Calling Ubuntu, Fedora and OpenMoko to Resolve Mono Problem, Not Ignore It
- Dr. Roy Schestowitz
- 2008-11-18 18:48:49 UTC
- Modified: 2008-11-19 11:56:18 UTC
Sweeping one's discomfort under the rug is only a temporarily solution
SOME short while ago we wrote about
the problem with Mono. We ought to have elaborated.
Ubuntu Openness Disputed
It is natural for a Free software distribution to accommodate open discussions, but
Mono 'guards' in Ubuntu Forums are trying to defend Mono because they have vested interests. They do Mono stuff for a living, so it's in their interests to have it spread inside Ubuntu, regardless of the consequences. Here is a discussion about Mono
being closed.
Thread closed, as it's been hijacked by a running spat from another closed thread.
Mark Shuttleworth's
stance on Mono, which
already fills Ubuntu, is not helpful and one
possible solution was proposed just a few hours ago.
When Java was not Open Source, it was not installed by default. Wine is not installed by default. Mono is installed by default, if we want it or not. This needs to change. If you want to build or run mono applications, then you can install it - it is your computer and we will not stop you.
Keeping Mono Away from the Fedora
Looking back at Fedora's decision-making, here is what
one finds.
The reasons Red Hat decided to go ahead [with Mono] have since been revealed. It's because of the Open Invention Network. http://gregdek.livejournal.com/4008.html
We also
wrote about this back in June. Additionally, the OIN is only protecting
the Linux kernel GNU/Linux and the approach of the OIN is
not terribly helpful.
Keeping Mono from the Moko
As we revealed the other day,
OpenMoko had come under patent attack by a 'front' of Philips. This ought to teach people a lesson about the use Ogg, but also troubling is this new observation that
OpenMoko is becoming OpenMONO because of Mono enthusiasts.
OpenMoko is a open-source development platform for cell phones. OpenMoko sells the hardware and offers a Linux-based operating system to run on top of it. OpenMoko is known amongst .NET developers as being the first phone to support the Mono runtime.
There are some more details about the patent dispute
in Heise.
The Openmoko project temporarily removed all firmware versions from its download section last weekend and replaced them with different ones this morning. Openmoko is developing a mobile Linux distribution and sells the open Freerunner Linux smartphone. The reason for removal of firmware appears to be a dispute with patent holding company Sisvel.
It's worth remembering that there were
attempts to contaminate Android with Mono (courtesy of Mr. de Icaza himself) and a Windows ISV
tried to contaminate LiMo as well, using Mono.
⬆
"There is a substantive effort in open source to bring such an implementation of .Net to market, known as Mono and being driven by Novell, and one of the attributes of the agreement we made with Novell is that the intellectual property associated with that is available to Novell customers."
--Bob Muglia, Microsoft
Comments
Needs Sunlight
2008-11-18 19:58:26
Joe
2008-11-18 21:22:14
Needless to say I learned the
sudo apt-get remove --purge mono-common libmono0
mantra in Ubuntu and everything feels better now.
Roy Schestowitz
2008-11-18 22:29:55
jo Shields
2008-11-19 00:18:46
But, much as it may pain you, they aren't lawyers and have no ineterst in discussions of phantom patents - only pure technical merit (i.e. which app is "better" overall) will be considered.
As you well know, the preferred programming language at Canonical is Python. But, as with Fedora, if the best app for a task means pulling in some stack, then fine. There is no "reward" for Mono, there's "reward" for superiority user experience.
Roy Schestowitz
2008-11-19 00:23:14
How can you continue to ignore the obvious pitfalls? I would be genuinely interested in hearing your response to this.
Best wishes.
jo Shields
2008-11-19 00:42:56
Roy Schestowitz
2008-11-19 00:48:42
jo Shields
2008-11-19 00:56:49
How, specifically, does years of support for a patent-hording proprietary software company like The Mathworks, when free alternatives to their proprietary MATLAB exist (such as GNU Octave), square with your "down with proprietary companies with patents" image?
Roy Schestowitz
2008-11-19 01:09:29
Remember RMS and the "Java trap" warning? Let's harness Java now.
By the way, Mathworks advised me to write a MATLAB book when I was #1 contributor and they also tried to hire me with a 6-figure wage. I declined both invitations.
jo Shields
2008-11-19 02:20:18
So, short version is "I'd rather not, but I'm pragmatic, and I want a degree"
That's fair enough. I've dealt with more obnoxious companies (I'd name names, but I'd be in violation of my license for mentioning them in a bad light or suggesting their Free competitor is better, and that would mean severe repercussions for my institution. Go figure.)
So, a reply to your little link. Section by section then:
>Mono includes: (1) the ISO standard parts, including C# and CLI, >and also (2) the Microsoft proprietary parts, including >Windows.Forms and ASP.NET and ADO.NET. > http://www.mono-project.com/WinForms > http://www.mono-project.com/ASP.NET > http://www.mono-project.com/ADO.NET
Correct.
>Mono is written by Novell.
Mono is written by a number of contributors, and sponsored by the current owners of original authors/sponsors Ximian, which would be Novell. It's not out of the ordinary for Free Software packages to have corporate sponsors funding full-time development.
>Novell has a patent deal with Microsoft, so that Novell has a >license from Microsoft to write these non-free parts of Mono, and >to include them in SLED.
No, technically, Novell have a patent deal with Microsoft, so Novell customers who 'violate' any undisclosed Microsoft patents in a specific list don't need to fear patent lawsuits (in countries where software patents are applicable). They've also agreed to collaborate on few projects such as Samba.
The patent deal is pretty obtuse legalese, but the following can be extracted: * Novell has no patent protection from Microsoft, Microsoft have no patent protection from Novell. This was the way the deal "worked around" GPLv2 - but means either party can still file patent troll suits against each other. * The list of what's "protected" is messy, but can be summed up as the following: everything is protected, except things which aren't. "Clone products" are not protected, unless they are an implementation of a published spec - or covered by a different agreement (such as OSP). Implementations of part or all of the other party's product's API, where it isn't given specific protection e.g. by being based on a published standard or OSP, is specifically NOT protected. The exception, however, is that if the product in question was already being shipped on the date the deal came into force, then it's grandfathered and counted as fine. Unless the product is one of Wine, OpenXchange, StarOffice, OpenOffice. Confused yet?
>They are indeed open source,
Free Software.
>but they are not licensed by Microsoft to run anywhere but in SLED.
Careful with terminology. Microsoft can't issue "Licenses" to Mono any more than to NTFS-3G (included by default in some distros, too). They can issue patent covenants and grants of course. If you believe in that kind of thing.
Strictly speaking, they're not *explictly* patent-granted anywhere in the published patent deal legalese - any moreso than other component of SLE which meets the criteria not to be excluded from the patent deal. I suppose the significance for you here is that whilst OOo is explicitly named as not being grandfathered out of trouble, Mono is not, so it's implicitly being counted as fine and protected. How they then turn around and name OOo as covered in the "open letter", I have no idea.
Adding to further confusion, however: "the exclusion of such products from such subsection (i) is without implication as to (and shall not affect the determination of) whether such products (or any features or functionality thereof) are Clone Products. Further, the Parties agree that (A) no inference shall be drawn from the reference to the above products in this subsection as to whether such products are Clone Products and (B) this subsection and subsection (i) shall not be admitted or referred to in evidence in any dispute regarding whether any of the products referred to in this subsection is a Clone Product."
Short version: the whole deal is a plate of bullshit and chips.
>Not in OpenSuSe, not in Ubuntu, not in Fedora, not in Debian, not >in Slackware, not in Gentoo, not anywhere but SLED.
