenOffice.org Posted Oct 3, 2007 20:18 UTC (Wed) by madscientist (subscriber, #16861) [Link] Interesting suggestion by Michael. Sounds very similar to how OpenSSH is managed: the OpenBSD folks are not interested in keeping code in their environment which is needed for portability to OSs other than OpenBSD, so another team takes the OpenBSD version and adds in autoconfiscation, patches for support for GNU/Linux, Solaris, etc., etc. to create the "portable" version.
Maybe something similar will be done with OO.o? It seems a shame.
Meeks on OpenOffice.org Posted Oct 3, 2007 20:33 UTC (Wed) by ajross (subscriber, #4563) [Link] I'm not sure it's directly analogous. The Open BSD folks have some software that they wrote, that does what they want it to do, and that they distribute on their terms. They refuse portability work because it doesn't interest them: they don't actually want the feature, so there's no reason to expect them to distribute it.The contention here is that Sun is refusing a new feature that it wants. But they don't want to include the software from the original source because the author wants to distribute it under the LGPL*, and won't assign copyright to Sun. But they still want it, and in fact want it so much that they are redeveloping it from scratch. The fact that OO.o is itself distributed under the LGPL is, I think, what makes this especially offensive. Sun isn't willing to honor the spirit of its own license, and wants the ability to sell the community-contributed code under a restrictive license as part of Star Office.
Meeks on OpenOffice.org Posted Oct 3, 2007 20:51 UTC (Wed) by man_ls (subscriber, #15091) [Link] I'm not sure I can condemn Sun's behavior in this case; it is (mostly) their code, they provide a very important service to Linux distributions and want to finance it with a proprietary version. Could you elaborate further?
Meeks on OpenOffice.org Posted Oct 3, 2007 21:22 UTC (Wed) by AJWM (guest, #15888) [Link] Sun could still sell a proprietary version including the donated LGPL code (which is a separate module) without the copyright transfer. The LGPL permits this.
If Sun doesn't want to do that -- indeed, if they're willing to expend the resources to re-implement it just so that they have copyright control -- then it would seem that there's something else involved.
Meeks on OpenOffice.org Posted Oct 3, 2007 22:30 UTC (Wed) by jmorris42 (subscriber, #2203) [Link] > ..then it would seem that there's something else involved.
Yes, Sun doesn't want Star Office to lack features in OO.o. Duh. And if they would have to pay internal developers to reimplement a version for Star Office it seems rather obvious they would prefer that version go into OO.o to prevent the two codebases from drifting apart.
Bottom line, if you want your code in OO.o you have to assign copyright to Sun and allow the code to be sold as part of Star Office. But anyone should realize that up front and decide whether to participate or not.
Just like the FSF won't accept a patch to packages they maintain (coreutils, bash, etc.) without a copyright assignment. This is how they can move their code to GPL3 without any of the problems less strict projects would face. For an example see Linux, the kernel which will always be GPL2 even if the braintrust in charge eventually desperately/needed wanted to relicense.
Meeks on OpenOffice.org Posted Oct 4, 2007 2:34 UTC (Thu) by stevenj (guest, #421) [Link] Well, the FSF can relicense because it licenses all of its GPL software as version 2 or later (now, version 3 or later). They do copyright assignments because they believe it will make enforcement easier (if it ever comes to a legal challenge).
Meeks on OpenOffice.org Posted Oct 4, 2007 9:16 UTC (Thu) by kripkenstein (subscriber, #43281) [Link] > Yes, Sun doesn't want Star Office to lack features in OO.o. Duh.
Sun can sell StarOffice while including a few LGPL modules to which Sun does not own the copyrights. So this isn't the issue. I am not sure what Sun's problem is, but this isn't it.
Meeks on OpenOffice.org Posted Oct 4, 2007 12:57 UTC (Thu) by webmink (guest, #47180) [Link] Sun doesn't actually have a problem here - the OpenOffice.org project is just being consistent with the way it has worked for years. There is a place for free-standing modules that don't need a contribution agreement to Sun - the extensions repository - and if Kohei had packaged Solver for use there, none of this would be happening and every distribution based on OpenOffice.org could benefit from his fine work.
