Slashdot Log In
What's Wrong With the FOSS Community?
Posted by
ScuttleMonkey
on Wed Nov 22, 2006 05:43 PM
from the 500-words-due-next-tuesday dept.
from the 500-words-due-next-tuesday dept.
An anonymous reader writes "Patrick McFarland, one of the major Free Software Magazine authors, has completed his second article on whats wrong with the Free/Open Source Software (FOSS) community, and what we face in this world. He touches on ESR's Cathedral and the Bazaar essay briefly, and warns against cherry-picking style software development."
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Full
Abbreviated
Hidden
Loading... please wait.
What's wrong with Santa Claus (Score:5, Funny)
Also he drinks all the milk and eats all the cookies!
People.. the same as any community (Score:5, Informative)
The major difference between FOSS and other communities are that the people in a FOSS community share far fewer specific goals than other communities. Some people want something fixed **now**. Others want it fixed **properly**, no matter how long that takes. Others just piss and moan.
Parent
Re:People.. the same as any community (Score:5, Insightful)
Sad really.
Parent
Re:People.. the same as any community (Score:5, Funny)
I'm not broken, just unintelligently designed.
Parent
Re: (Score:3, Insightful)
No... wait...
The difference between the FOSS and commercial software worlds is that in commercial enterprises, in the absense of leadership, someone will be unilaterally appointed, not to lead, but to dictate.
In the FOSS world, if there's no leader, there's no leader. People will choose their own direction until they find someone they want to follow or find others wanting to follow them.
I think that's a major rea
Re:People.. the same as any community (Score:5, Insightful)
Which in turn may or may not be successful. The mambo/joomla mess illustrates that some forks work and you end up with two relatively strong branches. Go the other way, and a fork splits its community, diverts resources, and eventually kills off one, the other, or both.
And while no one wants a moron in a suit yelling at them, OSS developers are notorious for chery-picking the "cool" aspects of the project and ignoring others, and generally being insensitive to things like schedules and deadlines.
As to "winning", you have some strange definitions. Get an OS with more than a percentage point or two of the average desktop, and "maybe" you can start waving that flag. Utill then...
Parent
Re:People.. the same as any community (Score:5, Insightful)
A lot of people who are given the label "self-appointed dictator" in this realm are really just people leading by doing, but leading in a direction different than those doing the labeling would prefer.
The "I'll donate some of my time to some project" developers like the "cool" features, yes, and the real leaders will take what help they can get. As they do most of the work.
Most successful FOSS projects seem to be based around a core group of people whose prime driver is their interest in fulfilling their vision of what the result should be, assisted in small ways by a large group of vaguely interested people.
This is leadership. You can tell the difference between a leader and a director with a simple comparison: If the person would eventually/theoretically get the project done even if everyone else left, they're leading, and if they wouldn't get anything done when everyone left, they're not leading.
Of course, there's no reasoning with people like that... they don't give a fuck about what you want, they're blazing trail.
As to "winning", which do you think most people care about, their desktop, or the Internet it connects to? How many people do you know these days who can't just sit down in front of any computer whatsoever, log onto whatever services they need, finish up and walk away? There are a lot of them. The services they're logging onto are the "Network is the Machine" effect Microsoft has been fearing and fighting all this time, and that network is pretty much owned by FOSS.
Linux might not be on the desktop, but the desktop is becoming more and more "That virus infested annoyance you're forced to deal with to get on the Internet", and the Internet is FOSS.
Parent
Re: (Score:3, Insightful)
Lord Vetinari, in "The Truth" by Terry Pratchett.
I think he's got a point.
Common sense says (Score:4, Insightful)
Nothing in this piece convinces that common sense is wrong.
Re:Common sense says (Score:5, Interesting)
The piece seems to be claiming that good > mediocre > no > bad leader.
That's somewhat true, certain kinds of software and features just won't get done without a leader. That nifty little project doesn't need a leader, it'll get done because personal motivation is enough to get it done and it's small enough that a single person can handle the entire workload. Boring stuff won't get done no matter how grand the end result unless there is a leader to make sure it gets done, no one digs ditches for fun - even if the end result will be the panama canal.
Parent
I can pay the coder to do it. (Score:5, Interesting)
A leader without the ability to fire someone or give them a pay raise isn't going to be able to provide much incentive.
But with FOSS, I (the end user) can email the coder and offer to pay him/her to finish a feature I'd like or do some other boring job. And that is one of the great things about FOSS. Once I pay for it, everyone benefits from it (including me).