Well, sure. But I don't have a license to NTFS-3G. Or to the WMV/WMA playback in Totem. Or the page up/page down keys that work in Firefox.
And you're assuming, of course, that patents are explicitly violated. Patents, even software patents (garbage as they are) need to specify an actual process. Even when implementing something over which patents are claimed, there are very clear-cut rules on what can and can't be claimed (hence the Freetype people escaping the wrath of Apple's patent lawyers, via simple workarounds to the very specific patent clauses).
Even if I install System.Data.dll on my computer, and the API is compatible with Microsoft's, there is no guarantee without extended legal analysis that I'm violating any as-yet undisclosed patents on the implementation thereof.
>When you install Mono 2 on any Linux system, you are installing >software which includes Microso[f]t proprietary technologies without >having a license from Microsoft to do so (unless you run SLED).
Well, this is where it gets more fuzzy. For those not on RPM-based systems, you don't "just" install Mono - the packages are highly segmented, and don't neccessarily do so (assuming you accept, which I know you don't Roy but I think this article's author did) that the ECMA sections of CLR are at least "fairly safe". Enough Mono to run "hello world" is not enough Mono to *potentially* violate specific patents on ASP.NET. Nothing in Tomboy's dependency chain is. F-Spot pulls in System.Data.dll (ADO.NET) as a requirement of the SQLite library it uses.
If you accept that the ECMA sections are safe, though, then Tomboy is safe on Debian/Ubuntu systems. I know you don't accept that, but the wording of the quoted article suggests its author does (perhaps as a suspension-of-disbelief device. I have no idea)
>What is worse, if you use Mono to port to Linux programs >originally written in .NET for Windows, then any such ported >programs on your Linux system will include and rely upon the >unlicensed Mono libraries on your system.
Depending on your degree of porting, and assuming you believe that ADO.NET/WinForms etc in Mono are implementational violations of specific patents, then that could be the case, yes.
>What exactly is Mono all about? I think this page sums it up nicely: >http://www.mono-project.com/Guide:_Porting_Winforms_Applications
Well, no. That's actually the intention of the DotGNU Portable.NET project.
I'll quote a little from the VERY early days of Mono (Feb 2002) here, source is http://www.mono-project.com/Mailpost:longreply
"Even if [not catching up with MS.NET] is the case, we still win, because we would get this nice programming environment, that althought might not end up being 100% .NET Framework compatible, it would still be an improvement and would still help us move forward. So we can reuse all the research and development done by Microsoft on these ideas, and use as much as we can."
"GNU is a free re-implementations of Unix. Linux is a re-implementation of the Unix kernel. Before the advent of Linux and the Berkeley Unix, Unix was a proprietary technology, built by ATT (which back in the day, was a monopoly).
Still, developers took what was good from Unix, and reimplemented a free version of it. Down to the Unix programming language: C (which was also invented at ATT). Even C++ was invented at ATT.
Think of Mono as following the same process: we are bringing the best technology out there to our beloved free software platform. And at the same time it serves to be a magnificent upgrade on the development platform. "
Does modern Mono ALSO offer a simpler migration path from Windows, e.g. by offering System.Windows.Forms? Yes - it does now. It's done the important part, offered a decent development platform allowing people to stop wasting their time in C, for ages. Quality apps (including non-Novell-sponsored, before you start) like Gnome-Do or Galaxium are a testament to this.
>Mono is all about getting existing Windows applications, >and their Microsoft-proprietary dependencies, installed >on to your Linux system, so that you will in the near >future require a paid-for license from Microsoft to run >programs on your Linux system.
Now, really, this is just tinfoil hattery. Up until here, the post is reasonably articulate, albeit with a few factual slips here and there. But at this point, it heads into "they're all mini Steve Ballmers" territory.
No, Mono is not "all about" any of the above. It never has been. If you believe that the people who contribute to Mono do so out of intentional hate and a desire to force you to buy Windows, then, well, that's just sad. You can believe that contributors are misguided - that's fine. I think that myself about some people I meet online. But thinking they're actively seeking to magic some mythical Windows apps which wouldn't be complete ass onto your Linux desktop, and make you buy licenses from Microsoft... well, is it so hard to see why people mock this kind of stance? Why not suggest they're really evil space-faring lizard-people and be done?
>Note: the argument built in this post is structured using >references only from the Mono project itself. It does not >rely on any potentially biased words from sources such as >the Boycott Novell website … only the Mono project’s own >words are quoted.
No comment.
There you go. Happy? -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (GNU/Linux)
iEYEARECAAYFAkkjd7AACgkQWQvNUd/Cr8GJHgCg2u2LCw4IhoIg0B6CPegjbXMM j0IAoIGSTU221gzxYWzK00Iwxg58xUDa =Zbvn -----END PGP SIGNATURE-----
jo Shields
2008-11-19 02:21:18
Roy Schestowitz
2008-11-19 02:29:17
Roy Schestowitz
2008-11-19 08:54:56
stevetheFLY
2008-11-19 11:13:46
So there is little to wonder here.
Note: comment has been flagged for arriving from an incarnation of a known (eet), pseudonymous, forever-nymshifting, abusive Internet troll that posts from open proxies and relays around the world.
AlexH
2008-11-19 11:20:08
http://www.openinventionnetwork.com/pat_linuxdefpop.html
It covers Mono. Basic research, yet again, as ever.
Roy Schestowitz
2008-11-19 11:30:33
I see a big list of applications/programs, but I very much doubt it based on the words of OIN's new leadership. See the bottom part of this recent post.
stevetheFLY
2008-11-19 11:41:00
There is a tipping-point when outright denial of facts just starts making you look sad, Roy.
Note: comment has been flagged for arriving from an incarnation of a known (eet), pseudonymous, forever-nymshifting, abusive Internet troll that posts from open proxies and relays around the world.
AlexH
2008-11-19 11:48:36
http://www.openinventionnetwork.com/pat_linuxdef.php
As I said before, basic research.
Roy Schestowitz
2008-11-19 11:55:27
It remains true that the approach they take can be more harmful then beneficial (it's a Maginot line).
stevetheFLY
2008-11-19 12:04:02
Note: comment has been flagged for arriving from an incarnation of a known (eet), pseudonymous, forever-nymshifting, abusive Internet troll that posts from open proxies and relays around the world.
Roy Schestowitz
2008-11-19 12:32:14
stevetheFLY
2008-11-19 12:37:10
Note: comment has been flagged for arriving from an incarnation of a known (eet), pseudonymous, forever-nymshifting, abusive Internet troll that posts from open proxies and relays around the world.
Roy Schestowitz
2008-11-19 12:38:34
stevetheFLY
2008-11-19 12:44:39
Don't proofread your posts, I'm fine with typos; but check your 'facts' to make sure they PROVE anything.
Note: comment has been flagged for arriving from an incarnation of a known (eet), pseudonymous, forever-nymshifting, abusive Internet troll that posts from open proxies and relays around the world.
Roy Schestowitz
2008-11-19 12:51:41
stevetheFLY
2008-11-19 13:11:06
Note: comment has been flagged for arriving from an incarnation of a known (eet), pseudonymous, forever-nymshifting, abusive Internet troll that posts from open proxies and relays around the world.
Roy Schestowitz
2008-11-19 13:58:55
twitter
2008-11-19 15:10:57
Jo, I agree when you say, "Short version: the whole deal is a plate of bullshit and chips," but don't understand how anyone could be comfortable working on that basis. Roy's work with Matlab taught him a valuable lesson in non free software. I'm also glad he's decided not to waste his talents furthering the cause of non free software. The only thing worse than not using your talent is being used to harm your neighbors.
