I agree that "an x86 Apple would probably be a proprietary Apple with an x86 in" but keep in mind that the reason that nobody ran competitors to Windows on x86 (Be, Linux, etc.) was that MS' licensing terms required licensees to pay for Windows even if they didn't ship it, and prohibited companies from shipping any competing OS installed. The result was that since companies were already paying for Windows, they installed it, and since they installed it, they couldn't install any other OS. This meant that even manufacturers that sold PC's specifically to run BeOS shipped them with Windows installed, and told people to take the BeOS CD's and do the installation themselves.
Needless to say, this wiped out BeOS in the OEM market. This tactic was of course eventually determined to be illegal, but by then Be was dead so it didn't matter.
"The transition to powerpc was not a change of architecture, it was an upgrade in architecture. x86 is an entirely different architecture with an entirely different instruction set."
Wow, is this wrong. The 680x0 and the PowerPC have pretty much nothing in common, other than that Motorolla is one of the manufacturers of the PowerPC chips. But the only way that PPC's run 680x0 code is that they run a software emulator.
When Apple moved from 680x0 to PPC they could have easily moved to any other chip architecture. In fact, they (kinda) had MacOS running on SPARC, x86, etc., two different ways. First, the MacOS emulator for UNIX, and second NeXTSTEP -- between the two, that covered running "yellow box" and "blue box". So the choice of PPC wasn't due to any similarity to the 680x0, but purely on other technical and strategic issues (i.e. PPC being faster than x86 at the time and with a cleaner roadmap, the possibility of launching PPC as a competitive platform to x86, etc.).
For perspective, think about that processor companies use when they have a choice. All of the next generation videogame consoles run PPC. Many embedded controllers (e.g. in cars, Tivo's) are PPC's. People only use x86's when they need to run legacy x86 software; PPC's cost less and run faster.
Finally, even if Apple switched to the x86 from PPC, they wouldn't ship MacOS for generic PC's, because there we far too many quality/user experience issues. They toyed it a while back (I used to run Rhapsody for x86) and the idea died because it was impossible for them to generate enough device drivers to have even the slightest change of working on a random PC -- Rhapsody for x86 only ran on specific motherboards, with specific SCSI controllers, specific ethernet interfaces, specific video cards, etc. This was fine for someone building a machine to run that OS, but made it impossible to support more casual switchers.
"Free speech does not include the "right" to deny others their rights of assembly and free speech by shutting their convention down. The protesters did not respect the rights of these others."
Speaking as someone who lives in NYC, and thus saw this all first-hand, this is a misrepresentation of the protestors, and of those that arrested the protestors. The protestors weren't trying to shut down the convention -- they were trying to make their opinions known to attendees of the convention. This is _precisely_ the kind of speech that is constitutionally protected (i.e. political speech).
The police and organizer's goal was to prevent the protestors from getting their message out, by (1) restricting their right to free speech in ways that are (IMO) unconstitutional, and (2) arresting as many people as fast as possible, simply to keep them tied up with the police, where the attendees couldn't see them. The result was that 90%+ of the arrests were later thrown out, often because independent videotapes showed that the police straight out lied...
"it would not have happened if the extremist thugs had not organized an effort to "shut down" the convention and proceeded to harass the convention-goers."
Yeah, free speech is sometimes inconvenient, but despite the efforts of some extremists, we still have constitution that protects our freedoms.
"Surely the can't expect that their raids of arrests will provide them with more sales."
If I understand your claim, I think you're wrong.
Yes, I think that they do in fact think that when people buy bootleg copies of DVD's that they don't buy legit copies of the DVD's, so shutting down illegal manufacturers (i.e. factories that manufacture DVD's that they don't pay royalties on, and street vendors who burn DVD-R's and don't pay royalties) they reduce the supply of bootlegs and thus increase sales of legit DVD's.
"Isn't this (if it actually works!) a type of denial of service attack?"
Nope. A denial of service attach means that the attacher is initiating connections to a server in order to consume the server's resources, denying the service to anyone else.
I believe that all this company is doing (if their technology works) is to join a public network and wait for requests, then sending back bad data. This is a well understood technique.
What this isn't: - A "denial of service attach" since they're not initiating any connections to anything. - A broad attack on the p2p networks, since the bad data is only served in response to specific files.
