One of the most prolific and active developers of FreeBSD, Matt Dillon, responsible for a lot of the latest advances in the FreeBSD kernel, lost his right to commit code to the source tree. The FreeBSD Project didn’t give a solid reason on what lead them to this decision. Last year, OSNews interviewed Matt about his (truly immense) work on FreeBSD.
Matthew Dillon Lost Commit Access to FreeBSD
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Eugenia you are genius. Thanks for bringing up such yummy open source quarrels and stuff. Shows us that geeks don’t need sitcoms, just mailing lists and cvs logs
This appears to be a symptom of the increasing fallout within the FreeBSD core team now that the project lacks a clear leader. I hope the public backlash from this will be enough to have this reversed, regardless of the surrounding circumstances. This has to be one of the most boneheaded decisions I’ve seen the core team make.
Shooting the messenger eh?
The FreeBSD Project didn’t give a solid reason on what lead them to this decision.
Really?
The short of it is that Matt was unable to treat many of his fellow developers with the civility and respect that they deserve.
Seems like a pretty solid/clear reason to me.
Adam
I would accept this answer if it was a statement from a company who would like to not give out more details.
But FreeBSD is not a company.
A more detailed reason and maybe a timeline would be better. Users DESERVE to know what the heck has happened, when they lose Matt, one of the **most** important contributors in the whole history of FreeBSD.
Matt is not just your next door wannabe-hacker.
The short of it is that Matt was unable to treat many of his fellow developers with the civility and respect that they deserve.
Yes, aside from the fact that most people who have had dealings with Dillon will tell you that he was usually quite friendly and helpful. This is certainly a series of events that has occurred behind the public face of the project, and Core’s secretive nature about this has resulted in some backlash in the community. After all, open development doesn’t necessarily end at the code. There is a large community who this decision affects, and for one I am put off by the way it is being dealt with currently.
Matt is not just your next door wannabe-hacker
and
Yes, aside from the fact that most people who have had dealings with Dillon will tell you that he was usually quite friendly and helpful.
Which may be the reason why Core has chosen to not potentially embarass Matt with the specific details.
Users DESERVE to know what the heck has happened, when they lose Matt, one of the **most** important contributors in the whole history of FreeBSD.
No they don’t… Core really has no obligation to any FreeBSD users outside of the stated goals of the FreeBSD project. Core is a democratically elected group of representatives of all the FreeBSD committers. If those committers don’t like the way Core is running things, they have every right to vote them out.
Adam
Basically, the FreeBSD core team is playing by the “can’t we all just get along” philosophy. I’m sure there was probably a better way to handle this.
But I guess the core team would rather put the whole FreeBSD project danger because a few developers were personally offended by some of Matt’s comments. I particularily liked how one message in the archive from a core team member said “I have seen that side of Matt too (the helpful and very nice side), but unfortunately, he isn’t always like that.” Tell me… Is there anyone who is always like that? Everyone has their bad days.
How much productivity did the FreeBSD project just lose by shit canning their #1 developer? All because a few people took personal offense at the way he talked to other developers some times. Basically, the core team is going to let their personal issues cause the decline of FreeBSD. Great…
> Core really has no obligation to any FreeBSD users outside of the stated goals of the FreeBSD project
That doesn’t give them the right to boot out a so important developer without giving a more refined explanation. Heck, if that was a political party, they would still issue a statement explaining a few things.
No need for snobbism in the Freebsd project please. These people were not voted up there just to behave like aristocrats.
“No they don’t… Core really has no obligation to any FreeBSD users outside of the stated goals of the FreeBSD project.”
Yes, they do. Because a lot of users are also volunteer developers. I have ports in the ports tree for example. I have binary software packages I created on the official FreeBSD CD distribution. Yes, I think the core team owes it to me and other volunteers who contribute code and ported software to the project to let us know what happened. This is our project too. We contribute software that ends up on the official CDs. It’s not just the pet project of the core team. They owe it to me and other volunteer contributors to let us know why they make the decisions they make.
That doesn’t give them the right to boot out a so important developer without giving a more refined explanation.
Yes, it does. Again, they have no obligations to any FreeBSD users outside of their stated goals. You may not like that (and you have every right to dislike that), but that’s the situation. You said it yourself: they are not a business, which means they have no obligations to a paying customer.
These people were not voted up there just to behave like aristocrats.