G. Michaels
2008-11-19 15:31:47
http://slashdot.org/~SockDisclosure/journal/214377
Yes, that's the same person. Stunning, isn't it?
Note: writer of this comment adds absolutely nothing but stalking and personal attacks against readers, as documented here.
AlexH
2008-11-19 16:25:45
There's nothing "trollish" about correcting FUD aimed at an organisation that protects (some) free software.
Bon Scott
2008-11-19 23:40:02
Under Debian/Ubuntu: $ sudo aptitude purge mono-common libmono0
Under Fedora Core 9: # yum erase mono-core libgdiplus
We, as end-users can yet choose what should be installed or not on our systems, but for how long ? If the main distro providers failed to preserve the safe haven that is GNU/Linux, what will remain to us, BSD ?
Victor Soliz
2008-11-20 19:42:49
lt does not seem to make your actions less trollish. Considering this company you defend so intensively has done nothing but harm free software rather than protect it.
Roy Schestowitz
2008-11-20 20:00:25
AlexH
2008-11-20 20:17:35
Roy Schestowitz
2008-11-20 20:24:33
AlexH
2008-11-20 20:41:24
Roy Schestowitz
2008-11-20 20:57:15
Ian
2008-11-20 21:43:25
Which again, is your opinion(just like Alex said), not fact. The all or none attitude, like what Victor showed above, is the general problem and why people on this site are some times labeled as zealots and can be dismissed outright. The fact is, Novell is neither angel or demon. They do some really positive things and some really stupid things in both the open source and "enterprise" realm. And even those "things" I refer to can be looked at from differing points of views and debated.
Roy Schestowitz
2008-11-20 21:48:51
http://techdirt.com/articles/20081118/0323432865.shtml
Groklaw goes further and expects legal action.
Also today:
http://slashdot.org/firehose.pl?op=view&id=1750919 ( Microsoft Moves Into Debt)
AlexH
2008-11-20 22:04:56
Roy Schestowitz
2008-11-20 22:07:46
AlexH
2008-11-20 22:17:00
If you think previous employees would join other companies, possibly. Canonical - not sure why they'd expand their desktop developers from the current 2/3 people. Red Hat - not sure their desktop products need developers :D
In short, I think you're dreaming. Novell disappearing would put a serious hole in key projects. Your example of "Samba relocating" isn't amazing, given only three(?) people were employed by Novell in the first place. Samba remains, sadly, chronically short of developers :(
Roy Schestowitz
2008-11-20 22:23:03
Josh Bell
2008-11-20 23:30:09
Roy you continually look at things in a very black and white way. The world is a lot of gray sir. Companies make choices, we don't have to like them but Novell does a lot for and with open source projects. While you have problems with things like Mono and Moonlight, obviously other distros don't.
I had a lot of respect for you and what you were dong but your twisting of stories, leaving out pertinent facts, ad-hominim attacks on Novell and the good things they have done has left me disappointed in your antics and your rhetoric.
Roy Schestowitz
2008-11-20 23:45:18
Dan O'Brian
2008-11-21 03:24:14
That would hurt Free Software.
Novell disappearing off the map would be a devastating blow to Free Software development, no matter how you look at it.
Roy Schestowitz
2008-11-21 03:29:04
Ian
2008-11-21 04:26:21
I guess the pragmatic follow up question would have to be, what does that severity translate into? What sort of real metrics can be used to determine the severity at which the patent deal is hurting FOSS. Are we concerned about lost revenues FOSS companies are experiencing compared to proprietary solutions? I'm not sure can we even nail down the exact financial impact the deal has had on anyone, Novell included. If we're concerned about possible litigation from Microsoft due to this deal, that's a train that's not going to stop anyway. Now, I think it certainly emboldens Microsoft to sabre rattle, but I don't think it has any impact on litigation. Basically, I don't know how the deal is going to work out in the long run. Moreover, I couldn't tell you in terms of real evidence what impact this has had to this point, good or bad. If you point to Mono, then Novell themselves are the most liable as the main contributers to the project. If they were to be sued by Microsoft because of Mono, that's their own fault.
As I see it, since we can't nail down real numbers on the deal's impact, it turns into an argument of ethics. Personally, I think Novell handled the deal poorly from a PR standpoint, not even getting their "open source movers and shakers" at the time involved apparently. That makes me realize that this was a purely a business decision. The logical follow up question to that would be, well why didn't a Red Hat follow suit? In my opinion, Red Hat certainly had nothing to gain in following suit. Novell's management probably believed and probably still believe that they(Novell) did. Now whether they were cognitive of whether or not what the patent implications were to their FOSS competitors and FOSS in general, I'm not sure and we can only really speculate. Maybe there was a healthy amount of customers who voiced concerns about legal issues involved Linux. Again, I'm not sure and I don't think any of us are.
Again, I think my main point is(and to echo Josh Bell), things are rarely black and white. I feel like you made up your mind two years ago about this deal and try to fit pieces of the puzzle together to create your own picture. Novell walking away from Hula shortly after the deal is something that still sticks in my mind. The claim was, at this site, that Microsoft influenced Novell to make that move completely ignoring GroupWise and the lack of developers(at the time) and how that claim just didn't add up at all. I think you'd find less detractors, if you even care about that, if you took a more even hand at these bits of news that come out. You'd certainly serve you cause much better.
Jose_X
2008-11-21 05:58:14
***
Roy, the link Alex provided listing all of those packages http://www.openinventionnetwork.com/pat_linuxdefpop.html is linked from the very bottom of this page http://www.openinventionnetwork.com/pat_linuxdef.php .
To verify it's what he is saying, read in particular the definitions for "Linux Environment Component" and "Linux System" from the definitions page (the second link).
I haven't read too much more beyond that, but note the use of "Linux System" in the top-level license page http://www.openinventionnetwork.com/pat_license.php as well as in other pages. "Linux System", as per the definitions page, basically can be expected to include all the packages listed on that other page.
***
[jo Shields] >> No, Mono is not “all about” any of the above. It never has been. If you believe that the people who contribute to Mono do so out of intentional hate and a desire to force you to buy Windows, then, well, that’s just sad. You can believe that contributors are misguided - that’s fine. I think that myself about some people I meet online. But thinking they’re actively seeking to magic some mythical Windows apps which wouldn’t be complete ass onto your Linux desktop, and make you buy licenses from Microsoft… well, is it so hard to see why people mock this kind of stance? Why not suggest they’re really evil space-faring lizard-people and be done?
"all about" is a simplification. Reread the comment you quoted. There is no accusation mentioned. [Maybe the author feels as you implied (or perhaps not), but in this write-up you quoted that wasn't suggested.]
>> So, short version is “I’d rather not, but I’m pragmatic, and I want a degree”
If you read over his example to refuse the job, it hardly suggests you summary quote would be an accurate description. Most "rather not but I'm pragmatic" people would not refuse a high paying job.
In the section where you include what is protected, lots of people are left out in the cold. I do hope the patent risks are diminishing, but trying for interop and to follow MS closely increases risks [see eg, http://www.groklaw.net/comment.php?mode=display&sid=20081114205538108&title=Attack%20of%20the%20patent%20orcs&type=article&order=&hideanonymous=0&pid=0#c737749 ].
My greatest dislike is for the spreading of "Microsoft protocols" [see http://boycottnovell.com/2008/11/18/explanation-of-mono-issue/#comment-37735 ] and generally supporting Microsoft. [This last link leads to a comment with 2 links. The second one takes you to a somewhat lengthy comment that mentions mono and how Novell is selling Linux short. I would very much appreciate a response to that piece if to nothing else.]