Instead, for reasons known to him and his boss at Novell, he decided instead to make a big fuss by claiming it should be core code and then refusing to follow the well-established process for upstreaming core code. That policy may be ripe for debate, but this issue is one Kohei has navigated into rather than around or away from. Since I am from Sun I suspect corporate politics, but your mileage may differ.
Meeks on OpenOffice.org Posted Oct 4, 2007 15:41 UTC (Thu) by k8to (subscriber, #15413) [Link] > Yes, Sun doesn't want Star Office to lack features in OO.o. Duh.
AJWM addressed this issue in the post you are replying to.
There's nothing that legally prevents Sun from distributing this module *as is* as part of Star Office. The LGPL lets you do that! So why are they being sticks-in-the-mud?
Meeks on OpenOffice.org Posted Oct 3, 2007 21:17 UTC (Wed) by daney (subscriber, #24551) [Link] The FSF requires a copyright assignment for contributions to its code. What are the moral and technical differences between what Sun wants, and what FSF does?
Meeks on OpenOffice.org Posted Oct 3, 2007 21:19 UTC (Wed) by AJWM (guest, #15888) [Link] The FSF doesn't sell proprietary-licensed versions of such contributed code.
Meeks on OpenOffice.org Posted Oct 3, 2007 21:47 UTC (Wed) by madscientist (subscriber, #16861) [Link] Right. The FSF's reason for wanting copyright control is so that they can more reliably defend the copyright in court. In the past it may have been so they could change the license if necessary, but most if not all FSF-owned code is licensed under "GPLvX or later" so they could just as easily relicense it by releasing a new version of the GPL. I guess having the copyright lets them change the license from "GPLv2 or later" to "GPLv3 or later" if they want, to encourage obsolescence of GPLv2.
Also, the FSF does give you back an unlimited license of your changes so although you are giving them ownership you can still take all your own changes (not any of the GPL'd code you didn't write of course) and put them into any proprietary, etc. program you like. I have no idea if Sun does the same thing or not.
At any rate, this is a moral difference. Other than the give-back clause (which Sun could have as well) I don't see any legal difference.
Old contributor agreement Posted Oct 3, 2007 21:53 UTC (Wed) by webmink (guest, #47180) [Link] I've responded at length on my blog so I'll not repeat it all here, but an issue that needs fixing is that the OO.o folks are using a very old version of the Sun contributor agreement that, among other things, doesn't have the give-back clause the current version includes.
Meeks on OpenOffice.org Posted Oct 4, 2007 13:44 UTC (Thu) by forthy (guest, #1525) [Link] The FSF has a very clear statement how new versions of the GPL will look like. They all will protect the "four freedoms". When you assign a copyright to them, they promise that they will release it only under a license which protects this four freedoms. And actually, the FSF doesn't relicense anything. The authors relicense.
FSF assignment Posted Oct 3, 2007 23:06 UTC (Wed) by michael-fig (guest, #47318) [Link] In fact, the FSF assigments require the FSF to make all publically available derived works free software (i.e. redistribution with source code available without royalty). I've excerpted the legal stuff on Kohei's blog (it'll be visible once it passes moderation). Michael.
Meeks on OpenOffice.org Posted Oct 3, 2007 21:50 UTC (Wed) by madscientist (subscriber, #16861) [Link] > The Open BSD folks have some software that they wrote, that does what they> want it to do, and that they distribute on their terms. They refuse > portability work because it doesn't interest them: they don't actually want > the feature, so there's no reason to expect them to distribute it.
Sure; I didn't mean the reasons were the same. I just meant that the end result could look the same: OO.o releasing their version of software, followed by a volunteer group adding extra enhancements that the OO.o folks didn't want or wouldn't take and releasing the version that most everyone actually uses.
Meeks on OpenOffice.org Posted Oct 3, 2007 21:52 UTC (Wed) by MattPerry (guest, #46341) [Link] > and won't assign copyright to Sun.
How is this different from the FSF requiring copyrights to be assigned to them for their GNU tools?
Meeks on OpenOffice.org Posted Oct 3, 2007 22:23 UTC (Wed) by johnkarp (subscriber, #39285) [Link] He states in a follow-up post that he is willing to contribute ownershipof his code to nonprofit organizations and consortiums. Unlike GNU, Sun is neither of those; they are a publically traded for-profit.