The only interesting bit is that if they've figured out how to create bad data that passes a hash check then the protocol's error correction won't work. This doesn't change anything with FastTrach (which effectively has no hash checking) but could be used to inject bad data into BitTorrent, eDonkey, etc.
I tried going to the site, and it's slashdotted. There's a nice irony in a company named Ioband not being able to deliever the I/O. Of course, they're being slashdotted by people who aren't using Ioband's technology...
"every country that has recognized gay marriage (among them Belgium and the Netherlands, if i remember correctly) has experienced a drop in the number of marriages"
I think you might have the relationship wrong -- it's well known that (as observed in dozens of countries for decades) as a country matures (and people are better educated, have more income, have longer life spans, lower infant mortality, etc.) their birthrate drops. Gay marriage is a very recent phenomenon, so it can hardly be the cause of a trend well established decades ago. I suspect that the real relationship is that as people get better educated, have better incomes, etc., they also tend to be more accepting of gays. So, for example, legalizing gay marriage last year in Canada didn't cause population growth to slow, because the rate of population growth in Canada peaked in 1951-1956 (see http://www.populationinstitute.ca/tables_and_chart s/table_steep_can_pop_curve.htm) and has been dropping since. You might as well argue that gay marriage causes low rates of divorce, since Massachusetts legalized gay marriage last year, and has the lowest divorce rate in the US. Of course, the real answer is that people in Massachusetts are (compared to the national average) better educated, and thus have better jobs with better incomes, and get married later in life, which leads to them being both being less likely to divorce, and also more supportive of progressive social issues. Contrast this with southern states with large "born again christian" populations, who strongly oppose gay marriage, and who get divorced far more often than the national average -- you wouldn't argue that opposing gay marriage causes high divorce rates, because the coorelation is caused by a third factor, "born again christian" social attitudes that lead to both opposition to gay marriage and a high rate of failed marriages.
"We have a separation of State and religion and that should be followed. The church has no right to impose law. Although like any other organization they try to influence the laws."
Actually, the deal is that churches get to not pay taxes to the government, in return for which they can't directly engage in politics. This is why when churches get too close to endorsing a candidate or telling people how to vote, they risk losing their tax free status.
"How do you know she wanted to die? Oh, you took the husband's word for it even though a court said there was a dispute."
Actually, as documented in the court cases, there were three other witnesses that documented her saying on six occasions things that indicated that she didn't want to have her life artificially sustained with no chance of recovery. Like most young people (she was young 15 years ago when this happened) she didn't have a written Living Will, so the court had to make a judgement based on the evidence.
The parents had no evidence other than their opinions of what she wanted. They later said that even if she'd had a signed Living Will they still would have fought it, which pretty much proves the point to me that they knew what she wanted, but didn't like it. I can understand why her parents would be in denial of the reality of her death (as a parent, I can think of nothing worse than going through 15 years of torture) but I can't understand why anyone else would support their delusions -- it seems unethical to me to encourage "hope" in desparate people given that she was quite clearly gone.
Another aspect of this that confuses me is that from what I can tell the Catholic Church disapproves of artificial means of sustaining life (because if it's God's will that you die, you shouldn't try to avoid it), but her parents claim that as a Catholic she would want to live no matter what. Does this make sense to anyone else?
"At least Republicans showed up for the vote. Your beloved Democrats didn't even appear."
Actually, 50 Democrats voted for it, too. Pretty embarassing that they'd vote for something so pointless. But as a friend put it, their calculation was that voting for it would thrill the 10% of the population that really wants it, would be ignored by the 90% that would oppose it, and didn't cost anything.
Of course, according to polls, a huge majaority of the country, and a majority of evangelicals and republicans, opposed the congressional meddling, and Bush's ratings are lower than ever. So worse than pathetic pandering to a republican fringe group, it was completely unsuccessful pathetic pandering that managed to not only not satisfy the fringe, but alienated the mainstream. There's hope for rationality to kick in yet.
"Because nobody else has developed such an API for Windows. It's not impossible for one to replace IE's API if they really tried."
If MS provided IE's underlying functionality as a DLL with a clean API, this would be true. Think of MacOS X's Safari engine, KDE's KHTML, etc. But in order to "embed" IE in the operating system, they split IE's functionality and mixed it across many DLL's, DLL's that can't be removed because they have other functions that the OS relies on. This was smart of MS, because it made it "impossible" to remove IE, and nearly impossible for any other browser to provide IE's API.