>> >They are indeed open source, >> Free Software.
Is there a reason you clarified this?
"Free software" includes lots of non-copyleft licenses that don't protect developers or users or (in short) copyright owners (eg, the BSD license, MIT, Apache) so which the FSF does not really prefer be used or use themselves in their works. [not to imply all copylefts would be good or that there aren't cases where nongpl might be best]
Were you trying to suggest you really view software licenses that closely to the FSF's views? If so, fine, I suppose you really would in that case stand up for the GPL against other open source licenses (I'll have to remember that about you); however, some people use the "free software" term when they really have no problem with or even exhibit a preference for BSD and friends. To use "free software" as a correction of "open source" when in fact one isn't clearly behind the GPL would be deceptive, to say the least.
>> Short version: the whole deal is a plate of bullshit and chips.
The benefits Microsoft claims are bullocks for the most part, I agree; however, the cooperative deal and money exchanges between Novell and Microsoft are very real and damaging to Linux/FOSS.
>> Well, sure. But I don’t have a license to NTFS-3G. Or to the WMV/WMA playback in Totem. Or the page up/page down keys that work in Firefox.
It's a very different thing when you follow Microsoft and strive for interop vs. when you mind your own business and then have some third party claim they invented rice and beans. Very different. The probabilities of violation and the probabilities of being able to code around a violation at a cost of X are very different. Microsoft was there first laying the bombs along the path you specifically seek to walk on, for the sake of the std or for "interop".
mono and some of the other Novell favorites keep growing outside their little boxes by definition. That's the whole point of development libs/tools. Contrast with a specific app's specific codec which you can easily avoid when you use that app or develop with it.
>> I’ll quote a little from the VERY early days of Mono...
All of these quotes I see people bringing up from mono devs have a common theme: we can't or won't improve upon Microsoft's offering in any real way, and we are willing to redo the wheel in order to spread Microsoft's flavor of what basically already exists.
I am willing to give some benefit of doubt for past mistakes, but at this point, all of these people have options but don't change.. and you know at least one reason why that likely is that they don't change? It's for the same reason MS should have been avoided in the first place if you care about FOSS and undoing the interlocking monopolies of Microsoft. The same link as above: http://boycottnovell.com/2008/11/18/explanation-of-mono-issue/#comment-37735 . Brain cells are a very prized item.
Miguel and others have constantly been quoted as saying how decent Microsoft is/ is becoming/ etc. Talk about blindness, lapse of judgment, or perhaps being really hooked on Microsoft's crack.
***
>> Samba remains, sadly, chronically short of developers
Alex, your opinion that more people should be developing some of these software that interface with Microsoft is not shared by everyone as a preferable way for a FOSS coder looking to contribute maximally to FOSS to do so.
Point of diminishing return. The goal of Samba, sadly, is for interop with a closed source product actively maintained and updated to defeat interop.
Let's really show enterprises and users everywhere what we can do, not how badly we fail at interop with the monopolist.
>> Forking SUSE doesn’t solve anything, that doesn’t create developers. ...Novell disappearing would put a serious hole in key projects.
Concerning the value of Novell to FOSS, I would not miss Novell (well, I would a little but would get over it). First, if one has to pick between Red Hat or Novell, it's clear Red hat (a) contributes more code to helping Linux (in absolute numbers and relative to other platforms, hence implying a net contribution), (b) combatting Microsoft (the biggest obstacle to Linux), and (c) pushing for the most FOSS/user/dev friendly business model. Second, if Novell was gone, those from Novell who really care about FOSS, rather than about helping Microsoft embrace and extend FOSS, would still work on their projects but sanely. They would have less pressure to follow the corporate direction aligned with Microsoft. They would likely get jobs quickly elsewhere.
Alex, did you reply to that long conversation we had started? I'm mostly interested in your views of how Novell is taking actions that almost completely aid Microsoft rather than help weaken them. You could possibly really help Roy show BN readers just how much of a liability Novell is to Linux/FOSS.
Jose_X
2008-11-21 06:24:16
It's very easy to exaggerate. It's also of no use. Please be specific so things can be corrected or explained more carefully.
>> Roy you continually look at things in a very black and white way. The world is a lot of gray sir. Companies make choices, we don’t have to like them but Novell does a lot for and with open source projects. While you have problems with things like Mono and Moonlight, obviously other distros don’t.
You are very correct that a lot of people *really* don't like the choices some companies have made (I have two companies in mind this second). Some grays are very dark. You obviously disagree here or you would not be trying to explain how the black is actually a bit grayish: "moonlight black" and not "100% black".
I just finished a post where (near the bottom) I disagreed about Novell's net value to Linux/FOSS.
A number of well-known and respected individuals have stated they do not like mono. Sure, you can find some that will say they do like it.
>> The same Red Hat whose market cap dropped below 2billion? Please. Yes you are the one who brought up Novell dropping below 2 billion in market cap.
I think Roy gets enthusiastic when Novell drops, but not when Red Hat drops. He also gets enthusiastic when Microsoft drops.
>> If any of the Novell users you mention such as myself use e-Directory where are we supposed to go?
..to an interoperable substitute!
What?
There are none?
Holy ... You should go with supported FOSS next time around.. unless you don't mind the risks of Novell going under.
... and what does that have to do with Novell contributions to FOSS? ..or why should that bother someone that doesn't use Novell products or has left them.
***
Dan, I won't repeat comments I just posted in my last batch of replies on this topic about Novell, except to say that I disagree with you as well.
How would you feel about Red Hat going under?
Just like Novell devs, many of them would find work or contribute part time or maybe their contributions weren't that valuable.
Microsoft "interop" technologies, which has become a very large percent of what Novell+Microsoft supports, will not be missed by me nor by many others.
I would hardly be able to contain myself when the day arrives.
Roy Schestowitz
2008-11-21 09:29:25
Josh Bell
2008-11-21 11:35:45
I'm not defending Moonlight, I don't use Moonlight and I don't use Mono either on my workstations or servers. There are many respected people that do use Mono and Moonlight but just not here.
I tried OpenLDAP but as I stated it is not e-Directory and it doesn't allow me to do what e-Dir does. Roy stated that Novell could go under and RH or Ubuntu could take over their projects. Well OES2 is SLES10 with Novell propietary technology. If the FOSS piece goes under so does everything based on it and with it.
Ian
2008-11-21 13:35:02
There’s a pattern where people defend Novell because of some proprietary technology they offer. There are rebuttals to this too.
Roy, you realize that you're referring to a comment that I responded to, only to have my follow up questions ignored and not be refuted in any way, right? That doesn't prove or lend credence to anything.
@Jose_X
Microsoft “interop” technologies, which has become a very large percent of what Novell+Microsoft supports, will not be missed by me nor by many others.
I would hardly be able to contain myself when the day arrives.
Could you elaborate on this please? Are you saying that Novell's entire product portfolio has changed over the past two years to serve as a Microsoft interoperability puppet? Or rather, are you saying that Novell has begun to support open source projects which look to be interoperable with Microsoft technologies?
Roy Schestowitz
2008-11-21 14:30:13
They can sell these assets and swap their SUSE implementation with something else that shares the same codebase.
Victor Soliz
2008-11-21 16:13:03
Their intention is to destroy free software, most of this money is going to power .net lock-in apps so Linux becomes dependent to their beloved Mono, other bunch of it is going to fork openOffice, regarding other things like Linux, gnome and KDE, well, sorry, but we can do without Novell, we didn't need their millions in the first time, and I don't really think we should whore ourselves to them and let them do what they are doing just because of money. Their contribution is severely overrated anyway.
