Excellent post overall. You are dead right that Microsoft has at their main market the basic tech needs of most enterprises. However, I do have to take issue with the following two points:
(3) A tech oriented make-over of MS based on innovation is a fantasy. An infantile fantasy: the kind that you're supposed to grow out of. They have a great business now, they just need to update it for the needs of 2005 instead of the needs of 1985.
(4) To do this, they need to become their customer's best friend, not the devil you know. People now have more time to be skeptical and demanding than they used to.
Yes and no. Absolutely, MS should not stake its future on some high-profile solves-all-the-worlds-problems technology, because their client base - corporate IT - will be the last to adopt this.
However, Microsoft's two core products are a one-time-only sale, and despite all their efforts to shepherd people into "Value Upgrade New Plus Value" plans, people only want to pay once for the same product. So far they've been able to slip 1 license per new computer in by demanding the sellers include it, but it won't be long before hardware vendors and corporate IT figure out how to come to terms and simply re-use existing Windows & Office licenses and cut MS out. This means MS needs to deliver something substantially different in order for people to buy again. Which takes us back to developing some kind of new technology. Merely getting more customer enthusiasm won't change that.
Some of it is plain bad luck. For example, they chose to view DRM as an integrated unit comprising both personal and corporate channels, when in fact buyers want very different approaches to DRM in those channels. Not to mention that the privacy and data control concerns are hitting public conciousness about 2 years after MS got into DRM. From here it doesn't really look like their visions of data privacy and control match up that well with what the market will need. The business consequences of your sales strategy escaping your corpnet pale in comparison with 5M customer profiles and credit card numbers making the nightly news because some sysadmin made a mistake or went home early. Perhaps if their DRM push had come at a better time, they would have gotten the product mix better and locked up the market early.
Some of it is baggage. Any tech company that lived through the 90's spent a fortune trying to plan for the content sales market (AOL-Time Warner anyone?). Google had the good fortune to approach content from a purely meta-level, like radio or TV, so they skip right by the difficult problem of how to make money actually producing and owning the content. Google is pretty close to immune to who owns the content - their plan is just to index it for searching, and cash in on corporate America's marketing budget as the content moves by.
I wouldn't rule out Microsoft getting into that business - over the years they've done pretty well as a second entrant into the market. Look for some kind of hooks for that in Longhorn.
But to rephrase your point 3, they need to stop swinging for the wall tech-wise, because that's a guaranteed strikeout.
The WinFS idea went way beyond desktop search. as I understood it, the idea was to make every filesystem a self-tagging database that could basically interoperate with other filesystem/databases.
This would head in the direction of a grand unified filesystem, which any corporation with document management issues would dream of (imagine being able to manage document revision control across a whole slew of desktops), or search to see when someone in your workgroup last wrote on subject X. It's turned out to be total vaporware of course, but the idea is definitely bigger than you give it credit for.
I never said the decision to go to war in Iraq had anything to do with protection, nor did I say we should be in Iraq in any way!
Fair enough - I overinferred...
What I did say is that our military is a deterrent force.
With states, I agree. For example, our military force unquestionable restrains Kim Jong Il, China, Taiwan, India, and others from much worse behavior than now (just imagine N. Korea _really_ on a rampage). With asymmetric threats I think that's taken too much as a given - after all, a lot of these folks get into jihadism for a chance to fight and die as radicals (whether they actually want to die when death is whizzing at them at a couple hundred miles an hour is probably a different story...), but overall I'm not totally convinced that they are restrained in their actions by our having a military. I don't have a better answer offhand, other than a strong and sustained investment in HUMINT and serious language training for any and all commanders in our military, but I don't think we can continue looking at it merely in terms of computer systems and boots on the ground.
Funny, I agree and disagree. It's amusing to see people try to get around the fact that the military is, on one hand a massive, protective, lifetime loyalty employer. If it's not a job program, ask your congresscritter why s/he spends so much effort trying to get and keep military bases, and what fraction of the population those bases and industries employ.
On the other hand, I do agree with your bulleted list - the military is often (though not always) an growing and learning experience for people in it, and it often provides excellent job preparation. And yes, if I had very limited job prospects, I'd find the military very, very appealing.
oh yeah, there is also that little added benefit of a trained military force being the only thing between you being able to post self-righteous crap like this and you being forced to obey the whims of some dictator.
As enjoyable as these rhetorical flourishes may be, they're mostly fluff. While it's probably not advisable to have just barebones military capacity, it's not clear that we need what we have (aside from military adventurism), or that a large military force is a particularly effective deterrent or response to "asymmetric" threats. We lost 50,000 lives in Vietnam figuring that out, and another 2000 in Iraq repeating disbelief. Not to mention the exorbitant cost of various "defense technology" projects whose efficacy is never put to a real test. SDI anyone?
And if you think Bush is a dictator, it really shows how ignorant and coddled you are in this country, protected by the troops which you disdain.
Both sides of this purported argument are fluff. Bush is not a dictator by any stretch, but to cast the Bush Administration's adventures in geostrategy in Iraq as "protection" is (and always was) a stretch as best. What real-world attack vectors were/are we being protected from? What evidence has ever been demonstrated that Hussein working with the network that attacked our, or that he was providing them with capacity to replicate an attack of a scale justifying the deployment of hundreds of thousands of troops? And what were the metrics for a successful outcome in Iraq? Hell, three years into this war, what ARE the metrics of a successful outcome?
Speaking for myself and not the grandparent, this is not a question of disdain, it's a question of asking for realistic assessment of threat and likely outcomes before sending a lot of our guys out to fight an ill-defined war against an ill-defined enemy. Regrettably we've broken this particular situation (possibly beyond repair), and now we own, leaving us with precisely 0 good options.
I think you're mistaking "desktop market" for "personal computer-related market". When MS controlled the desktop in the 90's it really controlled almost all of the personal computer market. It did fairly well in the corporate market, though it never achieved the same dominance as in personal computing. But you can easily rattle off multiple areas where Microsoft has not dominated the personal computing market: from phones to search to music, Microsoft hasn't been a big player. Yet their Windows/Office/Windows networking market is as solid as ever, barring a tolerable amount of self-competition.