That means Sun's officers are legally obligated to make decisions that
Meeks on OpenOffice.org Posted Oct 3, 2007 22:24 UTC (Wed) by johnkarp (subscriber, #39285) [Link] Oops... by 'GNU' I mean 'FSF'.
Greed not essential Posted Oct 3, 2007 22:34 UTC (Wed) by webmink (guest, #47180) [Link] While that's true, Sun's strategy is to grow stockholder value by nurturing FOSS ecosystems so that its users buy systems including support subscriptions for FOSS. That means it's in Sun's interests to act in ways that make those ecosystems trust Sun. My strategy there is thus to promote software freedom in order to grow stockholder value. A duty to create stockholder value is not synonymous with a duty to do evil, whatever prior examples may exist.
Greed not essential Posted Oct 4, 2007 9:50 UTC (Thu) by gavinmc (guest, #46316) [Link] That may be Sun's current _strategy_, but you cannot assume it will always be. Arguably, SCO once had a similar strategy.
At the end of the day, Sun will always do what's best for its shareholders and will change strategy where the shareholders' interest dictates. That strategy may or may not involve "nurturing FOSS ecosystems so that its users buy systems including support subscriptions for FOSS".
Greed not essential Posted Oct 4, 2007 12:59 UTC (Thu) by webmink (guest, #47180) [Link] Naturally people worry about the future, but the original comment to which I was replying was worrying about the present.
Meeks on OpenOffice.org Posted Oct 3, 2007 22:34 UTC (Wed) by liljencrantz (subscriber, #28458) [Link] Depends on whether you trust the FSF or not. The FSF claims that it needs the copyright attribution so that they can legally defend the software they distribute from copyright (left?) infringements. Sun wants to do that as well, but also needs the attribution so they can make money by selling oo.o under a proprietary license.
If you trust the FSF to act in _your_ best interest for all of eternity, there is a relevant difference of intention which might very well be relevant to you, but if you feel thet there is a distinct possibility that the FSF might one day be taken over by people with goals that you don't agree with (or feel that is already the case), then there should be no relevant difference between Sun and the FSF in this regard.
It should be noted in this discussion that the FSF is making money from selling printed manuals, and their manuals are under a non-free license (Not even dual licensed), so it could definitely be argued that the FSF is doing something a ot more dubious than Sun is when it comes to manuals.
Meeks on OpenOffice.org Posted Oct 4, 2007 0:42 UTC (Thu) by TxtEdMacs (guest, #5983) [Link] RE: "It should be noted in this discussion that the FSF is making money from selling printed manuals, and their manuals are under a non-free license (Not even dual licensed) ... "
That is quite an assertion, can you back it up with a citation proving this is indeed an action of the FSF? Until I see proof, I remain highly dubious of your claim.
Meeks on OpenOffice.org Posted Oct 4, 2007 1:12 UTC (Thu) by madscientist (subscriber, #16861) [Link] There are a group of people who consider the FSF's GFDL to be non-free, because you can add "immutable sections" which cannot then be removed or modified from the documentation.
In particular, the Debian community has rejected this license as not free according to the Debian free software guidelines, and as such all GFDL-licensed documentation (that contains immutable sections) have been removed from Debian proper and pushed to the "non-free" repository.
For more info on this, see the Debian General Resolution on this subject: http://www.debian.org/vote/2006/vote_001
Meeks on OpenOffice.org Posted Oct 4, 2007 12:41 UTC (Thu) by TxtEdMacs (guest, #5983) [Link] Your comment does not address the words written. In essence, you are forking the discussion to a philosophical splitting of hairs about the freest license of all. Unless you can resolve with finality, your comment is misplaced.
Moreover, that is NOT what s/he said. Let the author reply and defend his/her words, that are in essence quite nasty.
Meeks on OpenOffice.org Posted Oct 4, 2007 13:32 UTC (Thu) by madscientist (subscriber, #16861) [Link] Certainly my comment does address it. You quoted a statement:
>> RE: "It should be noted in this discussion that the FSF is making money
> That is quite an assertion, can you back it up with a citation proving this
The quote YOU provided makes two points: that the FSF makes money selling printed manuals, and that manuals are under a non-free license. You asked for the assertion to be backed up. You did not specify which assertion you were disbelieving, so I assume, since you quoted both, that you disbelieved both. I felt that the former (that the FSF makes money selling printed manuals) was quite obvious and did not need to be commented on, so I addressed the latter (that manuals are under a non-free license).