Victor Soliz
2008-11-21 16:15:57
Roy Schestowitz
2008-11-21 16:34:39
Shame on Novell. As Sam Varghese said last week, they just cannibalise GNU/Linux. To make matters worse, they try to expand that also to Free office suites. The pattern is so familiar.
Dan O'Brian
2008-11-21 16:49:44
KDE is meant to destroy Linux? How? Samba is meant to destroy Linux? How? OpenOffice is meant to destroy Linux? How? XGL / Compiz is meant to destroy Linux? How? The Linux kernel is meant to destroy Linux? How? ALSA is meant to destroy Linux? How?
You make no sense.
Roy: they /are/ contributing to OOo, they submit all of their patches upstream. Also, since OOo is Free Software at no cost, saying that Novell forked OOo to steal customers from Sun is the most retarded thing I've ever heard.
Linux users do not buy or even download OOo from Sun, nor have they in close to a decade. They get their version of OOo preinstalled from their Distro (Red Hat, Ubuntu, Debian, openSUSE, whatever). Most of those distros use Novell's fork.
Chris
2008-11-21 17:04:46
go-oo.org is no fork of openoffice.org - and that the guy from SUN says otherwise doesn't make it less false.
Point being that go-oo.org is based on openoffice.org and adds additional things. To be a real fork it had to be based on the same code base sometimes in the past and then had to go down a different development street which is clearly not the case.
There are many reasons why there are things in go-oo but not in openoffce - a few of them is that getting stuff upstream into openoffice is pretty bureaucratic, requires transfer of intellectual ownership to SUN, patches were proposed and rejected, were proposed and are waiting for decision, and so on, which has been discussed quite some times so you'll surely find them if you actually could be arsed to look them up instead of babbling your usual ramblings of "Bad bad Novell" ...
However, surely I would love to see a better relationship between the two of them but your silly claims that it would be an hostile fork (which it isn't) and for "selfish gains (stealing customers from Sun Microsystems)" (Novell and SUN both surely earn loads of money by investing time & personal into the development of an office suite that is given away for free ...) actually don't really improve the situation.
IMHO I would love to see an OpenOffice foundation but I guess that wont happen for some time at least.
Now regarding the "Mono problem": pretty strange that my system works totally fine and there isn't a single Mono package installed nor did I have to uninstall or deselect one during installation. And here it comes: I'm using openSUSE (and I verified that this still holds for openSUSE 11.1) so clearly "bad bad Novell" is pushing Mono down everyones throat ...
Cor
2008-11-21 17:10:05
Similarly, how about this whole 'SLED sources not freely available thing'? Have we heard one word of correction from Mr Schestowitz about that? No, we haven't.
Roy Schestowitz
2008-11-21 17:36:35
Chris
2008-11-21 17:41:49
AlexH
2008-11-21 17:42:24
Roy Schestowitz
2008-11-21 17:43:05
AlexH
2008-11-21 17:47:54
Cor
2008-11-21 17:50:32
And since when does a fork commit code upstream? Certainly not a fork by any known definition of the word.
Chris
2008-11-21 17:55:41
And that proofs what? Nothing ...
Besides that, the only one I ever heard of calling it a fork besides you and your shills was the SUN guy who posted here once.
So, it may be shocking to you, but it doesn't really matter if he, you, your shills or anyone else claims that it is a fork if it isn't according to the common definition.
Or, according to your argumentation, the earth would be a disk and no ball because I say so. Have fun in your little world ...
Roy Schestowitz
2008-11-21 17:59:15
AlexH
2008-11-21 18:01:09
Dan O'Brian
2008-11-21 18:13:39
Therefor it is not a fork.
A fork would mean that the 2 products are going in separate directions (they are not).
Also, if Go-OO was a fork, why do the Go-OO developers still submit their patches upstream?
Jose_X
2008-11-21 18:22:25
For the moment, you people should go read http://en.wikipedia.org/wiki/Fork_(software_development) . I'll quote a bit from the top:
>> 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.
Novell's and everyone else's code is a fork because they don't look the same. In particular they are all forks of oo.org.
Now skipping a bit to quote again:
>> A kind of fork that is standard practice in many projects is a stable or release version, modified only for bug fixes, while a development tree develops new features. Such forks are usually referred to as "branches".
Got that. Even within the same organization, a branch is considered a fork from main.
In legal matters if two works differ by a sneeze, they are different. It doesn't matter if one is version 1.2.2.1 and the other is 1.2.2.1b
And again later on:
>> In free software, forks often result from a schism over different goals or personality clashes. In a fork, both parties assume nearly identical code bases but typically only the larger group, or whoever controls the web site, will retain the full original name and the associated user community. Thus there is a reputation penalty associated with forking.
I think this describes the oo.org and go-oo situation to a T.
I think it is pretty clear that either wikipedia is wrong, and presumably wikipedia represents a lot of the community, or everyone here that said that go-oo is not a fork is wrong.
Personally, I'd rather think that the people here are simply taking a different definition of fork; however, what bothers me is the use of this stupid thing: a definition of fork (which they didn't even get right according to wikipedia) and using that to bash and bash and bash this web site.
The fork name is a tiny issue that has nothing to do with anything of substance here. It shows AGAIN, that people will come here, disregard the blog entries many times to focus on being wrong but pretending they are absolutely right so that they can repeat over and over that Roy and BN and shills are wrong.
It's going to get embarassing or tiring or something at some point in time one would hope cause nothing seems to stop this behavior. Anyone want to name names? Maybe we can start putting tick marks and links next to people's names so that they don't forget when they were wrong and Roy, shills, and BN were right. Maybe we can make a list and link back to this conversation because these people don't don't don't stop. They are wrong and the keep going and going and going.
People. Be adult. In general, understand that not everyone uses the same exact definitions as you do. It makes you look like you have an agenda when you ignore that and many other things so you can get you bashes in on Roy, shills, and BN.
I'm only speaking in this brief case to those that said go-oo was not a fork and then bashed/superbashed Roy, shills, and BN (or some subset thereof) for saying it was a fork, plus for anything else they could think of to add to the fire.
If you people don't have an agenda to knock on BN, you sure are fooling me.
Jose_X
2008-11-21 18:25:45
Roy Schestowitz
2008-11-21 18:25:58
Roy Schestowitz
2008-11-21 18:26:26
AlexH
2008-11-21 18:31:06
I'm very perplexed by the relationship you have with Sun. You're very quick to defend them, even though they have a patent agreement with Microsoft, develop OOXML, etc. And now you have some secret information about OOo?
I'm a Sun supporter - and I wish them all the best in these economic conditions - but it does seem that you have a bit of a bias yourself.
Jose_X
2008-11-21 18:31:54
Is there anyway, we can increase the percentage of times that disagreements fall into that bin and not the "sock puppet" bin?
Roy Schestowitz
2008-11-21 18:33:12
I'm a critic of Sun. Search the Web and you'll see.
Dan O'Brian
2008-11-21 18:33:38
I don't see where "looking the same" has anything to do with the quote you provided.
Go-OO is not a distinctly new piece of software, it is OpenOfice. It is even /called/ OpenOffice (which disproves your second quote, which requires that the fork be fall under a new name - it doesn't - Go-OO is noit the name of the product, it is just the website). Since all the patches made to Go-OO are submitted upstream to the official OO.o project, it can't even be considered independent.
To prove that Go-OO is not a fork, all I have to do is disprove one of the required criterion for defining a fork, which I just did... and then some.
Hence: QED.
AlexH
2008-11-21 18:39:15
Sun is important to defend. They mean well.
Doesn't sound very critical.