Correct, they were voted up there to handle situations such as this in whatever manner they see fit. Which is what they’re doing. If you don’t like it, you should feel free to become a FreeBSD committer and then vote during the next election.
Adam
This is our project too.
Then vote them out.
Adam
No, you don’t vote them out. You just change projects if you don’t like their (democratic
) system.
Linus has never booted out like that an _important_ developer without giving a reason.
Or, you go back to Microsoft Windows, and you don’t care who gets sucked. You just use the damned thing and you stop worrying about such incindents.
“Then vote them out.”
The vast majority of software contributors are not commiters, and therefore have no voting privilages. So they don’t have the option of voting out the core team.
Eugenia: “That doesn’t give them the right to boot out a so important developer without giving a more refined explanation.”
AdamK: “Yes, it does. Again, they have no obligations to any FreeBSD users outside of their stated goals.”
I think this is being argued from the standpoint of it being a moral imparitive, which is certainly something you can’t argue against.
I’d liken it to Bowling for Columbine, where the movie’s creator took two victims of Columbine to KMart headquarters, two kids whose bodies still contained bullets purchased at KMart, in one case lodged between his aorta and spine, and requested that KMart stop selling bullets. KMart was under no legal obligation to stop selling bullets, but when the creator and the two boys showed up with two large bags full of bullets they had purchased at a nearby KMart, KMart had an emergency meeting and called for the removal of all bullets from KMart stores within 90 days.
Whether the FreeBSD core team will buckle under the pressure of a public outcry for the details of this incident is another matter entirely. I think if the mailing lists were to become saturated with requests for what happened, eventually they will disclose the details.
No, you don’t vote them out. You just change projects if you don’t like their (democratic
) system.
Next time my elected representative pisses me off, I won’t be moving. Instead, I’ll be sure to help vote them out during the next election 🙂
Linus has never booted out like that an _important_ developer without giving a reason.
This isn’t linux though.
Or, you go back to Microsoft Windows, and you don’t care who gets sucked. You just use the damned thing and you stop worrying about such incindents.
And those are certainly options for you.
Adam
“No, you don’t vote them out. You just change projects if you don’t like their (democratic
) system.
Linus has never booted out like that an _important_ developer without giving a reason.”
I wonder if that’s what Matt will do? Take his substantial talents to Linux now that FreeBSD core team doesn’t want his skills anymore?
It might be where I start sending my code if FreeBSD core team doesn’t get off their high horse.
“Next time my elected representative pisses me off, I won’t be moving. Instead, I’ll be sure to help vote them out during the next election :-)”
That’s assuming you have voting privilages. As I said, a lot of volunteer developers don’t because they are not commiters. The vast majority are not commiters.
It might be where I start sending my code if FreeBSD core team doesn’t get off their high horse.
I just don’t see this high horse you’re talking about. Do you disttrust core that much? Would you prefer that they release potentially embarassing, spiteful, and hurtful information about Matt?
Adam
> I wonder if that’s what Matt will do? Take his substantial talents to Linux now that FreeBSD core team doesn’t want his skills anymore?
Matt is already part of the linux kernel. He has contributed there as well. He might switch altogether now.
BTW, these two conversations found here ( http://docs.freebsd.org/cgi/getmsg.cgi?fetch=92148+0+current/freebs… ) are VERY interesting:
“I have held the notion, from the very first day of my experience with FreeBSD, that the Project belonged to everyone involved. Core, Committers, Developers, Testers, and Users alike. By and large, this notion has been upheld by everyone I have had the opportunity to communicate with. Folks involved with the project take a special regard to the achievements the project has made.
Many people are now objecting to the removal of a valuable
contributor to the project. People who quietly participate in the background, without fuss. Their concern is not sport or muck-raking, but genuine concern for the well being of the project. To simply ask them to be quiet and trust that Core, like a wise parent, will know what’s best is degrading at best, and an insult to them.
Core is a group of very intelligent and hard working people, however the project does not soley belong to them. Locking out the rest of the community on an important issue such as this alienates many who may otherwise support Core’s actions. Obviously a great many people believe that they have a right to determine what’s best for the project, exclusion from Core not-withstanding. At the very least, all that they ask is for full information on an important issue like this.
If nobody from the inner circle will explain why such a move as this was made, many project participants will reconsider their involvement with the project. Secrecy within Core, membership changes without explaination, and cover-up of the events afterwards do little to convey an image of stability where itis needed most.