MS has competitive products in any of those new markets, but they don't come anywhere close to dominating them. And it doesn't seem likely they will. Google currently dominates ad-based search, and by all accounts seems to be using that to power a generation of applications that are basically disconnected from the desktop. Whether or not Google lives or dies, it's hard to see MS resuming control of the PC market in the same way as before.
Perhaps you're painting with a bit broad brush there? As someone born and raised on the East Coast, who also spent about 4 years in the Pacific Northwest, I can say with a good bit of certainty that people are much more blunt here in East Coast country, but most of the ones I've dealt with all their lives are just blunt, not jerks.
I'm sorry if people were assholes to you. Honestly, I am. But I also think it's a bit unfair to tar all of us with that brush as well as the "society sissies" label at the same time. There are benefits to each way of interacting - people are blunt here, but you don't have to worry about them being all silently offended. On the other hand, people in PNW are indeed usually quite polity, and they sure aren't gonna hassle you about your lifestyle, but it can be hard to tell what people really think. To each his own...
Don't give Linus short shrift. In spite of his occasional barbs, he is by all accounts an excellent leader and motivator of people. Part of the reason that megacorps can get involved with Linux is that they know they and their devs can interact sanely with Linus.
And the BSD control-everything mentality, while good for building secure routers, doesn't do much to encourage the hobbyists and hackers that are the backbone of the OSS movement. Linus and Stallman both deserve a lot of credit for understanding that you need to build a team (massive, distributed, uncoordinated, internet-based) to build a platform, and laid the groundwork to do that. That's more than just a tolerable personality, it's an understanding of people.
Finally, how about some jeers for Daniel Lyons, who will do just about anything he can to rain on the Linux parade, including publishing a rant by a left-out OSS leader widely reputed to be something of a crank.
Interesting... most of what you say I think is very astute, but a couple of them are, like Doc Love's columns, standard issue "The Rules for Men" type rituals. Mostly I'd just moderate what I think you and Doc overstate:
The "geek effect" is when smart men do stupid social things like acting too interested in a woman or acting like an idiot around her. Just walk up to her, talk to her, make her laugh, ask for her phone number, and if she gives it to you, wait [strike]at least a week[/strike] a little while to gather your wits before calling her. Don't wait too long, or there'll be unintended head games. Don't make a date on a weekend, make it on a weeknight. If she is otherwise taken, [strike]ask[/strike] wait a while, then discretely inquire if she has a friend that she'd like to set up on a blind date. [strike]Never walk away embarrassed or disappointed[/strike] A little disappointment shows you actually cared and that you don't ask out every pair of h**ties that walks by, but slinking away is extremely awkward -- [strike]that's weak and women don't like it[/strike] it telegraphs low self-esteem, and smart boys and girls quickly learn to steer away from that - not to mention it is unsexy on anybody.
This is spot on:
Women do like smart men very much.
I'd also add this: most geeks are quirky - if the woman you desire wants a really straight-laced, quirkless guy, I'd start looking for lady friend, cause to stay with her is to throw the baby out with the bathwater. You may not want to show her your Star Wars action figures for a long, long time, but if you truly sacrifice your personality for someone else, not only is she not that good a fit, but you've lost a piece of yourself.
Women don't want what is bad for them, they want what they think they can't have.
Insightful observation - I'll have to think about it more.
This is 100% false. Open Source relies upon the nature of markets: contributing to the market with the expectation of equal or greater returns.
Perhaps you're trying to fit this broad picture into a narrow lens? There many reasons why people do things. You can, if you like, choose to view these as an exchange of resources, or perhaps as a conversation between people, a social ordering phenomenon, or a variety of other ways. But I'd stop short of casting developers motivations as totally resource-oriented. I do agree with you that few actions are altruistic in the sense of having no benefit whatsoever to the actor (if nothing else, people do "altruistic" things for the satisfaction and sense of generosity they get in return).
It's not that I don't care about making the world a better place, but normal rules of classical economics (i.e. the Invisible Hand) ensure that outcome. Which brings me to Free Software.
Easy, Adam, the "Invisible Hand" sure doesn't ensure that there's never poverty, pollution, corruption, or restriction of free speech. The pillar of open source movements is a set of active government protections for various activities, including an innovative use of copyright law that allows others to create derivative works as long they release them. You may call such protections an extension of various natural right to expression, but it takes laws and good, transparent enforcement of them to corral the market and the Invisible Hand and make OSS work.
Innovations take a long time. Someone (or multiple people, together or independently) has a good idea and a bunch of people work on a bunch of different angles to make it happen. Ideas build on each other. It took years and years of work, between IBM, and UC Berkeley and elsewhere to make SQL a reality. And more years of work to make a rock-solid product.
So for anyone (McVoy or others) to contend that the "invention" of a type of software always happens all one place is absurd. People move stuff along and add to it, sometimes as proprietary works, sometimes using the academic tradition of open information sharing.
McVoy's point that open source does extensively reverse-engineer and thus commoditize many kinds of software is true. And I'm sure that's frustrating to a guy like him, raised in the golden days of virtually unchallenged product development. I would guess that BitKeeper had some new cool stuff - whether that turns out to be innovation kind of depends on where history goes.
Well, if a bit sardonically, put. It sort of surprises me that you still have to hash over these things with so many people. Basically, it takes a delicate balancing act for a government to do enough to keep a society from imploding and keep its grubby paws off of it enough to keep it from stalling.
In keeping with the 21st century being all about asking the economic questions, it seems pretty straightforward to me to wrap up some of these lessons from the 20th century.
Nobody likes taxes, but societies need to make judicious and thrifty investments in their future and their present.
It's bad for everyone to let people go desitute (not just morally bad either - a strong middle class drives a strong economy), but everyon can't have everything on silver platter, and trying to insulate people from every risk insulates them from effort and success as well.
If there's no economic rules, workers and consumers have no way of knowing when they'll get screwed. If there's two many, nobody can do anything new.
(nothing like a 3 bullet-point list to encapsulate 150 years of Western history)
None of this is rocket science, which is why it frustrates me to have to go through debates about whether government should solve every problem or just go drown itself in the bathtub. There's plenty left to debate without resorting truisms.
I agree with the general sentiment, if not with your choice of analogy.