Roy Schestowitz
2008-11-21 18:46:44
http://schestowitz.com/UseNet/2007/May_2007_1/msg00296.html http://www.schestowitz.com/UseNet/2007/June_2007_3/msg00214.html http://schestowitz.com/UseNet/2007/November_2007_1/msg00842.html
Jose_X
2008-11-21 18:52:22
>> You’re very quick to defend them, even though they have a patent agreement with Microsoft, develop OOXML, etc.
Maybe when you try and read and reply to the link I keep repeating and asking you to reply to, you will have a better understanding of why Sun does not fall into the same boat as Novell.
I know you addressed the question to Roy and specifically mentioned patents and ooxml development. Roy has stated he is against each of those very specifically and emphatically.
What I don't think you have figured out yet (and it has been pointed out to you multiple times, at least some of those by me) or you are ignoring in order to (a) put Roy on the spot to make or repeat a mistake, (b) waste his time, (c) distract, upset, discourage, or drive him mad, or something else is that, if Sun or the FSF or Canonical or anyone else did do something similar to Novell for which Novell is scolded verbally on this site, they overlap with Novell in these few items and not nearly in everything or to the same degree.
Novell is worse. They go beyond normal FOSS competition to heavily favor Microsoft in the competition. I know they are getting paid for this behavior by Microsoft. OK. So what is the problem? BN is just pointing out all the ways Novell is helping Microsoft. If people are upset by this, they should go send a letter to Novell at their headquarters. Bashing Roy is only making the bias carried by you people that much clearer.
This is a blog site that does a very decent job given the circumstances. If the entries get too heated for anyone's taste, it will help your cause to come here and not make it look like you don't have an agenda.
AlexH
2008-11-21 19:11:12
@Jose: I would respond more to you if you were concise. At the moment, your posts seem like teergrube. You talk about Novell being paid by Microsoft, but ignore other people who've also been paid by MS (e.g., Sun).
I have little problem with either Sun or Novell. They're basically very similar, in terms of both problems and contribution.
Ian
2008-11-21 19:22:59
This is a blog site that does a very decent job given the circumstances. If the entries get too heated for anyone’s taste, it will help your cause to come here and not make it look like you don’t have an agenda.
Does this mean that anyone who disagrees automatically has an agenda? The fact is, everyone has an agenda to a certain degree. If someone posts something here(Roy, you, me, Alex, whoever), that post is and should be scrutinized for the content rather than the murky underlying agenda. Put simply, if you can't back your statements up, don't get upset when you're called out on it. Doing the internet version of sticking your fingers in your ears and closing your eyes doesn't further your goals to skeptics either.
Is this site about one sided ranting or is there any value put on dialog? If it's the former, the comment system should be turned off. If it's the latter, then people should not be dismissed outright for their perceived(and quite possibly incorrectly perceived) agendas.
Roy Schestowitz
2008-11-21 19:27:04
Chris
2008-11-21 19:29:16
Yeah, right, that must be true ... Apparently I created quite some forks then ...
I dare to say that Wikipedia, who is, as we all know, the holy grail of objectivity and wisdom is wrong - not only - here.
Roy Schestowitz
2008-11-21 19:31:43
Ian
2008-11-21 19:37:14
Probably because it's wholly irrelevant to the discussion, regardless of where you come down on the issues. The content of the discussion is what should count the most.
Roy Schestowitz
2008-11-21 19:48:53
Do you want to me take stock of logs and show you Novell's activity in BN?
BTW, hot in the news right now:
"The PDF document holds a single paged scan of an internally distributed mail from German telecommunications company T-Systems (Deutsche Telekom), revealing over two dozen secret IP address ranges in use by the German intelligence service Bundesnachrichtendienst (BND). Independent evidence shows that the claim is almost certainly true and the document itself has been verified by a demand letter from T-systems to Wikileaks."
Jose_X
2008-11-21 19:53:11
Distinct: "Readily distinguishable from all others;" [ http://www.answers.com/distinct ]
So to repeat the quote for fork from wikipedia: >> 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.
And what I said: >> Novell’s and everyone else’s code is a fork because they don’t look the same. In particular they are all forks of oo.org.
I hope you now believe that all forks are distinct from each other. [I'm assuming various unnamed parties produce different binaries.. Whether that assumption is correct or not has nothing to do with the definition of fork.]
It should also be pretty clear that these people put their products together independently from each other. [again, this assumption can be met if we just look at those that don't share the oo.o repos.. again, if that assumption is incorrect it has nothing to do with the definition of fork.]
A more complete sentence on my part (in order to have been explicit about the independence) would have been: -- Novell’s and everyone else’s code is a fork because they don’t look the same *and are put together independently from each other*. In particular they are all forks of oo.org.
Did anyone besides Dan O'Brian get tripped up by this? It's difficult to know what people's backgrounds are. I don't speak/write perfectly in any language. Fortunately, I don't think I was wrong here, just "sloppy" in not being explicit with the full thought. Foolish me, I thought it was clear that an assumption was that the different distros work independently of each other.
OK, Dan. So you tried to show I had made a mistake in my post.
Now, I want you to go back to my comment (the one you quoted) and tell me if you deny that I was correct in implying that you got the definition of fork wrong, at least according to the wikipedia page.
If you agree that Roy did use the/a correct definition, it's really up to you if you want to apologize to Roy or to anyone else here. I don't need an apology, but then again, I am not hounded by you people. Roy might feel differently.
...Oh, wait. I am not done.
>> It is even /called/ OpenOffice (which disproves your second quote, which requires that the fork be fall under a new name - it doesn’t - Go-OO is noit the name of the product, it is just the website).
You are incorrect to claim that what you state here disproves that second quote from wikipedia [BTW, I think you meant "third quote" not "second quote"].
Please read it carefully again and pay careful attention to this word: "typically". If you need to look it up or if you have problems understanding the definition just holler.
>> Since all the patches made to Go-OO are submitted upstream to the official OO.o project, it can’t even be considered independent.
I think you need to review the definition of fork (the first quote). I think most people would agree that taking a copy of OO.o (I'm referring to the code base found on oo.o website at some point in time), developing it away from the oo.o main branch would qualify as independent development.
We can disagree over exactly how they might be defining "independent" in this context, BUT the second quote I provided, which appeared just a little below the first one in the wikipedia page, shows that the article considers even version branches to be forks (ie, independent). Thus if branches at oo.org are considered forks of each other, I'd suspect so too would "branches" developed under different leadership. Sure, these brances may all try and sync up every now and then, but they are still forks while they remain "independently developed and different".
Again, you might have a reasonable definition of fork, but it does really seem to me that it disagrees with my definition as well as with the definition on wikipedia.
What got me upset was the hounding of Roy and/or shills on that stupidity.
Do people that come here to hound Roy and/or shills pick different meanings for English words than what Roy uses purposely so they can hound .. or is this an accident?
In the future and in light of this incident here where it was shown that Roy and shills were attacked for being correct but alleged otherwise, may I suggest that you all consider approaching your comments here with a little more regard for your own limitations?
stevetheFLY
2008-11-21 19:50:04
Note: comment has been flagged for arriving from an incarnation of a known (eet), pseudonymous, forever-nymshifting, abusive Internet troll that posts from open proxies and relays around the world.
Ian
2008-11-21 19:55:31
What you say seems to suggest that it’s acceptable for Novell employees to leave comments here as “anonymous”
Yes.
even if the comments are a personal attack.
No. It doesn't matter if it's a Novell employee or not. It also doesn't matter if the person fully discloses their name or anything else. Again, it's the content of the messages. Furthermore, disagreeing doesn't always equate to personal attacks.
Do you want to me take stock of logs and show you Novell’s activity in BN?