"I mean seriously, if you hate IE that much, why are you even still using Windows?"
Because I can tell the difference between a web browser and an operating system. When I use Windows, I use Firefox and Thunderbird, so I don't get bothered (as much) by pop-ups, email scripting viruses, etc.
"I work retail, and our store received 60 PSP's and sold 10 the entire day."
I am a gadget fiend, and I looked at PSP's at two stores today and didn't buy one.
EB had the PSP in stock for $250, but had no demo units, though there were three employee units running movies against the wall behind the sales counter. Apparently they'd had a demo unit for people to test, and the unit got badly scratched so they stopped letting people see the PSP. As cool as the PSP looks from a distance, I'm not buying a $250 piece of electronics that I can't even hold first.
Then I went to KB Toys. All of their units were boxed, so I couldn't even see one one. Even more insulting, they were so convinced that people would go crazy over the PSP that they refused to sell it except as a "bundle" with two games for $350. To add insult to injury, you had to pick one game from list 'A' and one from list 'B', so you were forced to buy a crappy game!
If someone had let me hold a PSP for a minute, I'm sure that the unit would have sold itself. But with retailers this stupid, no wonder it's not selling.
It kinda reminds me of a visit to an electronic retailer long ago when the 3DO was hot. The store had signage and a demo kiosk set up, but off. When I asked the salesman in the nearly empty videogame area, he said that he unplugged it because it attracted such large crowds that he couldn't do his job (which apparently didn't involve selling things to people). His stupidity lives on.
So I can't say whether Sony did a good job launching the PSP (aside from NYC being blanketed in nice PSP posters), because the retailers were so braindead that the PSP never had a chance.
"The non-traceability of the Open Source process leaves any given product open to contamination from copyrighted/patented IP."
This claim gets the real world situation backwards.
All major open source projects retain a complete and precise development history through use of a a source code repository (e.g. CVS). This source code repository is open to public inspection, so anyone who wishes can determine the exact time and submitter of every line in that project. This has the effect of discouraging cheating, because the cheating is easy to detect, and the perpetrator is easily identified.
Proprietary software, on the other hand, may not have such a record o contributions, and even if one exists, it's certainly not open to public inspection (short of a lawsuit). So if you question the origin of some aspect of a proprietary system, you have to ask the company (i.e. sue them) for the information that you want. This is a bit of a catch-22, since you can't provide evidence of cheating until after you sue them in order to reveal the evidence.
It's certainly true that someone could illegally submit code that they don't own into an open source project, but the same is true for a proprietary project. And if someone thinks that their IP has been incorporated into an open source project, they can easily inspect the project's source code repository, and determine where the code came from and when, which should clarify the situation (and if someone submitted code illegally, smack them and remove the code from the project).
The only case where there's a problem is with proprietary code bases, where it's very difficult to determine whether IP has been illegally used, and if so it's extremely difficult to determine the source of the code.
Note that despite the theoretical risk of commercial IP getting into Open Sourrce projects, in practice I can't think of any cases where that's been shown to have occurred (even SCO gave up making all such charges against IBM), perhaps because the open source projects are open to inspection so such cheating is discouraged, while proprietary products are revealed fairly regularly to include open source software (or illegally used proprietary software), perhaps because the perpetrators thought that nobody outside their company would see the source code, so the cheating was "safe".
"no major distribution, and definitely no kernel, should leave the foundry without knowing who touched it."
You should be happy, because that's already the case for open source software. If the same were true of proprietary software, then we could put the whole issue to bed.
"A similar thing which happened to the story submitter happened to me once. My ISP got a call from the RIAA that someone was sharing loads of music online. So my ISP just looked for the biggest bandwidth user and shut them off. Well, that someone they shut off was me. But I wasn't the one sharing music. I was seeding Knoppix. A phone call to my ISP quickly resolved the situation."
Wow, what a lame ISP! The RIAA complaint would certainly have included an IP address, and they nail you instead of looking up who had that IP address? Amazing!
"Last I checked, apple pays 60c per song and resells them for 99c. That is approximately a 40% margin.... You can hardly say iTMS is next to nothing in margin.... You were pretty harsh for someone who is so totally wrong. "
To back up the analysis, let me provide some numbers.