It's pretty simply why IBM doesn't want to release a Linux distro: it thinks that most or all software is moving rapidly towards commodity, and OSS accelerates that cycle. You can make money selling (OSS product + support), or (OSS + consultancy) + (OSS product + non-OSS upgrades/add-ons + support), but the margins are much, much smaller than they used to be. So IBM doesn't really care if Linux supplants Windows - they just want to see MS' margins reduced. Linux is just a way to slam the door behind them so their competitors cannot follow.
When software is commodity the only places in IT to make money are:
a) hardware
b) services - we're talking massive organizational management services here, not just implementing a project or two.
c) crossover into other fields where IT is useful but not central.
And where exactly do we find IBM these days? Recall that article on IBM's research into management and social science? That's about being an organizational management provider. Blue Gene anyone? How far off is IBM Insitute of Psychohistory?
It's a straightforward and very smart strategy. And frankly it's got so many of the fundamentals right, I'm surprised there hasn't been more effort to duplicate it.
As for Microsoft - they've probably got one more rounds of software-as-product in them, but after that the market will likely go totally commodity. It will be interesting to see if they can snap out of their current schizophrenia and chase down some opportunities elsewhere.
OK, this is the first substantive reason I've heard for spending years re-implementing what Sun and others have made reasonably available. Evidently Sun has not quite reconciled itself to playing in the OSS sandbox when it comes to Java, and they will probably have the jitters for a while yet while they figure out what exactly they're trying to sell.
But aside from licensing holy wars, if the problem is that Java is hard to distribute w/ OSS, shouldn't the solution figure out how to pressure Sun or IBM into distributing Java with OSS - or shucks - just make the install process thoroughly transparent? This is like saying: there's no OS with using the Mozilla license! I'm gonna implement a new OSS, that walks and quacks like Linux, but is completely licensed under MPL rather than GPL! Where's the beef here?
Of course, the bigger the project the less of a monetary incentive might be necessary. Ask me to create a logo for your company and get paid $50, I'll pass. Ask me to do the next logo for Firefox 1.5, and I don't need $50, I'll do it for free. (Note I am not comparing my work to Burka & Desroches, or saying the logo needs a replacement, just using Firefox as an example).
I'd go a step further. $50 is less than free - it trivializes the work and reduces the non-monetary value as well. While you might well be an stellar coder or designer who contributed something incredibly valuable to the project and OSS in general, outsiders with marginal knowledge are gonna think: "$50? Either that project must have been really simple, or that coder's just not that good." Whereas before the stellar coder looked like a real saint for stepping in, now s/he has merely raised questions about his or her skills and ethics ("they paid him/her? isn't that really volunteer work?").
Pro bono law work revolves around this: the hours the lawyers _don't_ bill are worth plenty plenty to the firm, but if the pro bono lawyer billed at rate the client could afford (e.g. $50/hour) the whole thing would be a loss to the firm. If you get down to token, "frame and put on your wall" or "have a nice dinner on us" amounts, then it's still perceived as essentially volunteer work. $50 bucks sounds ot me like you'll get something like: a) a college student of unknown quality and follow-thru b) an enterprising Indian or Chinese coder for whom the value of $50 is different. c) somebody desperate or out of work
Note that Mark Shuttleworth is offering small but legitimate money for specs ($500), and real money for implementation (~$10000, and he's in S. Africa, ain't he?).
It's nice to think that you can have a range of incentives, but the reality is that you have to be very, very careful mixing volunteer work with paid work, or people start wondering what your motivations are.
As for LimeWire going in for this kind of work, it reinforces my impression that they don't have a clear business model. Hiring out for a couple hundred bucks (and no spec!) at a time for some bag o' features says to me that the plan is "let's make something really cool and it will sell itself", which is almost always a recipe for bankruptcy, and doubly so in a sector with an established track record of nobody making money on cool things that have already been invented.
In case you missed the boat on Economics 101, *you* are the polluter here. Companies wouldn't be producing goods that result in pollution unless *you* demanded them in the first place.
Not really. What the consumer wants is for the product to meet certain needs - they don't specify pollution or not pollution (although in terms of branding they are starting to do so). Your point that both are participating in the tragedy of the commons is well taken, though, and to protect the commons, you basically need public intervention.
Which means *any* cost for cleanup or pollution restriction is passed on directly to you, the customer.
Essentially this is correct, and basically as it should be - you want the good, you should pay for all the costs. But as often as not, the imposition of standards pushes the balance sheet towards investment in technologies that simultaneously improve performance, and have less environmental impact (as in the story about flexible concrete) - afer all, people still want the goods. And as people get better at making them, the real cost of the good dips back below the cost of the old technology (as is occuring now with LCD monitors). Not only that, but all that investment in new technology is great business for all us scientists and engineers, and good news for an economy increasingly based on automated production.
Full disclosure: I'm pretty secular, and I'm a scientist, child of a scientist, so my worldview is pretty deeply rooted in science. Quoth the parent:
Hey, I have no problem with people having faith in their religion, or believing things according to faith. But that's all it is - faith.
If you want to teach creationism, do it in religious studies class, not science.
I think what the creationists are broadly disquieted about is the notion that a secular lifestyle is implicitly endorsed by the separation of church and state (and some of them don't even want separation of church and state). What is clear constitutionally (and to me morally) is that the government has an obligation not to favor one religion or its practitioners over another.
Inasmuch as people take science as a worldview (and trust me, I know plenty of dogmatic scientists who try to view their whole life through the prism of science) it is a kind of faith itself (though in an odd twist, its practitioners deny the existence of the faith they pratice), and there's a potential non-discrimination issue there. So I think it's important for schools to really emphasize people's absolute freedom to form their faith and their worldview.
On the other hand, the scientific method (properly applied) is a profoundly effective strategy for basing action on empirical results, and deserves a lot of credit for providing a way for people to broker compromise among different worldviews without having to resort to violence. The scientific method really ought to have no view on the supernatural - its job is to put what people have observed into a good empirical order. The people who use science to categorically deny the existence of the supernatural are misapplying science as much as those who deny evolution are defying science. Put it this way - if angels are indetectable except in people's internal experiences, they are unprovable and hence not the domain of scientific explanation.
That said, I believe that many of those who are trying to put creationism into schools are interested in far more than just balanced teaching. These same folks have done everything they can to monkey with the scientific process and are attempt to block any research they think might contradict their literal, fact-for-fact interpretation of the bible. And given that empirical evidence is the best neutral ground to mediate between different interests, the government had damn well better stick to empirical evidence.