If nobody is willing to explain what happened, and why, perhaps it is time for the project to reconsider it’s philosophy as a project that belongs to whomever wishes to participate. Perhaps it’s time to redeclare the project as the domain of those who sit in the positions of influence. I don’t believe this to be the case, but events like this may persuade others that this is indeed the current state of affairs.
Regards,
Benjamin Krueger”
And then the FreeBSD -core guy replied:
“We’ve already given an explanation. The problem is, those of you in this thread don’t think it’s enough.
Yes, we could do our dirty washing in public. But that wouldn’t be fair to dillon. What I can say is that yes, I too have experienced dillon the way you describe him, friendly, helpful and effective. Unfortunately, he’s not always like that.
Consider that this was an extremely difficult decision to make. The fact that it was unanimous–the first time, I think, that we’ve been unanimous in anything–should give the clue that whatever the problems were, we considered them very serious, and they were impacting the project as a whole. In addition, we didn’t see any prospect for
improvement.
I expect that won’t satisfy you. Sorry about that, but I can’t see a way to satisfy you without being unfair to Dillon.
Greg
BOTH are right. BOTH.
Which means exactly one thing: The situation was handled wrongly by the -core team from Day 1. They should have issued an agreement with Matt that “Matt is resigning for personal reasons”, instead of all this that has happened and will continue to happen right now. Yes, that might have been a cover-up, but IF Matt would agree to this statement, this all wouldn’t be a public trouble now.
As I said, a lot of volunteer developers don’t because they are not commiters. The vast majority are not commiters.
And they are welcome to leave. If core thinks that this is going to happen on a massive scale, they may change their mind about posting a more detailed reason. However, this was still their decision to make.
Adam
“I just don’t see this high horse you’re talking about. Do you disttrust core that much? Would you prefer that they release potentially embarassing, spiteful, and hurtful information about Matt?”
I would prefer that they release honest information about why they chose to get rid of their most productive developer (by wide margin). I would prefer they release accurate information. And Matt should know as wells as anyone else that when you have a position like that, you should expect that your actions and such might be subject to public scrutiny. So I don’t see the issue here.
They should have issued an agreement with Matt that “Matt is resigning for personal reasons”, instead of all this that has happened and will continue to happen right now.
You’re assuming, of course, that Matt would have gone along with that.
Adam
Yes.
I would prefer that they release honest information about why they chose to get rid of their most productive developer (by wide margin).
You should take a look at the commits leading up to this. Someone posted them on the freebsd-chat mailing list. In fact, recently there has been very little commit activity from Matt.
Adam
I’m interested in what Benjamin Affleck has to say about this (j/k)
Although linux keeps getting brought up here, why wouldn’t he more likely go to the netBSD or OpenBSD teams? Or slightly change the mascot and call it DareBSD?
Maybe they didn’t mind losing him because if he goes to another BSD project, they can re-use his code. However, if he goes to a GPL’d project, they would lose that oppurtunity.
Who want’s to be he goes to work for Apple on Darwin/OSX?
When Theo was booted outta NetBSD despite *his* massive contributions to the project. There, as here, the reason cited was a lack of “civility”. I wish Matt the best in his future endeavors.
Eugenia has a good point. If core didn’t want this to happen, they should have said Matt was resigning for personal reasons. They shouldn’t have said they decided to revoke his privilages and then refuse to say why they did it.
Basically, core needs a decent PR officer, since obviously no one currently on the core team knows how to handle PR issues/
What makes you think Matt would have gone along with this? This is a group of individuals who unanimously decided to remove his ability to commit.
In fact, what makes you think that core didn’t actually make that offer to Matt?
Adam
If they made the offer and Matt refused, then the core team would be perfectly justified in telling their user base why they decided to get rid of Matt. They would have offered him an alternative and he turned it down.
If they made the offer and Matt refused, then the core team would be perfectly justified in telling their user base why they decided to get rid of Matt.
Or maybe they’d want to take the high road (unlike Matt) and not air any dirty laundry about Matt.
Adam
Two things stick out in there:
1. Matt can defend himself if he needs to. If he is so innocent then why hasn’t he defended himself yet?
2. It says clearly in there that his commit bit has been removed in the past
It seems to me that core doesn’t seem to be too worried about losing his talent. Yes, it’s unfortunate but it almost seems as though he’s already been replaced.