If a track sells for 99 cents, the record company gets perhaps 70 cents (which then pays royalties to artist, composer, label, etc.). Of the remaining 29 cents, the credit card company charges perhaps 15 cents (depending on how well Apple negotiated them down from the 25 cents that's their starting position). So now you've got 14 cents left, which has to pay for implementing a large scale e-Commerce system that can sell music, plus the hardware and bandwidth costs, plus of course the cost to have the engineers build and run the system, the customer support representatives answer phone calls, the cost of marketing, etc.
Given how iTunes' sales volume has ramped up, I'd hope that iTMS is profitable, but it's a rough business, and for a company that can make high margins on hardware, it wouldn't make sense to spend a lot of money to enter a low-margin business. Buf if they can brek even selling music, and by doing so enable the iPod business to grow (and somehat lock anyone who bought music into buying iPods forever), that's a very smart move.
Yes, there are a few stores selling music slightly cheaper than iTMS, but they do so in somewhat misleading ways. For example, Walmart sells music as a loss leader to generate traffic into the rest of their store. That's also how they sell CD's cheaper than everyone else -- they are willing to sell one heavily promoted CD at a loss, but they place the music section as far from the front door as possible, and sell everyone who goes to the store to buy the music $100 worth of other stuff. And the company advertising 79 cent tracks only sells _some_ tracks for 79 cents -- their tracks range up to $1.49 (last time I looked) so their average prices was about the same as Apple's, and the usage rules were all over the place (e.g. you can burn some music that you buy to CD, other music you can't, etc.). Then there was a company selling music at "reverse auction" prices, where music started cheap and got more expensive the more popular it was, which is cool, but the reality is that they paid the same for their music as everyone else, and decided that they could take the loss in order to get some PR for being a cool place to buy music. Napster is at least interesting, since they sell someting completely different -- essentially they're not sellig music, but are selling subscription access to a library on demand for as long as you keep subscribing, but once you stop subscribing all of the music goes away. Since they're not really selling music that you own, their costs are lower (or at least different). Their problem, though, is that what people want to do is buy music and own it, not rent it. In particular, while in theory the "infinite jukebox in the sky" sounds great, and accountants certainly like the idea of a permanent revenue stream, but in reality teenagers hate the idea of committing to a permanent $15/month subscription fee forever.
"Maybe all the great Windows developers all have jobs"
This could explain some of it, but I also went to universities and talked with students, and at least in the NY area CS students aren't getting trained in writing Windows software.
So, to be more precise, I'll say that we found: - a fair number of "ok" windows developers. But we're shipping an app that has to be "great" consumer software, so it has to look great, be very efficient, take advantage of all of the latest cool Windows API's to do things nobody else is doing, etc., which filtered out all of the corporate/VB developers, etc. - A few.Net programmers that were "great", but we don't want to require users to install.Net runtime (which doesn't support many older machines, and in any case is a very large download/install), so we're coding in C/C++. This filtered out developers that only want to use the latest cool stuff from MS. Similarly, we're not using Java on the desktop for the same reason (and which would have been _way_ easier to hire for and build than C). - We went to schools and talked with their best CS students. They're being taught Java and Python, and play with Ruby. I think one mentioned.Net. So in terms of what's being taught, C/C++ and Windows aren't on the radar, so that filtered out hiring anyone straight out of school.
So we did find a great Windows developer, but it was like pulling teeth. Perhaps if we were in Seattle instead of NYC it would have been different, but in NYC banks are all doing J2EE/web app's, web shops are using DHTML and Flash, etc., so nobody's doing Windows development here.
"what percentage of the market is Linux going to have to take before hackers switch over? I think it'll be well over 50%."
It's easy to show that market share alone doesn't cause security problems. Historically, there's a market where non-Windows software has clearly dominate Windows software -- web servers. The numbers vary over time, but Apache usually is about 2x more popular than IIS as a web server. And if you look at the data on web site exploits, despite the fact that most sites (and most high-profile sites) run Apache/UNIX, 2/3rds of the sites compromised run IIS/NT, and only 1/3rd ran Apache/UNIX (including Linux, BSD, etc.). So for that data (a few million web sites, over a decade), NT/IIS appears to be about 4x as vulnerable to compromise as Apache/UNIX relative to market share.
My conclusion from this data is that security problems are not caused by popularity (in which case, Apache/UNIX would have been less secure than NT/IIS), but by meaningful differences in the design and implementation of the different operating systems.