Try doing that with closed source products. You can't even find out the names of the coders working on it, much less contract them directly.
Parent
And I will wait for someone else to pay ... (Score:4, Insightful)
However progress will be slow because most of us will wait for someone else to pay for the changes we want. Most people will freeload if given the opportunity, Econ 101. Since you are reading this right now, I will thank you in advance for your future gifts to the community.
Parent
You know what the funny thing is? (Score:5, Insightful)
The bazaar model still worked when the pinnacle of software complexity were "cat" and "vi". That's it. It stopped working almost completely when complexity meant Open Office Org.
The Asperger's Syndrome kind of coder (and I'm one, so I can make fun of myself if I want to) which finds more joy in coding something cool instead of going out and flirting with a girl, also has a very narrow focus of attention and gets bored easily when he must deal with stuff either (A) outside that focus, or (B) which is basically homework instead of getting to the cool stuff. That's how we ended on the bad side of teachers in school, after all. Spending weeks understanding someone else's framework and code before you can even start on your cute "number paragraphs in Klingon" idea, is boring, and it's even more boring to understand and test all dependencies so you don't break something else.
So today in F/OSS the only ones making any progress nowadays are, sad to say, the Cathedrals.
Yes, everyone likes to use the Linux kernel and such as an example of why the Bazaar is strong, but have a look at the actual contributors some day. It's _not_ bored nerds like you and me working in their free time. Most of them are paid employees of Red Hat, IBM, etc. Linux as the work of bored nerds in their free time was a security shithole until Red Hat spent some real money doing a code and security review. And it was a joke in the enterprise arena until IBM started pumping some real money and formerly Cathedral-developped closed-source code into it. There's a reason why IBM looked like a believable target to SCO (as opposed to just a tempting target, by having deep pockets), and that's the sheer quantity of Aix code that IBM donated.
The same goes for OOo: practically all development is paid for by Sun, and it's bleeding Sun a ton of money. The same goes for Apache, which everyone uses as an example of why OSS is better than MS's software on a server: it, and most other Apache projects for that matter, is mostly IBM work. Go figure. IDE's? Both Eclipse and Netbeans are paid work by respectively IBM and Sun and a number of other corporate contributors. Compilers? You'd be surprised how much in GCC actually comes from Intel and the like. Browser? Mozilla was mostly paid work by Netscape, then AOL, and now it's mostly sponsored by Google. Etc.
So yes, as you aptly put it:
And that's why most of F/OSS nowadays is nothing more than a way for various corporate Cathedrals to pool their resources against MS. Sure, it's a good goal and I have nothing against benefitting from it. But let's stop pretending that ESR's Bazaar is anywhere _near_ relevant any more. The actual "Bazaar" projects are the thousands of unfinishet things on Source Forge that noone gives a damn about, either to help develop/debug or to use seriously or to pay the developper for features.
Parent
Re: (Score:3, Interesting)
That's exactly what I wanted to say.
I thought the article's comments about GNOME in particular were wrong. The problem with GNOME was not that it had no direction and therefore suddenly became bloated and unmanagable with feature creep. Far from it. It had a relatively popular leader who had an idea about how it should work. It gained feature creep because the "vision" of that OSS leader was to emulate a UI that itself was bloated, poorly designed, and suffering feature creep, both on the outside and the
In my opinion (Score:5, Insightful)
Re:In my opinion (Score:5, Informative)
Parent
Re:In my opinion (Score:5, Insightful)
A knowledgeable person who is simply inexperienced in an area will generally phrase a question better than a 12 year old kid demanding attention NOW.
"Gentoo is shit, it won't install why not?"
vs
"I attempted to install Gentoo on my computer (an aging P2 on an Acer motherboard) and came up with a number of problems during the install. It spent about 20 minutes compiling before it stopped saying 'The XYX system could not be compiled: missing file xyz.c'.
I tried looking around the furum but couldn't see where I am going wrong. Can somebody give me some assistance please?"
Parent
Re:In my opinion (Score:5, Insightful)
Ooops, I think you got that wrong.
There are a small number of very vocal people who are total assholes towards people.
Does it matter what the subject is?
Parent
Re:In my opinion (Score:4, Interesting)
Parent
Re: (Score:3, Interesting)
> from commercial closed source software.
Really? Because I use them on the same servers in the same data centers for the same purposes.
> The success of closed source / commercial software could
> simply be measured by the amount of money it makes for the
> creator.