I also think some of you are over-exagerating Dillon’s worth. I highly doubt you have counted all of his commits and compared them to that of others. I’m not saying that he isn’t a very influencial developer for FreeBSD I’m saying you DON’T know the facts yet.
So lets quit making stupid assumptions. Especially stupid assumptions like “FreeBSD is going to crumble and die now”. Look at what happenned to NetBSD when they blocked Theo’s commit access — we got OpenBSD out of the deal. So in the end it was to the benefit of everyone. Sometimes you just can’t force people to work together. It’s obvious that Theo has been a heck of a lot more productive working on his own with his own project.
I just refuse to sit back and judge the core team for their actions when, frankly, I know very little about the situation at hand.
I believe that this is a difficult situation to handle. If they give a reason, which seems spiteful towards Matt, they get blasted for being pricks. If they don’t give a reason, they get blasted for being an elitist aristocracy.
They had a job to do, they did it and, IMHO, they have no moral obligation to anyone but themselves and the individuals they represent.
Adam
Dunno .. but it seems this Dillon guy could be difficult. No matter how talented you are, or how valuable, you can’t just go about talking to be anyhow. If that’s why he was booted out, I would say that’s fair enough.
As for explanations, well, do they have to be that open? I guess they are following the spirit of the BSD licence
I give the over/under of the details coming out in 5 days. Only if Matt doesn’t want any of this to get out (due to embarassment or to protect a positive reputation, for example), we won’t see much of anything.
If he feels he was wronged, we’ll see a deluge of information coming out.
I’m betting on the latter.
The PR-correct move would be to wait until there was a press release sent, and all of core agreed on what they would say. Then the information would be announced in concentric circles.
IIRC, they could have restricted his commit access by scrambling his password. Then they could revoke the bit after the PR release.
“It seems to me that core doesn’t seem to be too worried about losing his talent. Yes, it’s unfortunate but it almost seems as though he’s already been replaced.”
Maybe they should be worried about losing a lot of other talent. It seems the majority of users think Matt was a good guy and are not really happy with the core team’s decision. A lot of people may just decide to jump ship and go somewhere else if they don’t get a better explination out of the core team.
“They had a job to do, they did it and, IMHO, they have no moral obligation to anyone but themselves and the individuals they represent.”
Once again, I think the core team has an obligation to all of the volunteers who contribute software to the FreeBSD project. If the core team thinks otherwise, than I may have to rethink where I contribute my work. I’m sure a lot of other volunteer contributors feel the same way.
Maybe they should be worried about losing a lot of other talent.
And perhaps they’ve decided that it’s worth the risk.
Adam
Uhm, uhm… Matt Dillon, is he the same guy? the actor? lol.
Why the heck do people believe that because they use open source or FSF software or BSD license they are entitled to stuff?
I use the software so I am entitled to know why they decided to take away commit priveleges to a developer?
Bull.
An official explanation to true contributors not just folks with commit priveleges is probably warranted.
However, just because someone creates some tool to cover some intellectual itch they have does not instantly entitle people to a bunch of other things.
People get free software off the web. But that is not enough.
If people have to compile it, they fuss about that. If the developer is nice enough to make rpms and even point people to the dependencies through links, people still complain about the dependencies and about dependency hell to the folks on the devel list. Not a quick comment or a detailed bug but an absolute bitch fest about how the rpms suck.
If they provide a mozilla style installer, people turn around and complain they are no rpms.
That is not all though. Just because someone created a quick tool for themselves and they posted on their website does not entitle somebody downloading it to a complete help system, manual, tech support or actually any freakin’ thing else. They do it in their free time and they give the bloody software away!
Why in the heck do they owe the == users == anything else?
The developers deserve a proper explanation probably. But the users — come on.
It is so funny that no one would demand an explanation from a private company but because it is open source the developer mailing list and bitch fights have to be in the open?
I don’t buy that at all.
Actually he answers that question, and others at
http://apollo.backplane.com/ . Amiga users know him from way back.
Any user which switches systems without knowing the entire story is an IDIOT. None of us know the details so lets quit making stupid assumptions.
Simba: As previously stated, Dillon hasn’t really made any major contributions in the last 12 months. He’s only really been doing maintenance.
Lets try and remain reasonable here. I see a lot of irrational and impossible-to-backup claims being made with little or no backing.