I'm not saying that I agree or disagree with this assertion, just that there are a number of people who have looked into the issue and thought about it a lot; it's not just a few malcontents.
No one is talking about the "freest" license. Although the OP should have been clearer, they were talking about an objective definition of "free" as described by the Debian free software guidelines not a subjective philosophical hair-splitting.
Meeks on OpenOffice.org Posted Oct 4, 2007 19:54 UTC (Thu) by k8to (subscriber, #15413) [Link] Agreed, not hair-splitting. The GFDL with invariant sections does not guarantee the four freedoms. The current versions of GNU manuals all include invariant sections.
The deviation from fully four-freedoms free text is, contentwise, minor. For an organization as strongly ideological as the FSF though, it's quite surprising.
Meeks on OpenOffice.org Posted Oct 4, 2007 22:10 UTC (Thu) by liljencrantz (subscriber, #28458) [Link] I disagree with the assertion that the deviation from the four freedoms is minor, as they make it impossible to reuse the same documentation text in another manual, unless you are ok with having a large blurb on both the fron and backside of your manaul stating that your manual is a Gnu manual. What this all boils down to is that if anyone ever wishes to fork a Gnu project, they have to start by throwing out all the documentation. See the Xemacs example I gave in my post below for a realworld example of this problem.
Meeks on OpenOffice.org Posted Oct 5, 2007 5:39 UTC (Fri) by k8to (subscriber, #15413) [Link] I meant that the percentage of the text which deviates is minor. Eessential the covers are invariant. That's about it.
I don't make any claims about whether the implications are minor or the deviation from the freedoms are minor.
Meeks on OpenOffice.org Posted Oct 5, 2007 8:42 UTC (Fri) by liljencrantz (subscriber, #28458) [Link] Fair enough. Though that would be true as well if you took the GPL and added a 'you must sacrifice your firstborn to the flying spagetti monster to use this software'-clause.
Not that the invariant clause is even nearly as bad as mass murder of newborn babies, my point is only that half a sentence can easily be enough to move a license from free to non-free.
Meeks on OpenOffice.org Posted Oct 4, 2007 21:59 UTC (Thu) by liljencrantz (subscriber, #28458) [Link] I am guessing (correct me if I'm wrong) that the assertion that you feel I should prove with citations and such is the one about the FSF manuals being non-free, not the one about making money.
I would like to begin by stating that what I meant was not that the manuals where unavailable online. When I sad non-free I was referring to free as in freedom, as in the freedom to modify and redistribute the manuals. They are indeed available free of monetary charge online, and other people are free to print them, and even to charge money for printed copies. The problem with GFDL-licensed manuals are much more subtle than that. Sorry if this was the cause for the confusion.
I could simply ask you to read the conclusion of the Debian project had on the subject that another poster was kind enough to link to, as the theoretical arguments are in my opinion valid. But I will do you one better and give you a real world example of why the non-freeness of the GFDL is in my opinion a big mistake that I hope the FSF will admit to and correct one of these days.
You should probably start by reading the GFDL at http://www.gnu.org/licenses/fdl.html. More specifically, the parts about invariant sections are crucial.
As far as I know, all GNU manuals have invariant sections, including but not limited to a front cover text and a back cover text. This is the case for the Emacs manual. These make it impossible to reuse even smaller bits of a GNU manual in your own project unless you want to put a blurbs about this being a Gnu manual on your front and backpage. Thus you can't recombine and reuse documentation in any meaningful way. All the text is forever locked into the format of a manual for that specific software.
And this is not a mere theoretical problem. Richard Stallman refused to relicense even parts of the Emacs manual in such a way that the Xemacs project could use it because he has a strong personal dislike for the Xemacs project and the people involved in it, see this email for reference:
http://calypso.tux.org/pipermail/xemacs-beta/2004-Decembe...