No it can't. That's not success of the project, it's success of the product. It's a whole different question. Microsoft Bob was a successful project, because it did what it set o
Re:In my opinion (Score:4, Insightful)
Now, it may be that you lack the experience to discern what portion of TFM is the information you are seeking. If so, say so! Say that you have looked in TFM and not found an answer. Ask for help explaining specific parts of TFM, or ask for a more specific pointer to what part of TFM you should be looking in.
Reading TFM is an important skill, and one that must be acquired. If you have that skill, then there is no call for you (or anyone else, of course this entire post is directed generally) to go demanding that other people use their energy and time to do what you are perfectly capable of. If you don't have that skill, then the greatest ROI for people responding to your question comes when they encourage you to acquire that skill. If you have trouble acquiring it on your own, then generally you can still find someone who is willing to help you acquire it. But not many people want to spend their time and energy doing something that either you can do or that you should be learning to do, unless such an expenditure will help you learn to do it yourself. If you expect someone to put down little arrows on the ground in front of you when you are lost in an unfamiliar city, then you'd better have some cash in hand. Similarly, many distros offer paid support contracts.
When you spend 5 minutes saying exactly how to do something in detail, you are often setting yourself up to spend another 5 minutes saying exactly how to do something else in detail later. If someone figures out the answer themself, even if it is with guidance and aid (think Socrates), then they are much more likely to be able to figure out the next answer as well.
Required reading (or it should be): http://catb.org/esr/faqs/smart-questions.html [catb.org]
Parent
When did the community become an entity (Score:5, Insightful)
Hmm.. (Score:3, Insightful)
Nothing really is wrong except one thing. (Score:5, Insightful)
Not every software need will be be solved with FOSS.
There needs to be freedom to write Open and Closed source software. That is what bugs me are people that think selling a closed source package is evil. I just don't think that the FOSS model can work for every program.
Re: (Score:3, Insightful)
If you don't like it, don't patronize the people who enforce it (RIAA, MPAA, etc.). Nobody is forcing you to listen to music or watch movies either.
Re: (Score:3, Insightful)
F(L)OSS (Score:3, Insightful)
no leadership? (Score:3, Insightful)
Unlike in the Cathedral, the Bazaar has no official leadership.
Isn't this what enables FOSS? Most of the FOSS don't have official leadership (other than the creator of course
Re:no leadership? (Score:5, Funny)
Unlike in the Cathedral, the Bazaar has no official leadership.
Sometimes that is preferable to Archbishop Balmer flinging the Holy Chair of Antioch.
Parent
I read the article (Score:5, Interesting)
I challenge thee to summariser it.
This is what (/usr/bin/ots) a text summariser said (interesting to note it tents to focus on cathedral-style, bazaar-style, and gnome bashing)
A few years back, Eric S. Entitled The Cathedral and the Bazaar, he wrote about how the Free/Open Source Software (FOSS) community does what it wants when it wants to. In Cathedral-style projects, your not-so-friendly neighborhood PHB (fueled by the lies from various ugly hunch-backed minions), although wrong 120% of the time, says what goes in a project. Backed by the Free Software Foundation and the FOSS community as a whole, the GNOME project for many years just added lots and lots of feature creep and otherwise unnamed bloat.
The GNOME project lacked true vision for those years, and feature creep and other long term development problems rushed in to fill that hole. Problem is, many projects are just like GNOME. Incidentally, few Cathedral-style projects suffer from lack of vision: those that do simply die off and are never heard from again. Bazaar-style development allows projects to be in a zombie state for long periods of time, where it is vastly expensive for a Cathedral-style project to do the same. someone with vision (corrupt or not) would control a project, driving development behind it, and have the project reach goals in specific time frames.
Of course... (Score:5, Insightful)
Here's my rimshot: (Score:5, Insightful)
Re: (Score:3, Insightful)
Re: (Score:3, Insightful)
Interesting bit about XFree86 (Score:5, Insightful)
> way of everyone who did have vision.
That's rather well said. If you're the author of a successful open source project and you find yourself unable to keep working on it, do you have a duty to turn it over to the other developers for continued maintenance? I can't think of a reason not to, and if you don't, it'll either die or get forked, both of which aren't pleasant outcomes.
Coming along just fine (Score:5, Insightful)
Perhaps the real problem is the plethora of side-liners, pundits, philosophers, and magazine authors who have nothing better to do than sit around and draw erroneous conclusions. I call these people OSS arm-chair experts. We don't need 'em. Seems the people with most to say write the least amount of code. Maybe they should learn to program and get involved rather than digging too deeply into what's wrong. Be positive.