Why the heck do people believe that because they use open source or FSF software or BSD license they are entitled to stuff? I use the software so I am entitled to know why they decided to take away commit priveleges to a developer?
Bull.
It doesn’t matter whether it’s free or paid. All stakeholders have the right to know details if it involves the success, failure, or direction of a project. One thing that does concern me about this is whether a fair discipline and grievance procedure was followed. This last one is a worrying issue for potential contributers, and given the open nature of the project makes it a public issue in any case.
As a rider to my last comment, I’d like to add that the poor handling of this issue, including the “we’re not saying, though we’ll drop some biased hints” comment, shows a lack of management expertise in the core team.
You almost hit it right on the head.
If you are a real stakeholder (ie a contributor to the projects in question), I already said that those people should receive some proper explanation.
However, as a user, I do not believe I am entitled or should really even care why the people actually doing the work made a decision on the commit rights of one contributing developer who’s sheer number of commits have decreased a great deal in the last year anyway. It is pure nonsense to desire all such developer fights to be handled and handed over to the public.
You gave me something for free now I have rights to know why you won’t let someone commit changes directly?
Once again, that is pure crack.
You gave me something for free now I have rights to know why you won’t let someone commit changes directly? Once again, that is pure crack.
No it’s not, and without anything more than a simple assertion off you, there’s no good reason to change think any different.
Whether free or paid, people who use a product are customers, who in turn are stakeholders, who in turn have a vested interest in the quality and direction of the product. In the absense of any formal and verifiable discipline proceedure, all stakeholders gain the natural right of inquiry.
Good, now maybe he can get back to all those Amiga projects he left behind.
Jim Steichen, Author of AmigaTalk
There is a natural right of inquiry into the private dealings of a group of people creating software?
Hold on let me fire off a letter to my congressman. I do not know of any legal let alone philosophical reason why I should have such a right.
Just because I use a wrench made by a company designed by a guy does not mean that I have a right of inquiry into why that designer no longer works on making wrenches. Just because I play a free to use java game off of a website I have no right to know why the developer of that app moved onto another project and no longer updates that little java applet.
Just because you download a piece of software from the net or some place else does NOT make one a stakeholder in the dealings into the work of the people making that software.
Also I looked through the definition of the word customer and I think that you labor under a misconception about the word.
“Webster’s Revised Unabridged Dictionary (1913)”
Customer Cus”tom*er (ku^s”tu^m*~er), n. A doublet of
customary, a.: cf. LL. custumarius toll gatherer. See
Custom.
1. One who collect customs; a toll gatherer.
2. One who regularly or repeatedly makes purchases of a
trader; a purchaser; a buyer.
A customer actually buys the software, makes a purchase of the software in question. The person who uses a piece of free software are not customers and unless they contribute back to the community they are not even stakeholders. They are users, plain and simple.
You have not explained in any real fashion why someone who merely uses a piece of software has some natural right to know why the developer of that software makes a decision on who he/she/they will allow to make commits.
Just because you say you have a right does not make it so.
In fact, I view this as just another reason why I cannot imagine why good programmers bother giving back to the community.
All one has to do is post a piece of alpha code that you use on your machine on a website (or give to a friend that ends up doing this) and suddenly everyone has a “right” to bitch, moan and complain about everything having to do with the project and the code and how it is packaged and the documentation regardless of whether or not they have even touched the bugzilla setup you took the pains to setup. If you tell people why you don’t let so-and-so commit changes, then you are defaming and flaming that person online. If you do not say anything, then suddenly you are infringing on some “natural right of inquiry” which simply does not exist.
If this person was unfairly censored in an academic institution, or unfairly fired from a place of work, there is such a right of inquiry. If you are Matt Dillon, then you naturally have a right of inquiry to know why you are no longer allowed to make commits (but that is not the same as being censored really since he can still contribute code for review to be committed by others). Even as a contributer I can see where a right of inquiry might possibly exist but as a user?
If a academic was censored and he wanted to know why due to a right of inquiry, I can see that.
If another academic wanted to know why that original academic was censored I can possibly see that but it might or might not be embarrassing and an unnecessary infringement on privacy rights.
If the guy who checks out a book at a library printed by the university, wants to know why that academic was censored I do not necessarily see how that person has any rights of inquiry.