I should mod you up, but instead I'll reply in support.
I think that it's a fairly high probability that the Earth's climate is a chaotic system, and as such, our contribution *might*, in fact, affect the future climate. Either way, I see no sin in reducing the profligate amount of pollution we contribute to the environment, and think that a reduction can only have a positive effect on the future.
Somebody give this dude a MacArthur grant for a brilliantly spoken statement. Let's take a step back from absolute consensus. There's no doubt that we are putting a whole bunch of greenhouse gases into the atmosphere. There's little doubt that greenhouse gases are a controlling element for an important temperature balance cycle: solar radition -> atmospheric transmission -> reflection (i.e. albedo) -> reabsorption. There is some disagreement - I'll stay away from whether there's "consensus" - over whether this cycle is the primary driver of short term and long-term variability. There's may or may not be consensus on whether the earth has any long-term equilibrium states, nor on exactly what those states are - that's the chaos the parent alludes to. To be honest, I'm not we've gotten far enough to even use those kinds of mathematically-oriented descriptions. Fair enough.
But to cast this in terms of construction - if you were operating a crane without knowing entirely what the levers do or what the crane might hit, and some people said there was a big chance of catastrophic results, and other you can successfully keep the lever down, would you pull these levers to their max, or would you try to make smaller motions? Prudence and realism clearly dictates the latter.
right on, right on. Especially #s 4 and 7. It took people years to realize that durable software is stuff that can withstand people's constant impulse to change things. After finally realizing that doing this means keeping things as distinct as possible, we've essentially arrived at the wisdom of 3-tiered (n-tiered) architectures. There are definitely problems with that architecture, but it's hard to believe that it's not important to have the option to keep some of the logic outside the database.
On #7, I totally agree, but let me see if I understand the basic issue well enough to paraphrase. Databases basically work by using a variety of indexes plus hashing to quickly determine whether data meets a set of (exact) criteria. Different indexes work better for different kinds of exact queries, but you can relatively quickly work out a balanced set of indexes that do the job. In other words, some indexes are great for "field > number" queries, while others are better at "field matches string_pattern" queries.
So in order to have effective heuristic or "fuzzy logic" queries, somebody will need to work out indexes and hashes for each fuzzy logic matching operator, or write an algorithm that figures out how to make those indexes and hashes. And that's ummm... rather more difficult. So until then, we have to catch as catch can with with analyzers and aggregators doing underlying exact queries and applying as many optimizations as they can.
So that's my understanding of it - please correct me if I'm wrong.
How on earth does somebody think that a FREE model will be ever replaced by a NON FREE one??
How on earth does somebody imagine that people will pay to get around places when they can bicycle everywhere for free, free, free!
I use and like Rhapsody for the following reasons: 1) It's not too expensive: $10 a month is really pretty cheap for constant access to a sizeable music library. 2) It has a decently broad library - not everything I'd like to hear is there, but certainly it's got a lot more than I would have on my machine if it were up to me to carry it around whereever I go. 3) Though its interface has some definitely flaws, its advertising is pretty unintrusive (haven't seen this new release), so its just about listening to music. 4) When I want to listen to a song, I fire it up, buffer and listen. I don't have to come back and tend my downloads, or throw away clipped, duped, mislabeled or poisoned songs. The worst that ever happens is that the connection drops, and I can't listen for a bit.
All of this rests on having a high-quality internet connection wherever I am. But most of the times I want to listen to music on a computer, the connection is no problem.
So there you have it: time is money, and $10 a month saves me an lot awful lot of time and headache.
You, and the article make a bunch of good points, but you miss one key one that will keep things in play: cost.
The article, although apparently from the POV of the purchaser (business, in this case), it actually speaks from the point of view of the software industry. End purchasers care about continuity, proper performance, and price, in roughly that order. All the article's arguments are valid when the end user interacts directly with the software producer. If they have to screw around with Linux kernel changes themselves, yes they get pissed. But when there's an intermediary (packager, vendor, consultant, etc.) who can provide continuity and performance, there's a nice opportunity to capitalize on the massive bag of Legos which those intermediaries obtain for $100 to several thousand less per copy than Microsoftware. If the intermediary/alternate vendor can figure out a way to split that $100 + between themselves and the consumer, there's an incentive for the consumer to _consider_ change. That's precisely the niche that IBM's in. I wouldn't be surprised to see IBM start opening up more of their core products, if and when: a) transition that revenue to services (services is always the top of the stack) b) address the platform/continuity issues the article brings up c) doing so would represent a kick in the teeth to a competitor
This is not to be a OSS triumphalist, but I think there's a decent enough balance in there that it might be just a wee bit early to call the OS market sewn up by Microsoft.
OK, finally somebody cuts to the metacommentary back to the heart of the issue:
But to claim that it is reverse engineering the way that Tridge is reverse engineering the Bitkeeper client is definitely not correct.
Agreed. Reverse-engineering is permissible legally (and morally, though truthfully I'm not sure exactly how morality ought to fit in here) when it's an effort to make a widget you've already made work with somebody else's widget. Linux was a reimplementation of existing standards (i.e. creation of an OS that worked with existing standards), plus a bunch of drivers that do so. If anything comes close to the kind of reverse engineering Tridge is doing, it would be the recreation of the utilities and shells that made the Linux kernel viable, which were reverse engineered versions of their UNIX counterparts.
I'd say the ethics and legality of this depend on what Tridge is trying to accomplish by reverse engineering. When you're essentially attempting to duplicate someone else's widget, to do so against their wishes is dicey. This of course, is where a functional patent system ought to come in - if BitKeeper is truly a novel, useful idea, then McVoy should be able to ask for, and get a patent for his invention.
I don't really have a good enough sense of what Tridge is doing to know where his work falls, but if it is, as McVoy and Linus represent, a complete recreation of the product using reverse engineering, clean room or no, I'd say Tridge is out of bounds unless he has a pretty compelling argument for why he should be able to recreate the product against McVoy's wishes.
Seriously, you must have a license, to drive a car, fish, hunt, act as a business, and so on. No such requirement for being a parent. Not saying that we should license people to be parents, but does it not strike anyone else as odd that a certain level of competency must be demonstrated before you can get in an automobile and drive, but not to have a child.