RMS may have some good reason to dislike the Xemacs people, they may have been rude to him (In my opinion they where not rude or threatening on the mailing list, but that may refer to private conversations I am unaware of), but that is entirely beside the point. RMS willingly used the non-freeness of the GFDL in order to make it impossible to share and reuse the contents of the Emacs manual in any meaningul way outside the Emacs project, and the GFDL license let him do it. That makes it a bad license as well as a non-free one.
Meeks on OpenOffice.org Posted Oct 11, 2007 16:52 UTC (Thu) by TxtEdMacs (guest, #5983) [Link] It was my intention to reply to this post on the 5th, but other pressing matters took precedent. Indeed, I would have preferred to put this off further. I am setting aside a task to answer. if I delayed any longer there would be no purpose in writing any response.
I do not intend to take time to read what constitutes a really free system, partially due to the implicit constrains I alluded to above. However, the core reason is I tend to distrust some of the voices making those claims.
For now, let me state that in principle we may agree more than we disagree. Debian principles are, in their words, those nearest to aligning with my own. Moreover, pertaining to having doubts about RMS, I have many. I think he has an authoritarian bent I distrust, I think free software issues have become too personalized around his view. Moreover, I have used tools he rejects that at times are superior to his, <i>e.g.</i> xemacs.
I thought xemacs group wanted to be merged back into emacs proper. It is my understanding that RMS made a short sighted decision to reject their request. In the longer term his obstinance harms free software. But he was against selling documentation for free software <i>à la</i> many O'Reilly books. He obvious changed his view there. So perhaps there is hope he can be induced to become more reasonable.
Recently Moglen's suggestion that it was time for new blood leading FSF, is idea I can subscribe. RMS may not see that as a necessity. Thus, in the interim, I purposely say and write linux, never GNU/Linux.
Despite our agreements, there is too much in your initial comment and your response I cannot accept.
No you cannot now take the money assertion out. It was a major part of your statement. Moreover, I think its purpose was to mislead the reader.
Now on to my other points, I think many of the loudest voices for this abstract concept of freedom are the very ones that do a disservice to the real ideals of freedom. My suspicion too is that many are part of the problems I describe below.
Why are the loudest voices and quickest responders on the Debian help forums those that spew abuse? The example I know of personally was what my son encountered. He was helped but by others that did not have the need to demonstrate their inherent superiority. Who are these stalwart, pretenders for the ideals of freedom that is cloaked in abuse and nastiness?
These defenders of truth have won an odd victory, but not for freedom. My son underrates himself, because computers and surrounding issues are not his professional goals. He has run a Debian server, been involved in cryptography (security issues), he has written documentation, he has helped others on forums. Indeed he really set up the Debian desktop I used and got the dual monitor system to work. Nonetheless, I suspect the nastiness left some scar tissue.
A portion of the consequences may be that many users may be taught to appreciate the features in MS Vista OS, that he finds very well designed. Be assured he will transmit that message. I could say more, but I hear the voices " ... good riddance ... who needs someone like him? ... etc." However, I wonder how many of those voices have been invited into projects and actually contributed?
Then there were the screaming about how those millions that went into Ubuntu would have been better spent going directly to Debian. Anyone that believes that probably is tempted by the exclusive penny stock investments that arrive in their email boxes. Debian is a great server, but ...
Yes RMS may be a hypocrite about not allowing total access, however, in essence is that any less free than the attribution in the BSD licenses?
I would have given more credence to your citation had you used it upon your first writing, now it is just a bit too convenient with others making the same claims. Both a bit too late and too minimal for the initial assertion, sorry.
Meeks on OpenOffice.org Posted Oct 4, 2007 1:00 UTC (Thu) by TxtEdMacs (guest, #5983) [Link] Here is another reason I doubt your claims, from the Preamble to the license for emacs documentation there are these choice words:
"The purpose of this License is to make a manual, textbook, or other functional and useful document free in the sense of freedom: to assure everyone the effective freedom to copy and redistribute it, with or without modifying it, either commercially or noncommercially. Secondarily, this License preserves for the author and publisher a way to get credit for their work, while not being considered responsible for modifications made by others.
This License is a kind of copyleft, which means that derivative works of the document must themselves be free in the same sense. It complements the GNU General Public License, which is a copyleft license designed for free software."