Don't bother because it doesn't matter.
Roy Schestowitz
2008-11-21 19:57:30
Roy Schestowitz
2008-11-21 19:59:05
Jose_X
2008-11-21 20:06:45
I don't know why you have that opinion. It's so easy to convincingly show that *what* is "contributed" must be taken into account and not just *that* something is contributed.
Then again, you also have stated or implied that all companies equally pose the same threat from patents to all FOSS projects. That too makes little sense. Grep for "probability" in this thread. What I said is basically something I have already told you on more than one occasion.
Do you want to adjust your position or add clarity on either of these two issues?
Do you think *this* comment is long and like teergrube? [Please answer the above question first before tackling this current one about teergrube.]
Jose_X
2008-11-21 20:14:02
>> If it’s the former, the comment system should be turned off. If it’s the latter, then people should not be dismissed outright for their perceived(and quite possibly incorrectly perceived) agendas.
I agree with the second part to a large degree.
I disagree with the first part. You can come here and leave your input, and that is preferable to not having that option. Many articles out there done by people that get paid money and whose work are considered professional products do not individually address all concerns in replies even when they appear correct. I don't see why Roy should be held to a higher standard considering he does take many things into account and this is sold as a blog where numerous fairly distinct entries are posted daily by the same individual.
People should be contributing constructively if possible and doing less hounding. That's what I think makes the most sense.
AlexH
2008-11-21 20:14:29
I'm not totally sure what your question regarding patents is; that seems to be a very general hand-waving. If there are patents covering a piece of free software, that's a problem.
I'm also not sure what you mean by "good" or "bad" free software contributions. If you don't like Mono, that's your choice - other people don't like Perl. It doesn't stop Mono being free software.
Roy Schestowitz
2008-11-21 20:17:30
AlexH
2008-11-21 20:19:01
Roy Schestowitz
2008-11-21 20:25:34
AlexH
2008-11-21 20:27:33
Therefore your definition of what might be a contribution is entirely irrelevant to the discussion.
Perhaps you should shoot from the hip less and think a little more. Ears:mouth in their usual ratio, yes?
Roy Schestowitz
2008-11-21 20:29:25
Dan O'Brian
2008-11-21 20:34:23
As someone else also said, most distros patch a huge number of the software they ship with the distro to some degree. Even if only changing icons, etc.
Does this mean that Red Hat has forked Evolution, for example? I would argue that no, they did not fork it. Even though they have patches that are not upstream.
Red Hat's version of Evolution is not a distinctly new version of the software, just like the OpenOffice that comes shipped with my Linux distros is not a distinctly new product than the official OpenOffice. And as we all know, most distros ship the Go-OO build of OpenOffice.
You can argue semantics all you want, but fact remains that it is not a fork.
Ian
2008-11-21 20:36:02
Who is holding him to a higher standard than anyone else? I specifically stated that everyone here should be held to the same standard. You can extrapolate beyond the confines of this blog as well. I think you're disagreeing with a point that was never made, at least by me anyway.
The only difference here, is that Roy posts to the main page and his volume of posting is obviously much higher than anyone else here. None of that, however, should make his or anyone elses' comments scrutinized less or more.
Dan O'Brian
2008-11-21 20:37:46
Of all the people that comment on this site, you shoot the messenger more often than anyone else - I would even go so far as to say that you shoot the messenger more than everyone else combined.
Roy Schestowitz
2008-11-21 20:41:50
Jose_X
2008-11-21 20:56:57
Well, how do you differentiate Joomla from Mambo? Did we not have a fork as soon as Joomla was started? Wouldn't it stop being a fork if we merge the two and cease development in one of them?
Independent and modifies.. yes.. a fork.
If you read my comment, I acknowledged that people's views of forks differ a bit. This is true for most things in life. Context matters.
Victor Soliz
2008-11-21 20:57:08
Novell may contribute to those projects, that's irrelevant to the discussion, what I say about they dumping most of their money in projects meant to destroy Free software does not magically become false with that, not to mention how Novell's contribution to these projects is not exactly as full of Holy Light and benevolence as you Novell zealots would like to make it look.
What I said is that most of the money Novell is contributing is done to Mono, the so called 'interoperability' and the little projects like F-spot and Tomboy meant to make OSS dependent on Novell. And that whatever contribution they make to actual benefical OSS projects, in no way should make them immune to criticism or even out of any responsibility for what they do and keep doing, that is all but negative to free software, of course, you keep repeating everything like a parrot, that Novell contributes to this project, etc.
They are after all not the only company that does it, and no, they are not vital or whatever, I see no reason to be so fanatically thankful of them as you are, that would force me to look away when they attack free software's future with their silly plots, I am sorry for that.
Victor Soliz
2008-11-21 21:02:27
Dan O'Brian
2008-11-21 21:06:56
Chris
2008-11-21 21:14:36
What is go-oo? A patchset against oo!
Some of these patches - were rejected from upstream - weren€´t yet pushed upstream - were pushed upstream and are waiting for integration into upstream - were ... (I guess you get the idea)
So go-oo is a project which is based on oo and extends its features while trying to push its additions upstream for integration.
IMHO this is no fork because - IMO - a fork is one project that got split into 2 independent projects (which is e.g. the case for Joomla and Mambo iirc). But not here, since go-oo is based upon oo (as in it is a patch set applied against the latest oo version) and therefore it is - IMO - not a fork.
Now say what you want, I€´m done with it ;P
Roy Schestowitz
2008-11-21 21:15:09
So the other 4,000 or so can 'negate' their work.
SCO Also Used to Contibute to Linux, Just Like Novell.
Dan O'Brian
2008-11-21 21:24:52
If the Mambo developers contribute all their patches upstream to Joomla (like Go-OO does with OOo) then I'd argue that it might not be a fork at all.
If you launch Go-OO, the splash screen says "OpenOffice.org 3" and even has Sun's logo in the lower-right corner. The about box mentions that it has modifications by Novell, but that's the only way to actually tell that I don't have the official OOo.
As far as what was meant by "typically" in the wikipedia article, it seems to describe who "typically" gets to keep the original name. "Typically" the originator keeps the name, meaning that the forked product gets a new name. However, that is not always the case (Epiphany & Galeon, when Galeon was forked, the original developer renamed his branch of the fork Epiphany while the other branch kept the name Galeon).
One of the forked branches always has to have a new name, unless it is the type of fork that is referred to as a "branch" as in "development branch" vs "stable branch".
Jose_X
2008-11-21 21:26:36
That is your opinion. I can accept it; however, my opinion seems to me to be closer to what is stated on wikipedia (you seem to agree on this point).
I now know you don't view go-oo as a fork but using a meaning of fork that contradicts what is on wikipedia. I'll try to remember that or point it out again in the future if necessary. I hope you or no one else will form a grand case out of Roy or anyone else calling go-oo a fork from now on.
Note that even the commonest definition of "fork" implies something along the lines of a branching or a separating into 2 or more distinct parts where each part naturally will diverge.
One thing I want to point out. It's clearer with this wikipedia definition what is or is not a fork. The definition you use has lots of gray (doesn't it?).. which implies a greater likelihood of there being arguments and misunderstandings.
How would you be able to check if something is a "fork" (your def) or not?
In fact, could you define what you mean by fork a little clearer. I have an idea of what you don't consider a fork. I admit, it doesn't make sense to me because all you appear to be saying is that the person independently developing what I and wikipedia call a fork is not forking if they keep tabs on some other project and intend to keep tabs.
Dan O'Brian
2008-11-21 21:29:57
Also, I said *over* 400 engineers working on Free Software.