There is no gag order on Matt Dillon. If the censored soul wants to let the world know the deal, then he can go ahead. However, I do believe that there is a place and a reason for private developer lists and this is one of them. People should have a right to have a private conversation about their work and unless the person inquiring actually contributes to the progression of that code they have no natural right of inquiry.
Eugenia writes:
Linus has never booted out like that an _important_ developer without giving a reason
He’s never been in that situation.
He has never given anyone else commit access to the “blessed” kernel – except when it is in maintenance mode, and then only to the 1 maintainer
You can’t be kicked out of “linux”, because you can’t ever get into it. The best you can do is run your own tree and submit patches to Linus.
Matt can still do that in FreeBSD, he just no longer has write access to the main tree.
(Based on past history) he will never get any greater set of priviledges to the linux tree than he currently has to FreeBSD. And nor will anyone else unless they are either Linus, or his nominated maintainer.
Actually, Andre Hedrick (the IDE maintainer) was kicked out of the Linux project for several months… then they had to take him back after whoever took over the IDE code trashed it.
Great post Johnathan Bailes ! You hit the nail on the head on this issue.
Linux’s VM is inferior to FreeBSD’s. In all other aspects, esp usability, core approachability, and organization, Linux has already overtaken FreeBSD. e.g. Linux kernel source tar.bz2 can be downloaded at http://www.kernel.org – simple. FreeBSD is….well try /usr/src/sys after you’ve installed from the CD – not so simple.
Now the FreeBSD’s core VM person is kicked out. He is already active in Linux. Which means that he will take his excellent work to Linux. Therefore Linux’s last weakness i.e. VM will get fixed. Threading, SMP, and scaling is getting fixed in 2.5 series. Clustering is getting taken care of OpenMosix.
Somebody please give me 2 good reasons why FreeBSD is going to matter beyond some amount of time. Mind you, I am a FreeBSD AND Linux kernel developer who prefers FreeBSD’s response characteristics under heavy system load, but it seems like Linux’s community is more fun, more open and more responsible towards its users. And an open-source OS owes EVERYTHING to its users. e.g. FreeBSD has more users than PicoBSD and hence FreeBSD is more important than PicoBSD.
As for why it matters – well, when I tell my Boss to use FreeBSD or Linux as the OS for product development and deployment, I am implicitly assuring him that there will be no open-source monkeys wrecking the shop 6 months down the line.
So if the core tells users that their opinion is irrelevant, then the core’s efforts will eventually fail cos no one wants to support such an OS in their official jobs where asses are on the line and families need to be fed.
The excellent FreeBSD VM-system is the work of John Dyson (and David Greenman) not that of Matt Dillon. Matt only maintained Dysons work.
Linux kernel source tar.bz2 can be downloaded at http://www.kernel.org – simple. FreeBSD is….well try /usr/src/sys after you’ve installed from the CD – not so simple.
WTF? You can download the FreeBSD kernel source from http://ftp.freebsd.org, or use CVS.
As for why it matters – well, when I tell my Boss to use FreeBSD or Linux as the OS for product development and deployment, I am implicitly assuring him that there will be no open-source monkeys wrecking the shop 6 months down the line.
And perhaps that’s the very reason they removed Dillon’s commit bit.
Adam
Well, if opensource continues this way; big OpenSource projects, like FreeBSD, start fighting with each other, throwing people out and removing the dialog from their list, opensource is going to die.
This happened again with NetBSD and Theo, the openbsd project leader.
I believe its time opensource projects to re-shape and start working AS A PROBER TEAM. Opensource will suffer again from this, we all know it.
Hm…
tpv on Linus: He’s never been in that situation.
Benevolents dictators has its advantages
Charles:
Whether free or paid, people who use a product are customers, who in turn are stakeholders, who in turn have a vested interest in the quality and direction of the product. In the absense of any formal and verifiable discipline proceedure, all stakeholders gain the natural right of inquiry.
Excellent point! The Core team has to give some excuses. Even an empty one is better than no-one.
—
And what’s the bad of forking? People seem to fear fork as a waste of energy/time. I wonder if OpenBSD, NeXTStep and Darwin are waste of energy/time, not talking about the most famous fork of all time – MS-DOS – because it’s not Open Source
In the absense of any formal and verifiable discipline proceedure, all stakeholders gain the natural right of inquiry.
Core has the natural right of refusing to give any more of an explanation than they have already.