I picked emacs, because of my familiarity with this tool. Are there others that you can cite that have non-free licensing? Making money is allowed. Using fees to support the foundation should be permitted when commercial ventures can use the same materials for that purpose. Hence, I suspect you may be repeating uncritically what you heard, or at least, that is my hope.
Meeks on OpenOffice.org Posted Oct 4, 2007 2:30 UTC (Thu) by njs (guest, #40338) [Link] Sure, that text is in the GFDL, but whether a license includes text that sort has little to do with whether it is, in fact, a free license. Unfortunately, the GFDL is (in the opinion of many people, including myself) a non-free license. See http://people.debian.org/~srivasta/Position_Statement.xhtml for details. That page doesn't actually give the official Debian position on the GFDL; the majority of developers voted to consider the GFDL free iff it the document in question contained no invariant sections. Having read the arguments on that page, I think the majority was wrong and the GFDL is always non-free, but whatever -- all FSF manuals *do* contain invariant sections, so...
However, this hardly makes the FSF and Sun situations equivalent. The GFDL stuff appears to be a weird lapse of sanity at the FSF; they've done some stupid things, but I can see where they're coming from given their principles, the lapses are minor in the grand scheme of things, and they have little practical effect. I don't think I understand the Sun/OO.org situation well enough to take a side, but the allegations are that Sun's commercial interests in out-and-out proprietary software are causing them to dramatically mistreat FOSS community members, causing clear and direct harm to both their volunteer developers and also the rest of us who benefit from using their software.
The GFDL and OO.org stuff have totally different issues at stake.
Meeks on OpenOffice.org Posted Oct 4, 2007 13:09 UTC (Thu) by TxtEdMacs (guest, #5983) [Link] There is one more issue, beyond the licensing: I was reading the documentation without being billed. You would not have gotten the impression that was even possible reading the words of the comment I was responding. Yes, the FSF, would like to get some cash in return for their efforts, but unlike a commercial site I was unhindered in digging into the documentation without payment.
I have not read enough to argue either cogently or with belief on the details of the licensing, hence, I know not what side I would take. In my defense, I would say I have too many other tasks that I am not attended to due in part to my lack of language skills. [I am a slow reader, etc.] As a result I pick up on hints on the veracity I assign to some voices that take part in the arguments. As a result I may conclude, perhaps too quickly, that from some no matter what ever the content I distrust the source.
As a example when a few Debian package managers were paid for their efforts, the volume of the outraged outcry embedded a bias against those that I thought attempted to be too pure. I think true freedom for some is taken as a license to run amoke. Take some of the Debian participants that populate some of their help forums, to a too common a degree many are there to belittle those that ask help. I have seen similar behaviour elsewhere where those that bemoaned the fact that so few were adopting lInux they were so quick to harass an individual that made an honest mistake and started wtih a distribution much too difficult for a beginner. Another RTFM, with absolutely no aid or even a link to that mythical font of knowledge. So perhaps your are right that this license is not as free as many think it should be, however, on average have they been a good, ethical player?
We all make mistakes, but I sense the words written and so far not supported are too extreme to jump to a quick defense. Let the author speak, please.
Meeks on OpenOffice.org Posted Oct 4, 2007 13:34 UTC (Thu) by madscientist (subscriber, #16861) [Link] > There is one more issue, beyond the licensing: I was reading the> documentation without being billed. You would not have gotten the > impression that was even possible reading the words of the comment I was > responding.
I don't think that's true. The comment very clearly states "printed manuals". There is no ambiguity and no suggestion that the manuals are not available online for free.
I think you are overreacting to a comment that is not, after all, very controversial.
FSF assignment Posted Oct 3, 2007 23:08 UTC (Wed) by michael-fig (guest, #47318) [Link] See my post above. Michael.
refusing the thing that it wants Posted Oct 4, 2007 0:35 UTC (Thu) by bkoz (guest, #4027) [Link] > The contention here is that Sun is refusing a new feature that it wants.
Or, that Sun refuses to link another LGPL library, even conditionally?
ldd /usr/lib/openoffice.org2.0/program/swriter.bin
Shows a couple already.
;)
Or, that Sun doesn't know how to conditionally add this feature? Seems unlikely.
On a more serious note, this seems like a very dysfunctional development community. Besides this, there are other issues: the build issues, some bugs unresolved for years, even when probed on a monthly basis? Why is everybody (IBM, Sun, openoffice.org, others) doing separate versions?