Re: (Score:3)
2. Ease of use isn't everything. In fact, it's 1% of everything at best. People here act as though a learning curve is the end of the world. It is, if your goal is market domination. That is NOT an FOSS goal. FOSS cares not one bit about market domination. It cares about powerful tools that can be made to do complex work that
someday ... (Score:5, Funny)
Someday pigs will fly
Someday hell will freeze over
Someday bears will be catholic and popes will shit in the woods
Someday poeple will stop using weasel words
This Blog Seems to be Spot On (Score:3)
HOWEVER - I think it is very good that such a review exists. As the benevolent dictator of my staff, I encourage ideas and help move software projects forward. I can learn from the FOSS community and their mistakes.
I certainly hope that the kernel development and many other such projects (KDE) follow this type of path.
Too violent? (Score:5, Insightful)
Off the top of my head... (Score:4, Insightful)
- Users, What Users?: Coding for yourself is nice, but if you want users to flock to your app, you might want to actually consider what they want. Don't bitch and moan at them when they offer suggestions, even if said suggestions don't fit your own personal vision, or even if they are downright stupid. That doesn't mean you have to implement them, but it means you have to be weigh them equally with your own ideas. Try to be inclusive and open to your userbase. "Go code it yourself" is a great way to keep OSS in the geek ghettos of the computing world.
- But It Looks Pretty: That's a snazzy looking interface you just whipped up, is it consistent? No? Does it follow standard UI principles? No? I'm sure people won't become frustrated and dismissive of your hard work. You can say that UI standards impinge on your freedom as a developer, but they make a user's life much easier, and makes people much more likely to actually use your software.
- Ask, Don't Beg: Asking companies and organizations to open code is nice and helpful, but be careful how you go about it. It can easily come across as "The OSS community could never dream of putting something like that together. Gimme!" Don't act like you *expect* the code, and that they are evil incarnate for withholding it. Don't make it seem as if the OSS community is incompetent and needs privately-developed projects turned over wholesale to get anything accomplished. Sure it helps a whole lot, but don't make it seem as though OSS is just mooching off the investment of others.
- Vendettas: If two projects can fight over something, no matter how petty, they will. Try coding, it's more productive and makes you appear like a mature, competent project that might help win over those hesitant to support OSS. Or you could just continue the pissing matches and flamefests over icons and licensing minutiae that could probably be settled if egos were set aside for a few moments. Public wars of words, endless forking....nothing gets accomplished but the stroking of egos. Well, except the whole "OSS developers come across as immature, childish amateurs" thing.
Mildly interesting, but shallow. (Score:5, Informative)
While, of course, the OSS community doesn't have a single vision for any piece of OSS software, quite a lot of OSS projects do and, as his story alludes, OSS projects that have a following but languish either for lack of vision or because the project owner has misguided vision—unlike closed-source projects which, while they may not tend to lack vision, are no less likely to have a misdirected vision than their open-source counterparts—can be rescued by forking.
And plenty of OSS projects do have a vision, direction, roadmap, etc. Sure, there's probably a lot of stuff that gets released under an open-source license (or straight into the public domain) because the author is essentially "done" with it and throwing it out to the community to do with what they will, but certainly open-source players like Apache, Mozilla, etc. have a vision for their main projects, and members of the community are attracted to and contribute to projects, no doubt, largely because of how they see the project's vision as compatible with their own. The "solution" McFarland offers is what it seems to me almost every major open-source project is already openly trying to do: allow the community to contribute, but institute a degree of top-down control in terms of timelines, roadmap, and assignments to make sure that the grunt-work necessary to have a polished project gets done.
I probably wouldn't call it "acting like the Cathedral", the openness of many successful projects to community process and innovation, while retaining a kind of top-down vision, is something of a synthesis: the Cathedral harnessing the energy of the Bazaar, the Bazaar borrowing the focus of the Cathedral. And you see something like it in the embrace by some commercial, formerly closed-source vendors of both open-source software and increasing community involvement. If I had to name the model, I'd call it the "Congregation" or "Assembly", a less-propietary Cathedral, a small portion of the Bazaar united by a common purpose and direction to accept, in the context of a project, some degree of authority and leadership (but not the exclusive ownership and control of the Cathedral.)
The article is all wrong. (Score:4, Insightful)
In OSS much more than anywhere else, the best floats on top. That's why Outlook mail sucks and KMail sucks considerably less. Linux works because NOBODY doubts that Linus is the chief, Blender works because NOBODY doubts that Ton is the chief, because they both do an excellent job at what they do: leading large OSS projects.