I started a company a few months ago that's building consumer software that runs on MacOS X and Windows (and Linux, etc., eventually). Our strategy is to build the core in tight C code, and then build platform-specific applications in the appropriate language, so the result is a great ObjC Mac app, a great C++ Windows app, etc. While I like Java, Ruby, etc., our goal is to make the app small and efficient, so asking people to install 30 MB runtimes is out. Interestingly, it was easy to recruit first-class Mac and Java (server) developers, and nearly impossible to recruit a really great Windows developer. It turns out that the best CS students are _all_ working in modern cross-platform environments (e.g. Java, Python, Ruby), most use Mac's, almost none are using C++, and nobody even _considers_ writing Windows applications any more. While this is kinda neat in one respect, it's a bit surreal that the vast majority of great developers won't write software that runs natively for the platform on 95% of desktops. Weird.
I thought about using VMware, but I wanted to make sure that I saw the actual behavior on a normal PC, and I suspected that low level networking running within VMware might not match.
On a more serious note, I think that this is a fantastic piece of analysis. I did a simple version of this last year (nothing formal enough to publish, but interesting) and it took days, because KMD, etc., so thoroughly destabilize a PC on installation that you have to spend hours cleaning/reinstalling/etc., each time.
">> One suggestion is that labels want to introduce variable pricing - so they can charge more for top selling tracks.
> You know what? I'm all for it... if it means they are willing to discount the less-popular stuff in exchange.
So, you think that the record companies, who are already in a frothy panic because they think they're losing all their profits to those "p2p pirates" are going to suddenly become magnanimous and cut you a break by lowering their profit margin?"
Record companies don't discount back catalog sales because they're generous -- they do so because they have a huge inventory of older music that people won't buy for as much as the "hot new hits", so they can either sell it discounted or not sell it. So the best way to maximize their income is to dicount catalog sales, either directly (e.g. lower track and album prices) or indirectly (e.g. selling boxed sets, "best of" albums).
"Lets see, it is using an over hyped keyword based on an Apple product when just about any mp3 playing device works. And all it is is using RSS feeds to find and pull down new music/shows."
I think you're missing the point of why PodCasting is so cool. It's not interesting because it's hard technology, it's interesting because it's trivial technology that is trivial to implement, thus has been widely implemented, and has (and will have) a huge impact on people's lives.
Remember when the Enterprise crew list was leaked, and everyone thought that it had to be a fake because it was so derivative of the previous Trek shows? I wish that analysis had been correct -- the cast feels like the cast of previous Trek's rotated slightly (this time the southerner is the Weapons Officer instead of the Doctor, and we'll have alien instead of a hologram). It's a shame that the writing is so bad -- the acting is first rate, the effects are great, etc., but until the last few episodes I was _hoping_ that they'd cancel Enterprise (and fire B&B) and try again in a year or two.
Me, I'm annoyed that they stripped out the dock and cable because it lets them drop the price of the 60 GB iPod Photo by $150 but call it a (trivially) new product by stripping out $60 (list) of cable and dock, so price protection won't apply to the 60 GB iPod Photo that I bought on Monday.
So far (AFAIK) in every one of these cases where the RIAA was "suing the wrong person" it was not because they were wrong about what was going on -- the documentation of copyrighted files being shared over p2p file sharing networks from an IP address is pretty straightforward -- it was due to "the wrong person" being the person that the ISP billed, and the person who was illegally sharing copyrighted music was someone else using the internet connection. As a legal defense, "my mom didn't do it, I did" is pretty weak.:-)
Does anyone know if the owner of an internet connection is legally responsible for the actions of people who use that connection (i.e. the way the owner of a car is liable for speeding tickets when the polics look up his license plate number, even if he claims he wasn't driving at the time)? It seems to me that if the owner's liable, then liabilityt issues will force people not to run public wireless access points. And if the owner's not liable, then enforcement for illegal online activity will become nearly impossible. Messy either way.
Slashdot Mirror
I agree that "an x86 Apple would probably be a proprietary Apple with an x86 in" but keep in mind that the reason that nobody ran competitors to Windows on x86 (Be, Linux, etc.) was that MS' licensing terms required licensees to pay for Windows even if they didn't ship it, and prohibited companies from shipping any competing OS installed. The result was that since companies were already paying for Windows, they installed it, and since they installed it, they couldn't install any other OS. This meant that even manufacturers that sold PC's specifically to run BeOS shipped them with Windows installed, and told people to take the BeOS CD's and do the installation themselves.