That means that you can't do simple math like: "4000 - 400 = 3600 engineers working on proprietary technology".
I know, I know, we shouldn't let simple things like facts get in the way of a good smear.
Dan O'Brian
2008-11-21 21:37:46
I said no such thing.
The common definition of a fork in Free Software is 2 products emerge from a single codebase, where neither branch consumes patches added by the other branch and each branch ends up with separate names.
For example, NeoOffice is a fork of OpenOffice, but Go-OO is not a fork of OpenOffice by the standard definition of a fork. If Go-OO is a fork by your definition, then the Evolution that Red Hat ships is a fork of the official Evolution. Red Hat's Firefox is a fork of the official, etc.
I don't know too many people who would maintain that the Evolution and Firefox as shipped by Red Hat are forks of the official version. They are repackaged with some patches, yes, but they are not forks.
This is the same thing as Go-OO - it is a smallish patchset against official OOo.
If Novell started calling their version Novell Office, then I might be more inclined to agree that it is a fork, but they don't do that - they call it OpenOffice and give credit to Sun and only mention that it has custom Novell modifications in the about box.
Go-OO could be considered a *development branch* of OOo, but it is not a fork.
Chris
2008-11-21 21:39:12
IMHO go-oo is no fork because it is based on oo (as in it is just a big patch set) and once oo will release a new version go-oo will release a new version built upon the last oo release and thereby always keeping the oo codebase as its base against which its patches, that are, once again, apparently against common knowledge by some of the readers of this site, pushed upstream.
On the other hand It would be a fork if they had taken the source from oo at some point in time - e.g. oo 2.0 - and developed it independently from then on which had resulted in 2 different codebases but which isn€´t the case here.
Not really this hard to understand, or is it?
Roy Schestowitz
2008-11-21 21:47:19
The difference in terms of numbers of still a reflection -- upon the fact that Novell's F/OSS agenda is a minority.
Jose_X
2008-11-21 22:07:17
Novell can call it "openoffice" if they want. They don't own that name. They are allowed to call it that, but that doesn't change what the product is or isn't. Tomorrow they might not be allowed to use that name and it would not, right then and there, change the status of their project to a fork from not being a fork [at least not based on the definition of fork from the article].
Look at that sentence again: "...typically *only* the larger group [keeps the name]...." This suggests (in agreement with the rest of the article) that the name can be *shared*; otherwise, "only" would not be in that part of the sentence. Just ask yourself what difference exists when "only" is there vs. when it is removed.
So each party of a fork can potentially share the project name (but this would not be "typical" for forks). You still have the fork whether the name is shared or not.
I was about to suggest that maybe the go-oo project only supplies a set of patches so the fork is implied (ie, only a fork once you put everything together), but they actually appear to offer for download the full patched product. http://go-oo.org/download/
And the distros likely also maintaining their own full fork themselves as go-oo appears to be doing as well.
Roy Schestowitz
2008-11-21 22:14:26
Dan O'Brian
2008-11-21 22:44:57
Jose: yes, you can download prebuilt packages of the modified OOo with the Go-OO patchset (which they have done for convenience) on the website, but that doesn't prove that it's a fork.
You realise that Go-OO originally started out as a set of build fixes and a set of build scripts to make OOo easier to build on Linux, right?
Jose_X
2008-11-21 23:15:48
>> On the other hand It would be a fork if they had taken the source from oo at some point in time - e.g. oo 2.0 - and developed it independently from then on which had resulted in 2 different codebases but which isn€´t the case here.
While mainline oo.org is being developed, go-oo does not depend on it. Rather go-oo depends on a fixed earlier version of oo.org -- that would be the "common codebase" that is a part of every fork. All patching on top of that common codebase that differs from the main/other project is part of a distinct fork.
When you update the oo.org reference codebase to a newer version, you are simply copying over the codebase. Once you add the patches, you have then created a fork from that common codebase.
Jose_X
2008-11-21 23:18:19
I already explained why it is a fork (wikipedia). Which part are you not understanding?
Jose_X
2008-11-21 23:38:50
While mainline oo.org is being developed, go-oo does not depend on it. Rather go-oo depends on a fixed earlier version of oo.org — that would be the “common codebase” that is a part of every fork. All patching on top of that common codebase that differs from the main/other project is part of a distinct fork.
When you update the oo.org reference codebase to a newer version, you are simply copying over the codebase. Once you add the patches, you have then created a fork from that common codebase.
It's simple. When you add to code, you fork that code at that point.
Dan O'Brian
2008-11-21 23:41:42
That is the source if your misunderstanding.
Go-OO does not depend on a fixed earlier version of OOo, it depends on the latest revision of OOo at all times. It is a patchset that follows OOo cvs (or do they use svn these days?)
Dan O'Brian
2008-11-21 23:42:12
Jose_X
2008-11-21 23:45:45
Let me cross the t's and dot i's a little better.
If you take a snapshot of a code base that is developing outside your control, and you then make any modifications to that code, you have just forked that code base from that point.
You don't need to have a 20 developer army, and it doesn't matter what you hope or expect to do in the future. It also doesn't matter what you want to call your Frankenstein. You have created a fork because the other code base does not have the changes you added to your copy at the copy/fork point.
Wikipedia calls "branches" special cases of "forks", in case you people missed that. I'll quote it again:
>> A kind of fork that is standard practice in many projects is a stable or release version, modified only for bug fixes, while a development tree develops new features. Such forks are usually referred to as "branches".
http://en.wikipedia.org/wiki/Fork_(software_development)
You can go and try to merge the forks, but until such time, there will continue to exist one or more (perhaps private) fork to the copied code base.
Dan O'Brian
2008-11-21 23:47:43
Seriously?
Because I highly doubt many people would accept that definition.
Roy Schestowitz
2008-11-21 23:51:20
Jose_X
2008-11-21 23:51:53
The speed at which you sync up doesn't matter. Once you apply the patches to this new updated codebase you are forking from this updated code base. Every time they sync, they are creating a new fork which differs based on the patches applied.
Hello?
Jose_X
2008-11-22 00:01:25
>> Seriously?
>> Because I highly doubt many people would accept that definition.
Those are forks. They are not developed widely, but it's a fork. It is a divergence, a fork, from the main project. You can even have a private fork. The divergence don't have to exceed a certain threshold of changed lines to be a fork, a divergence.
Look up "divergence" and "fork" in any dictionary you want. There is no threshold. Any changes create a divergence or a fork.
Again, wikipedia calls "branches" forks. What you call forks and what you call branches are called forks by wikipedia. These all are a "forking" from a base code base.
Jose_X
2008-11-22 00:04:14
Perhaps we should talk about the divergences that go-oo is building up. Will everyone be happy with "divergence".
Of course, "fork" is going to be used and people will once again get angry......
Jose_X
2008-11-22 00:07:47
I got forked today.
Dan O'Brian
2008-11-22 00:17:39
Not that understood why you guys labeled it as evil before, but now it makes even less sense to be against it.
Oh, wait, I forgot... you guys have double standards wherever Novell is involved. If X does it, it's ok as long as X doesn't equate to Novell. Right. I forgot about that little loophole in your world of logic.
Victor Soliz
2008-11-22 00:46:31
Cor
2008-11-22 07:13:07
Dan O'Brian
2008-11-22 14:42:37
Has he posted the article he mentioned yesterday? Not as far as I can tell.
Roy Schestowitz
2008-11-22 14:52:15
Dan O'Brian
2008-11-22 15:00:22
In either event, I believe I found the article you were referring to which is just more misinformed tripe about Moonlight.
You don't even get the basic facts right.
Roy Schestowitz
2008-11-22 15:04:04