Of course there's weedy stuff in OSS that's buggier and more twisted than Autodesk Converter and Macromedia Director together, but that sinks to the lowest bottom, and does not get pushed onto the market by monopolies and marketing budgets of galactic proportions (Windows XP anyone?).
The article is bogus and has it all backwards. I want my 5 minutes back.
FOSS reminds me of Canada (Score:3, Interesting)
It does somethings better, others worse. It will never admit what it does worse or will even defend it as a strength.
It's arrogant and sanctimonious even though it often has its heart in the right place. Other times it's naive in thinking that because it believes *its* way is right, it *must* be so.
It mocks the U.S. as backwards, even displaying a near pathological hatred for it, yet it secretly wishes it could hold the same lofty perch.
Now replace "It's/It" with "FOSS" and "U.S." with "Microsoft".
Frankly, what turns me off about the FOSS community in general is it reminds me of the science acolytes in the recent South Park episode when Cartman traveled to the future. Meet the new boss, same as the old boss. If Linux were king it would be Microsoft redux.
Feel free to mod me down, I have plenty of karma to burn.
As opposed to what hard data, precisely? (Score:3, Interesting)
Some things to quantitatively evaluate are:
Failures per release, time to release, bugs discovered, function points derived, cost-benefit, TCO, testability, verifiability, number of severity one bugs, number of severity one bugs never fixed, number of abandoned projects, time to next version, rate of customer abandonment.
There are probably 50 more I can't think of right now but the only sure way is to apply engineering and project management discipline to the criteria and comparison of those criteria. Then one must capture a candidate group of commercial and FOSS projects and track them over a multiyear period.
In other words we've been looking at the development experience instead of the results experience. How you build something is less important than what it does. Anyone who's ever seen the movie 'Apollo 13' understands this. More to the point though, development modalities reflect more the cultural aspects that the development team has almost no control over. Even in FOSS communities, they will self organize and operate according to features that have little to do with development.
We really don't know or care that much what the differences between good and mediocre closed source projects are. They are unverifiable in either case. So one cannot focus on the method. It's a black box. Instead we need to focus on the outputs and metrics that we can see.
Re: (Score:3, Insightful)
But they aren't selling anything ... FOSS developers code because it's fun; that's their compensation, not money. It's no excuse to be an ass, but I don't really see why they should necessarily cater to anyone that isn't contributing in a tangible way.
Re:Developers who ignore users (Score:5, Insightful)
While this may be OSS development in the sense that people work on OSS code, it isn't about this topic - the "FLOSS Community" and the coders that form part of this community. Those coders tend to fall in the second camp. They tend to work for reasons other then direct cash. They do it for fun, peer recognition, whatever. For the majority of these people, "non-coding end-users" are the same bunch of clusterfucks they deal with everyday during the dayjob, and tend to not feature very prominently in the motivation chain. The things that drive them are project that are "fun" to code, "pet projects" and all that kind of stuff. They have little motivation to work on projects that are "boring", "seen as difficult" or "of no interest to the developers". This camp of OSS developers "must" do nothing, and more importantly, owe you nothing.
You then bring in some muddled argumentations about the "market" and "running out of business". Unless the OSS coder in question is pretty incompetent, and gets fired from his (quite possibly non-OSS related) dayjob, there is no "business" to be run out of. Most of these projects *are* pet projects, and they only reason you can use them for free is because the coder in question has an urge to tell the world: "Look what I can do!!"
Now its time to bring market drivers / basic economics into the picture. You, as a non-coding end-user, want an application. There are some half-way-there projects out there, but non really fit your bill. You are angry because all the selfish devs only think of their pet projects and having fun. Some entrepeneur, somewhere, will know this, and hire a bunch of devs to create a project you, and hopefully many others, will pay good money for. Only now, once renumeration has entered the picture, can you speak of a market in a meaningful way. Now you are a paying customer, and you can vote with your wallet and feet.
Unless you are a cheapskate, and don't want to pay for anything, but still want every little piece of functionality handed to you "just so". If you ain't paying the cash, either do it yourself, or STFU.
Parent
Where is Eclipse in this picture? (Score:3, Informative)
Then there is Java -- long time criticized by RMS but not trumpeted by SUN as being GPL'd. And how about Open Office/StarOffice with a
Re:what is wrong (Score:5, Insightful)
Parent
Re: (Score:3, Insightful)