While I completely agree with the sentiment that parents ought to get help and support in learning parenting, there's a slippery slope on this side of things. Driving a car or hunting are licenseable because the activities in question are not considered fundamental rights - and rightly so, in my opinion. Having children, on the other hand is considered such a right - and rightly so, in my opinion. After all, I can't imagine many Americans of any political persuasion want to pass laws about when people may and may not have children. So the best you can hope for is a lot of outreach programs that teach people how to parent.
My civil libertarian side says that government intervention is not a good way to help keep kids from playing games that are too mature for their age, but there's a pretty limited windows to learn decent parenting skills. If it takes a parent until 8th grade until to learn the he has to keep an eye on his kids' video games as well as teaching them about violence in reality vs. in fantasy, anger management, I'd wager there's a pretty good chance that kid will have some developmental hurdles ahead of him or her.
I also think things with kids go way amuck when they start to run around in groups that are involved with bad influences (excessive drug use, real-world violence, crime, etc). I'd be really curious to know how video games relate to those kind of social situations and to the decision-making skills kids need to navigate those situations.
Hear hear! Frankly, if this guy wants guaranteed-quality computer repair, perhaps he ought to look for it first, and see what it costs.
I do think there's an interesting conundrum for companies in here. In the IT industry, and numerous other engineering industries "repair tech" or any kind of frontline support is basically an entry-level job. Try calling the phone guys to repair your line - inevitably, up shows some guy doing his best to sport some peach fuzz. Not infrequently the guy doesn't know what he's doing and hence makes a huge mess. Worse yet, when he does start to know what he's doing, and up and away he goes, to be replaced by another entry level drone.
The problem is that this is an important part of the customer's interactions with the company, particularly with long-term contracts. So unless they can figure out how to invert that hierarchy, they're be dealing with pissed off customers forever.
I think the idea of establishing incentives for fuller release of data and methods is a great one. Not only can it speed work, but it can finally start to break down the "build-it-yourself" mentality that seems to pervade science (or physics at least), and get people to think in terms of platform compatibility when they build software to solve a particular problem. The amount of repeated software work is simply staggering.
On the other, there are legitimate reasons to want to withhold your code for a while, from commercialization to future work to simply not being confident that your code is in a releasable state, even if your work is.
The goal of greater openness and sharing of data and algorithms is generally a good one, but let's give it a little time to develop, rather than forcing people's hands and creating new problems.
If the question is "how do I sell my product when so many others are giving similar product away for free?", somebody's obviously doing something right over at HotSweatyStinkyBooty.com. Pr0n vendors are already selling content on the internet - and giving it away as well, and making a HotSweatyStinky pile of money in the process.
Slashdot Mirror
Yes and no. Absolutely, MS should not stake its future on some high-profile solves-all-the-worlds-problems technology, because their client base - corporate IT - will be the last to adopt this.
However, Microsoft's two core products are a one-time-only sale, and despite all their efforts to shepherd people into "Value Upgrade New Plus Value" plans, people only want to pay once for the same product. So far they've been able to slip 1 license per new computer in by demanding the sellers include it, but it won't be long before hardware vendors and corporate IT figure out how to come to terms and simply re-use existing Windows & Office licenses and cut MS out. This means MS needs to deliver something substantially different in order for people to buy again. Which takes us back to developing some kind of new technology. Merely getting more customer enthusiasm won't change that.
Some of it is plain bad luck. For example, they chose to view DRM as an integrated unit comprising both personal and corporate channels, when in fact buyers want very different approaches to DRM in those channels. Not to mention that the privacy and data control concerns are hitting public conciousness about 2 years after MS got into DRM. From here it doesn't really look like their visions of data privacy and control match up that well with what the market will need. The business consequences of your sales strategy escaping your corpnet pale in comparison with 5M customer profiles and credit card numbers making the nightly news because some sysadmin made a mistake or went home early. Perhaps if their DRM push had come at a better time, they would have gotten the product mix better and locked up the market early.
Some of it is baggage. Any tech company that lived through the 90's spent a fortune trying to plan for the content sales market (AOL-Time Warner anyone?). Google had the good fortune to approach content from a purely meta-level, like radio or TV, so they skip right by the difficult problem of how to make money actually producing and owning the content. Google is pretty close to immune to who owns the content - their plan is just to index it for searching, and cash in on corporate America's marketing budget as the content moves by.
I wouldn't rule out Microsoft getting into that business - over the years they've done pretty well as a second entrant into the market. Look for some kind of hooks for that in Longhorn.
But to rephrase your point 3, they need to stop swinging for the wall tech-wise, because that's a guaranteed strikeout.
The WinFS idea went way beyond desktop search. as I understood it, the idea was to make every filesystem a self-tagging database that could basically interoperate with other filesystem/databases.
This would head in the direction of a grand unified filesystem, which any corporation with document management issues would dream of (imagine being able to manage document revision control across a whole slew of desktops), or search to see when someone in your workgroup last wrote on subject X. It's turned out to be total vaporware of course, but the idea is definitely bigger than you give it credit for.
Fair enough - I overinferred...
With states, I agree. For example, our military force unquestionable restrains Kim Jong Il, China, Taiwan, India, and others from much worse behavior than now (just imagine N. Korea _really_ on a rampage). With asymmetric threats I think that's taken too much as a given - after all, a lot of these folks get into jihadism for a chance to fight and die as radicals (whether they actually want to die when death is whizzing at them at a couple hundred miles an hour is probably a different story...), but overall I'm not totally convinced that they are restrained in their actions by our having a military. I don't have a better answer offhand, other than a strong and sustained investment in HUMINT and serious language training for any and all commanders in our military, but I don't think we can continue looking at it merely in terms of computer systems and boots on the ground.
On the other hand, I do agree with your bulleted list - the military is often (though not always) an growing and learning experience for people in it, and it often provides excellent job preparation. And yes, if I had very limited job prospects, I'd find the military very, very appealing.
As enjoyable as these rhetorical flourishes may be, they're mostly fluff. While it's probably not advisable to have just barebones military capacity, it's not clear that we need what we have (aside from military adventurism), or that a large military force is a particularly effective deterrent or response to "asymmetric" threats. We lost 50,000 lives in Vietnam figuring that out, and another 2000 in Iraq repeating disbelief. Not to mention the exorbitant cost of various "defense technology" projects whose efficacy is never put to a real test. SDI anyone?