Adam
I think the situation was explained pretty clearly.
**They felt Dillon was abusive towards other developers and hard to work with.**
This is the reason they removed his commit bit. Now if it just so happens that this _isn’t_ true, then Dillon can defend himself and say otherwise. Then and _only_ then can you bitch about core’s decision.
This happens all the time in the real world. You be an asshole to your fellow employees and your boss fires you (not to make premature assumptions about Dillon’s personality). This doesn’t just happen in the open source world. Lets quit spreading FUD around, k?
“Adam”
It’s not necessary to sign your name with each and every post, especially when your name is given immediately above those posts.
Whatever. As you can see, I’m easily irritated. Are you one of the people Matt D. had to deal with on a regular basis?
…Linux, TeX, and OpenBSD are development autocracies–managed and maintained by individual personalities (Torvalds, Knuth, and de Raadt respectively). There’s lots of henchmen and viziers in those but they really revolve around their central authority (see also Perl and Python). There are many open source projects like these. But the projects that I think will last longest are not ones that revolve around an individual. A lasting project is going to require a constitution with a checks and balances system, possibly even separate legislative, executive and judicial branches. Some of these projects sort of have something like these already (GNU and Debian) and hopefully they’ll survive the loss of their center because what seems to work with one personality in charge isn’t necessarily stable when other personalities have to take charge.
I expect many open source projects, no matter how good, are going to suffer from ozymandias syndrome when their creators abandon their projects for one reason or another. It’s that or succession wars (of the largely pointless, flame variety). So, in a way, the more incidents of NetBSD and no FreeBSD have of putting project ahead of personality the better they look as lasting software development communities.
It seems that people tend to overlook the fact that Open Source is a social experiment, not just (or even mostly!) a software development experiment. I think that developers and developer communities are going to have to start thinking about this sort of thing carefully. Of course the only way to find out what will really last and what won’t will be to wait and see.
saying WTF does not automatically make your opinion correct. Its a childish way of throwing a tantrum when you are not able to make your point any other way.
ftp://ftp.freebsd.org/pub/FreeBSD has 50 different directories under it. Compare that mess to the simplicity of http://www.kernel.org – Linux is more approachable.
And yes, users like me DEMAND answers and have a right to them. I have been working with FreeBSD since 1995, which might be longer than some of the core team. Being core doesn’t make them better or smarter than any other user, it only means that they are under a social contract to interact with the rest of FreeBSD’s users. without users, FreeBSD is nothing, e.g. FreeBSD v/s PicoBSD.
And Mr. WTF Adam K WTF, please drop the attitude. You come across as if you’re licking core’s ass. Dillon might wrong or core might be, I dont know. But we need answers. Right now it looks like core is hiding something. Maybe its Dillon’s mistake. Either way we need to KNOW. Else I cannot support a movement whose core is elitist and secretive.
Don’t quote me, but I don’t think this is the first time Matt has lost his commit priveledges. And he’ll probably get it back. Thanks for trying to still up a shit fight where there is none.
Perhaps he can join the OpenBEOS development team. After all he is probably itching to work on another awsome operating system
Cheers,
Clint
While it’s understandable that the FreeBSD core developers don’t want to air their dirty laundry in public, the fact is that we’ve already caught a glimpse of the laundry and can still smell it. Worse, the incomplete picture that we do have is going to spur more distrust. The core team has already hinted that Matt Dillon was being discourteous or something like that, yet others have dealt with Dillon, found him a fairly nice guy. Naturally, those others are now wondering why the Dillon with whom they worked is so different from the picture of Dillon that the core team presents, and are wondering if the core team is somehow biased against Dillon. Right or wrong, people are wondering what’s going on–and they are suspicious.
At this point, the core team might as well come out with the full story. We’re already imagining some ugly soap opera scenarios ourselves, and they are probably all worse than the truth. To not come clear will only make the core team look worse, even if they were justified in their decision.
ftp://ftp.freebsd.org/pub/Fre eBSD has 50 different directories under it. Compare that mess to the simplicity of http://www.kernel.org – Linux is more approachable.
Really? So if I want to know more about Linux, should I go to linux.com? linux.org? debian.org? kernel.org? linux-mandrake.org?
With FreeBSD, it’s quite simple: http://www.freebsd.org
Now which is more complex?
And yes, users like me DEMAND answers and have a right to them.