Perhaps a new start for all parties involved would be a good idea.
Intentional conflict? Posted Oct 6, 2007 1:12 UTC (Sat) by webmink (guest, #47180) [Link] There are two ways to make Solver available to everyone who uses every derivative of OpenOffice.org. One involves integrating it into the core OO.o code tree. The other is to package Solver as an add-on package.
The standard process in OpenOffice.org for integrating in the main code tree involves submitting the code under the terms of the SCA (Kohei has already signed it by the way so that's not the issue). For better or worse, it's been that way for years and expecting to vary it on a case-by-case basis is unreasonable.
To create an add-on package, there is no need to share copyright with Sun. Once he had changed his mind about the SCA, I have no idea why Kohei didn't choose to do this rather than knowingly heading into conflict with the established process.
As it is, his code is only available to Novell's downstreams since it is neither integrated nor an add-on. If I were a conspiracy theorist I'd have to guess his change of employment (and his employer's need to mend a damaged reputation) might have a part to play.
Meeks on OpenOffice.org Posted Oct 4, 2007 20:54 UTC (Thu) by vonbrand (subscriber, #4458) [Link] The fact that OO.o is itself distributed under the LGPL is, I think, what makes this especially offensive. Sun isn't willing to honor the spirit of its own license, and wants the ability to sell the community-contributed code under a restrictive license as part of Star Office. Sorry, I fail to see any difference to what the FSF is doing with the GNU programs...
Meeks on OpenOffice.org Posted Oct 4, 2007 0:26 UTC (Thu) by dkite (guest, #4577) [Link] Maybe Sun just has go get out their checkbook and start writing. 6figures will go a long way.
This isn't free software. This is business.
Derek
Meeks on OpenOffice.org Posted Oct 4, 2007 13:01 UTC (Thu) by webmink (guest, #47180) [Link] Actually, we already did that if you recall, back in 2000 when we rescued StarDivision and made all their code Free software in the first place.
Meeks on OpenOffice.org Posted Oct 4, 2007 22:16 UTC (Thu) by liljencrantz (subscriber, #28458) [Link] Which was a wonderful thing to do. Nicely done!
What is missing today is a more streamlined and open development process. If Sun doesn't help in making that happen, I'm pretty sure Oo.o will eventually fork, as has happened many times in the open source world (XFree86, Glibc6, egcs, Xemacs, etc.). But performing a fork and doing double development for a while before a 'winner' emerges means a huge amount of extra work with very little benefit, so I think everyone is hoping it doesn't come to that.
Change in progress Posted Oct 6, 2007 0:53 UTC (Sat) by webmink (guest, #47180) [Link] What is missing today is a more streamlined and open development process. I completely agree. I believe however that all the historic issues in this regard are in the process of being addressed. The guy who employs all the Sun engineers working on OO.o just blogged summarising the most recent changes for example, and prior to these there have been other changes that are starting to have an effect like the improvements in patch integration or the add-on repository which allows people not wanting to sign the SCA to participate freely.
But for me the real question is... Posted Oct 4, 2007 7:42 UTC (Thu) by m.alessandrini (guest, #36991) [Link] Who is buying StarOffice anyway??
But for me the real question is... Posted Oct 4, 2007 19:59 UTC (Thu) by k8to (subscriber, #15413) [Link] You know, I am curious. I wouldn't be surprised if there were areas where the sales were significant. I wouldn't be surprised if it was a fairly small sales volume. I guess I'm really not at all sure what StarOffice sales figures look like but I'd be interested in seeing them.
I'm not interested enough, though, to get off my ass and dig through sun's corporate site to try to find the answers.
XFree86 vs X.org anyone... Posted Oct 4, 2007 17:50 UTC (Thu) by arjan (subscriber, #36785) [Link] Reminds me quite a bit about the XFree86 thing... it was really hard to contribute (lots of hassle) and the code was quite hard to hack.
X.org made the codebase a lot more modular, and actively encouraged contributions... result is that X.org is moving ahead quite a lot faster and innovation has picked up bigtime. (and the original XFree86 project is effectively irrelevant nowadays)
Sounds like OpenOffice is undergoing a similar transition...
|
|||||||||||||