Both sides of this purported argument are fluff. Bush is not a dictator by any stretch, but to cast the Bush Administration's adventures in geostrategy in Iraq as "protection" is (and always was) a stretch as best. What real-world attack vectors were/are we being protected from? What evidence has ever been demonstrated that Hussein working with the network that attacked our, or that he was providing them with capacity to replicate an attack of a scale justifying the deployment of hundreds of thousands of troops? And what were the metrics for a successful outcome in Iraq? Hell, three years into this war, what ARE the metrics of a successful outcome?
Speaking for myself and not the grandparent, this is not a question of disdain, it's a question of asking for realistic assessment of threat and likely outcomes before sending a lot of our guys out to fight an ill-defined war against an ill-defined enemy. Regrettably we've broken this particular situation (possibly beyond repair), and now we own, leaving us with precisely 0 good options.
I think you're mistaking "desktop market" for "personal computer-related market". When MS controlled the desktop in the 90's it really controlled almost all of the personal computer market. It did fairly well in the corporate market, though it never achieved the same dominance as in personal computing. But you can easily rattle off multiple areas where Microsoft has not dominated the personal computing market: from phones to search to music, Microsoft hasn't been a big player. Yet their Windows/Office/Windows networking market is as solid as ever, barring a tolerable amount of self-competition.
MS has competitive products in any of those new markets, but they don't come anywhere close to dominating them. And it doesn't seem likely they will. Google currently dominates ad-based search, and by all accounts seems to be using that to power a generation of applications that are basically disconnected from the desktop. Whether or not Google lives or dies, it's hard to see MS resuming control of the PC market in the same way as before.
Perhaps you're painting with a bit broad brush there? As someone born and raised on the East Coast, who also spent about 4 years in the Pacific Northwest, I can say with a good bit of certainty that people are much more blunt here in East Coast country, but most of the ones I've dealt with all their lives are just blunt, not jerks.
I'm sorry if people were assholes to you. Honestly, I am. But I also think it's a bit unfair to tar all of us with that brush as well as the "society sissies" label at the same time. There are benefits to each way of interacting - people are blunt here, but you don't have to worry about them being all silently offended. On the other hand, people in PNW are indeed usually quite polity, and they sure aren't gonna hassle you about your lifestyle, but it can be hard to tell what people really think. To each his own...
Don't give Linus short shrift. In spite of his occasional barbs, he is by all accounts an excellent leader and motivator of people. Part of the reason that megacorps can get involved with Linux is that they know they and their devs can interact sanely with Linus.
And the BSD control-everything mentality, while good for building secure routers, doesn't do much to encourage the hobbyists and hackers that are the backbone of the OSS movement. Linus and Stallman both deserve a lot of credit for understanding that you need to build a team (massive, distributed, uncoordinated, internet-based) to build a platform, and laid the groundwork to do that. That's more than just a tolerable personality, it's an understanding of people.
Finally, how about some jeers for Daniel Lyons, who will do just about anything he can to rain on the Linux parade, including publishing a rant by a left-out OSS leader widely reputed to be something of a crank.
Perhaps you're trying to fit this broad picture into a narrow lens? There many reasons why people do things. You can, if you like, choose to view these as an exchange of resources, or perhaps as a conversation between people, a social ordering phenomenon, or a variety of other ways. But I'd stop short of casting developers motivations as totally resource-oriented. I do agree with you that few actions are altruistic in the sense of having no benefit whatsoever to the actor (if nothing else, people do "altruistic" things for the satisfaction and sense of generosity they get in return).
Easy, Adam, the "Invisible Hand" sure doesn't ensure that there's never poverty, pollution, corruption, or restriction of free speech. The pillar of open source movements is a set of active government protections for various activities, including an innovative use of copyright law that allows others to create derivative works as long they release them. You may call such protections an extension of various natural right to expression, but it takes laws and good, transparent enforcement of them to corral the market and the Invisible Hand and make OSS work.
I would mod you up to 6 for that one...
Innovations take a long time. Someone (or multiple people, together or independently) has a good idea and a bunch of people work on a bunch of different angles to make it happen. Ideas build on each other. It took years and years of work, between IBM, and UC Berkeley and elsewhere to make SQL a reality. And more years of work to make a rock-solid product.
So for anyone (McVoy or others) to contend that the "invention" of a type of software always happens all one place is absurd. People move stuff along and add to it, sometimes as proprietary works, sometimes using the academic tradition of open information sharing.
McVoy's point that open source does extensively reverse-engineer and thus commoditize many kinds of software is true. And I'm sure that's frustrating to a guy like him, raised in the golden days of virtually unchallenged product development. I would guess that BitKeeper had some new cool stuff - whether that turns out to be innovation kind of depends on where history goes.
In keeping with the 21st century being all about asking the economic questions, it seems pretty straightforward to me to wrap up some of these lessons from the 20th century.
- Nobody likes taxes, but societies need to make judicious and thrifty investments in their future and their present.
- It's bad for everyone to let people go desitute (not just morally bad either - a strong middle class drives a strong economy), but everyon can't have everything on silver platter, and trying to insulate people from every risk insulates them from effort and success as well.
- If there's no economic rules, workers and consumers have no way of knowing when they'll get screwed. If there's two many, nobody can do anything new.
(nothing like a 3 bullet-point list to encapsulate 150 years of Western history)None of this is rocket science, which is why it frustrates me to have to go through debates about whether government should solve every problem or just go drown itself in the bathtub. There's plenty left to debate without resorting truisms.
I agree with the general sentiment, if not with your choice of analogy.
It's pretty simply why IBM doesn't want to release a Linux distro: it thinks that most or all software is moving rapidly towards commodity, and OSS accelerates that cycle. You can make money selling (OSS product + support), or (OSS + consultancy) + (OSS product + non-OSS upgrades/add-ons + support), but the margins are much, much smaller than they used to be. So IBM doesn't really care if Linux supplants Windows - they just want to see MS' margins reduced. Linux is just a way to slam the door behind them so their competitors cannot follow.