You can demand all you want. Core is still under no obligation to tell you any more than they already have (which is more than anyone outside of the committers are actually entitled to).
Oh, and if you don’t like my attitude, don’t read my posts. It’s quite simple.
Adam – I was talking of access to the kernel source code. http://www.kernel.org solves that without confusing a user. Dont muddy the water by talking of http://www.linux.com and other nonsense.
as for confusion in the bsd ranks, there’s netbsd, openbsd, freebsd, picobsd, bsdi, bsd4.4, other legacy bsd stuff. heck for all i know maybe openbeos might also be using bsd stuff.
at least in the linux ranks, you CAN guarantee the kernel. the *bsd kernels are different from each others.
And Adam, please learn to shut up
Have you untarred the linux kernel tarball yet or are YOU just trolling?
Adam – I was talking of access to the kernel source code. http://www.kernel.org solves that without confusing a user. Dont muddy the water by talking of http://www.linux.com and other nonsense.
It’s hardly muddying the water. It’s a valid concern. You said “Linux is more approachable”. Maybe you meant “Finding the linux kernel source code is easier,” but that’s not what you said.
Assuming, however, that’s what you meant, I might point out one slight difference which explains why it’s easier to find the linux kernel than the FreeBSD kernel: FreeBSD is developed as a whole. Go to the linux-mandrake website and try to download *just* the kernel. Go ahead and let me know how you make out.
at least in the linux ranks, you CAN guarantee the kernel.
Since when? There are probably as many versions of the linux kernel as there are distributions… RedHat has their own, Mandrake has their own, etc…
And Adam, please learn to shut up
It’s funny that you’d criticize my “WTF” (a very common phrase used as an exclamation these days) as a childish tantrum, yet you’d resort to the juvenile tactic of telling your opponent to shut up.
Adam
Users have no need for kernel sources, trust me.
People who have got even the faintest need for source code must at least be competent enough to find the source code on the FreeBSD FTP archive, don’t you think?
This is downright silly. The last thing you could accuse any BSD of is lack of organisation, on any level.
There are three main BSDs plus BSDi/BSDOS. Bloody confusing, isn’t it. Compare that to the mess that is Linux. I sometimes went to Linux.org many years ago. It turned out that it wasn’t the “official” Linux site, despite the deceptive URL. And it was rather sparse in its information. Kernel.org? Well, that’s really the most obvious place for a Linux kernel, isn’t it?
Note that this is in no way an attack on Linux. It’s an attack on CitizenKane’s trolling.
Commit-rate is not the thing one should use judging one’s project activity. Matt did fixed several nasty bugs in TCP stack very recently, and, most notably, he was probably the only one who did very in-depth and extensive review of Jeff’s new scheduler work, finding/fixing bugs again! The man is being superb as bug-hunting machine! Moreover, Matt is currently in the process of writing his own scheduler AFAIK, and should things go on fine, FreeBSD in some near future would be full of a great improvements and stuff came directly from Matt! And they’ve screwed him. Ass-holes.
maybe i need to talk about mbuf utilization in *BSD or about how Linux’s Netfilter macro NF_HOOK calls okfn
if nf_hooks is empty but translates into nf_hook_slow() for non-empty nf_hooks or under debugging conditions. Maybe I should compare ip_input in Linux to FreeBSD or comment on how NetFilter’s one-pass packet processing is faster/superior to other available solutions.
OK ? Not a troll. I’ve unzipped kernel sources for as long as I can remember, grokked *BSD and Linux, and can puke it all out in my sleep if I have to. No problemo.
By now, it is obvious that I am quite knowledgable about the internals of FreeBSD AND Linux kernels, as I have been saying. Which also explains my strong interest in this issue and my accurate comments about Dillon’s VM solutions.
I asked you to shut up cos you keep on raking up new meaningless issues without answering the important questions that people are asking. Calling me a troll is not going to answer my questions, or those of other people or those of my BOSS. You can please continue talking all you want.
Given the elitist stance that you (Adam K, James) and FreeBSD core seems to be taking viz. “you dont need to know jack shit about how we run FreeBSD”, when most of you are NOT more capable/talented than many of us non-committers, I think I will be choosing Linux exclusively in the future.
Have a nice day. FreeBSD just lost a supporter who has been around since 1995.
Have a nice day. FreeBSD just lost a supporter who has been around since 1995.
Bye.
Adam