When software is commodity the only places in IT to make money are:
a) hardware
b) services - we're talking massive organizational management services here, not just implementing a project or two.
c) crossover into other fields where IT is useful but not central.
And where exactly do we find IBM these days?
Recall that article on IBM's research into management and social science? That's about being an organizational management provider. Blue Gene anyone? How far off is IBM Insitute of Psychohistory?
It's a straightforward and very smart strategy. And frankly it's got so many of the fundamentals right, I'm surprised there hasn't been more effort to duplicate it.
As for Microsoft - they've probably got one more rounds of software-as-product in them, but after that the market will likely go totally commodity. It will be interesting to see if they can snap out of their current schizophrenia and chase down some opportunities elsewhere.
OK, this is the first substantive reason I've heard for spending years re-implementing what Sun and others have made reasonably available. Evidently Sun has not quite reconciled itself to playing in the OSS sandbox when it comes to Java, and they will probably have the jitters for a while yet while they figure out what exactly they're trying to sell.
But aside from licensing holy wars, if the problem is that Java is hard to distribute w/ OSS, shouldn't the solution figure out how to pressure Sun or IBM into distributing Java with OSS - or shucks - just make the install process thoroughly transparent? This is like saying: there's no OS with using the Mozilla license! I'm gonna implement a new OSS, that walks and quacks like Linux, but is completely licensed under MPL rather than GPL! Where's the beef here?
Pro bono law work revolves around this: the hours the lawyers _don't_ bill are worth plenty plenty to the firm, but if the pro bono lawyer billed at rate the client could afford (e.g. $50/hour) the whole thing would be a loss to the firm. If you get down to token, "frame and put on your wall" or "have a nice dinner on us" amounts, then it's still perceived as essentially volunteer work. $50 bucks sounds ot me like you'll get something like:
a) a college student of unknown quality and follow-thru
b) an enterprising Indian or Chinese coder for whom the value of $50 is different.
c) somebody desperate or out of work
Note that Mark Shuttleworth is offering small but legitimate money for specs ($500), and real money for implementation (~$10000, and he's in S. Africa, ain't he?).
It's nice to think that you can have a range of incentives, but the reality is that you have to be very, very careful mixing volunteer work with paid work, or people start wondering what your motivations are.
As for LimeWire going in for this kind of work, it reinforces my impression that they don't have a clear business model. Hiring out for a couple hundred bucks (and no spec!) at a time for some bag o' features says to me that the plan is "let's make something really cool and it will sell itself", which is almost always a recipe for bankruptcy, and doubly so in a sector with an established track record of nobody making money on cool things that have already been invented.
Essentially this is correct, and basically as it should be - you want the good, you should pay for all the costs. But as often as not, the imposition of standards pushes the balance sheet towards investment in technologies that simultaneously improve performance, and have less environmental impact (as in the story about flexible concrete) - afer all, people still want the goods. And as people get better at making them, the real cost of the good dips back below the cost of the old technology (as is occuring now with LCD monitors). Not only that, but all that investment in new technology is great business for all us scientists and engineers, and good news for an economy increasingly based on automated production.
Quoth the parent:
I think what the creationists are broadly disquieted about is the notion that a secular lifestyle is implicitly endorsed by the separation of church and state (and some of them don't even want separation of church and state). What is clear constitutionally (and to me morally) is that the government has an obligation not to favor one religion or its practitioners over another.
Inasmuch as people take science as a worldview (and trust me, I know plenty of dogmatic scientists who try to view their whole life through the prism of science) it is a kind of faith itself (though in an odd twist, its practitioners deny the existence of the faith they pratice), and there's a potential non-discrimination issue there. So I think it's important for schools to really emphasize people's absolute freedom to form their faith and their worldview.
On the other hand, the scientific method (properly applied) is a profoundly effective strategy for basing action on empirical results, and deserves a lot of credit for providing a way for people to broker compromise among different worldviews without having to resort to violence. The scientific method really ought to have no view on the supernatural - its job is to put what people have observed into a good empirical order. The people who use science to categorically deny the existence of the supernatural are misapplying science as much as those who deny evolution are defying science. Put it this way - if angels are indetectable except in people's internal experiences, they are unprovable and hence not the domain of scientific explanation.
That said, I believe that many of those who are trying to put creationism into schools are interested in far more than just balanced teaching. These same folks have done everything they can to monkey with the scientific process and are attempt to block any research they think might contradict their literal, fact-for-fact interpretation of the bible. And given that empirical evidence is the best neutral ground to mediate between different interests, the government had damn well better stick to empirical evidence.
Somebody give this dude a MacArthur grant for a brilliantly spoken statement. Let's take a step back from absolute consensus. There's no doubt that we are putting a whole bunch of greenhouse gases into the atmosphere. There's little doubt that greenhouse gases are a controlling element for an important temperature balance cycle: solar radition -> atmospheric transmission -> reflection (i.e. albedo) -> reabsorption. There is some disagreement - I'll stay away from whether there's "consensus" - over whether this cycle is the primary driver of short term and long-term variability. There's may or may not be consensus on whether the earth has any long-term equilibrium states, nor on exactly what those states are - that's the chaos the parent alludes to. To be honest, I'm not we've gotten far enough to even use those kinds of mathematically-oriented descriptions. Fair enough.
But to cast this in terms of construction - if you were operating a crane without knowing entirely what the levers do or what the crane might hit, and some people said there was a big chance of catastrophic results, and other you can successfully keep the lever down, would you pull these levers to their max, or would you try to make smaller motions? Prudence and realism clearly dictates the latter.
right on, right on. Especially #s 4 and 7. It took people years to realize that durable software is stuff that can withstand people's constant impulse to change things. After finally realizing that doing this means keeping things as distinct as possible, we've essentially arrived at the wisdom of 3-tiered (n-tiered) architectures. There are definitely problems with that architecture, but it's hard to believe that it's not important to have the option to keep some of the logic outside the database.
On #7, I totally agree, but let me see if I understand the basic issue well enough to paraphrase. Databases basically work by using a variety of indexes plus hashing to quickly determine whether data meets a set of (exact) criteria. Different indexes work better for different kinds of exact queries, but you can relatively quickly work out a balanced set of indexes that do the job. In other words, some indexes are great for "field > number" queries, while others are better at "field matches string_pattern" queries.
So in order to have effective heuristic or "fuzzy logic" queries, somebody will need to work out indexes and hashes for each fuzzy logic matching operator, or write an algorithm that figures out how to make those indexes and hashes. And that's ummm... rather more difficult. So until then, we have to catch as catch can with with analyzers and aggregators doing underlying exact queries and applying as many optimizations as they can.
So that's my understanding of it - please correct me if I'm wrong.
How on earth does somebody imagine that people will pay to get around places when they can bicycle everywhere for free, free, free!
I use and like Rhapsody for the following reasons:
1) It's not too expensive: $10 a month is really pretty cheap for constant access to a sizeable music library.
2) It has a decently broad library - not everything I'd like to hear is there, but certainly it's got a lot more than I would have on my machine if it were up to me to carry it around whereever I go.
3) Though its interface has some definitely flaws, its advertising is pretty unintrusive (haven't seen this new release), so its just about listening to music.
4) When I want to listen to a song, I fire it up, buffer and listen. I don't have to come back and tend my downloads, or throw away clipped, duped, mislabeled or poisoned songs. The worst that ever happens is that the connection drops, and I can't listen for a bit.
All of this rests on having a high-quality internet connection wherever I am. But most of the times I want to listen to music on a computer, the connection is no problem.
So there you have it: time is money, and $10 a month saves me an lot awful lot of time and headache.
You, and the article make a bunch of good points, but you miss one key one that will keep things in play: cost.
The article, although apparently from the POV of the purchaser (business, in this case), it actually speaks from the point of view of the software industry. End purchasers care about continuity, proper performance, and price, in roughly that order. All the article's arguments are valid when the end user interacts directly with the software producer. If they have to screw around with Linux kernel changes themselves, yes they get pissed. But when there's an intermediary (packager, vendor, consultant, etc.) who can provide continuity and performance, there's a nice opportunity to capitalize on the massive bag of Legos which those intermediaries obtain for $100 to several thousand less per copy than Microsoftware. If the intermediary/alternate vendor can figure out a way to split that $100 + between themselves and the consumer, there's an incentive for the consumer to _consider_ change. That's precisely the niche that IBM's in. I wouldn't be surprised to see IBM start opening up more of their core products, if and when:
a) transition that revenue to services (services is always the top of the stack)
b) address the platform/continuity issues the article brings up
c) doing so would represent a kick in the teeth to a competitor
This is not to be a OSS triumphalist, but I think there's a decent enough balance in there that it might be just a wee bit early to call the OS market sewn up by Microsoft.
Agreed. Reverse-engineering is permissible legally (and morally, though truthfully I'm not sure exactly how morality ought to fit in here) when it's an effort to make a widget you've already made work with somebody else's widget. Linux was a reimplementation of existing standards (i.e. creation of an OS that worked with existing standards), plus a bunch of drivers that do so. If anything comes close to the kind of reverse engineering Tridge is doing, it would be the recreation of the utilities and shells that made the Linux kernel viable, which were reverse engineered versions of their UNIX counterparts.
I'd say the ethics and legality of this depend on what Tridge is trying to accomplish by reverse engineering. When you're essentially attempting to duplicate someone else's widget, to do so against their wishes is dicey. This of course, is where a functional patent system ought to come in - if BitKeeper is truly a novel, useful idea, then McVoy should be able to ask for, and get a patent for his invention.
I don't really have a good enough sense of what Tridge is doing to know where his work falls, but if it is, as McVoy and Linus represent, a complete recreation of the product using reverse engineering, clean room or no, I'd say Tridge is out of bounds unless he has a pretty compelling argument for why he should be able to recreate the product against McVoy's wishes.
While I completely agree with the sentiment that parents ought to get help and support in learning parenting, there's a slippery slope on this side of things. Driving a car or hunting are licenseable because the activities in question are not considered fundamental rights - and rightly so, in my opinion. Having children, on the other hand is considered such a right - and rightly so, in my opinion. After all, I can't imagine many Americans of any political persuasion want to pass laws about when people may and may not have children. So the best you can hope for is a lot of outreach programs that teach people how to parent.
My civil libertarian side says that government intervention is not a good way to help keep kids from playing games that are too mature for their age, but there's a pretty limited windows to learn decent parenting skills. If it takes a parent until 8th grade until to learn the he has to keep an eye on his kids' video games as well as teaching them about violence in reality vs. in fantasy, anger management, I'd wager there's a pretty good chance that kid will have some developmental hurdles ahead of him or her.
I also think things with kids go way amuck when they start to run around in groups that are involved with bad influences (excessive drug use, real-world violence, crime, etc). I'd be really curious to know how video games relate to those kind of social situations and to the decision-making skills kids need to navigate those situations.
Hear hear! Frankly, if this guy wants guaranteed-quality computer repair, perhaps he ought to look for it first, and see what it costs.
I do think there's an interesting conundrum for companies in here. In the IT industry, and numerous other engineering industries "repair tech" or any kind of frontline support is basically an entry-level job. Try calling the phone guys to repair your line - inevitably, up shows some guy doing his best to sport some peach fuzz. Not infrequently the guy doesn't know what he's doing and hence makes a huge mess. Worse yet, when he does start to know what he's doing, and up and away he goes, to be replaced by another entry level drone.
The problem is that this is an important part of the customer's interactions with the company, particularly with long-term contracts. So unless they can figure out how to invert that hierarchy, they're be dealing with pissed off customers forever.
Hear, hear!
I think the idea of establishing incentives for fuller release of data and methods is a great one. Not only can it speed work, but it can finally start to break down the "build-it-yourself" mentality that seems to pervade science (or physics at least), and get people to think in terms of platform compatibility when they build software to solve a particular problem. The amount of repeated software work is simply staggering.
On the other, there are legitimate reasons to want to withhold your code for a while, from commercialization to future work to simply not being confident that your code is in a releasable state, even if your work is.
The goal of greater openness and sharing of data and algorithms is generally a good one, but let's give it a little time to develop, rather than forcing people's hands and creating new problems.
Pr0n.
If the question is "how do I sell my product when so many others are giving similar product away for free?", somebody's obviously doing something right over at HotSweatyStinkyBooty.com. Pr0n vendors are already selling content on the internet - and giving it away as well, and making a HotSweatyStinky pile of money in the process.
So how do they do it?