Needless to say, this wiped out BeOS in the OEM market. This tactic was of course eventually determined to be illegal, but by then Be was dead so it didn't matter.
"The transition to powerpc was not a change of architecture, it was an upgrade in architecture. x86 is an entirely different architecture with an entirely different instruction set."
Wow, is this wrong. The 680x0 and the PowerPC have pretty much nothing in common, other than that Motorolla is one of the manufacturers of the PowerPC chips. But the only way that PPC's run 680x0 code is that they run a software emulator.
When Apple moved from 680x0 to PPC they could have easily moved to any other chip architecture. In fact, they (kinda) had MacOS running on SPARC, x86, etc., two different ways. First, the MacOS emulator for UNIX, and second NeXTSTEP -- between the two, that covered running "yellow box" and "blue box". So the choice of PPC wasn't due to any similarity to the 680x0, but purely on other technical and strategic issues (i.e. PPC being faster than x86 at the time and with a cleaner roadmap, the possibility of launching PPC as a competitive platform to x86, etc.).
For perspective, think about that processor companies use when they have a choice. All of the next generation videogame consoles run PPC. Many embedded controllers (e.g. in cars, Tivo's) are PPC's. People only use x86's when they need to run legacy x86 software; PPC's cost less and run faster.
Finally, even if Apple switched to the x86 from PPC, they wouldn't ship MacOS for generic PC's, because there we far too many quality/user experience issues. They toyed it a while back (I used to run Rhapsody for x86) and the idea died because it was impossible for them to generate enough device drivers to have even the slightest change of working on a random PC -- Rhapsody for x86 only ran on specific motherboards, with specific SCSI controllers, specific ethernet interfaces, specific video cards, etc. This was fine for someone building a machine to run that OS, but made it impossible to support more casual switchers.
"Free speech does not include the "right" to deny others their rights of assembly and free speech by shutting their convention down. The protesters did not respect the rights of these others."
Speaking as someone who lives in NYC, and thus saw this all first-hand, this is a misrepresentation of the protestors, and of those that arrested the protestors. The protestors weren't trying to shut down the convention -- they were trying to make their opinions known to attendees of the convention. This is _precisely_ the kind of speech that is constitutionally protected (i.e. political speech).
The police and organizer's goal was to prevent the protestors from getting their message out, by (1) restricting their right to free speech in ways that are (IMO) unconstitutional, and (2) arresting as many people as fast as possible, simply to keep them tied up with the police, where the attendees couldn't see them. The result was that 90%+ of the arrests were later thrown out, often because independent videotapes showed that the police straight out lied...
"it would not have happened if the extremist thugs had not organized an effort to "shut down" the convention and proceeded to harass the convention-goers."
Yeah, free speech is sometimes inconvenient, but despite the efforts of some extremists, we still have constitution that protects our freedoms.
"Surely the can't expect that their raids of arrests will provide them with more sales."
If I understand your claim, I think you're wrong.
Yes, I think that they do in fact think that when people buy bootleg copies of DVD's that they don't buy legit copies of the DVD's, so shutting down illegal manufacturers (i.e. factories that manufacture DVD's that they don't pay royalties on, and street vendors who burn DVD-R's and don't pay royalties) they reduce the supply of bootlegs and thus increase sales of legit DVD's.
Is there something there that you disagree with?
"Isn't this (if it actually works!) a type of denial of service attack?"
Nope. A denial of service attach means that the attacher is initiating connections to a server in order to consume the server's resources, denying the service to anyone else.
I believe that all this company is doing (if their technology works) is to join a public network and wait for requests, then sending back bad data. This is a well understood technique.
What this isn't:
- A "denial of service attach" since they're not initiating any connections to anything.
- A broad attack on the p2p networks, since the bad data is only served in response to specific files.
The only interesting bit is that if they've figured out how to create bad data that passes a hash check then the protocol's error correction won't work. This doesn't change anything with FastTrach (which effectively has no hash checking) but could be used to inject bad data into BitTorrent, eDonkey, etc.
I tried going to the site, and it's slashdotted. There's a nice irony in a company named Ioband not being able to deliever the I/O. Of course, they're being slashdotted by people who aren't using Ioband's technology...
"every country that has recognized gay marriage (among them Belgium and the Netherlands, if i remember correctly) has experienced a drop in the number of marriages"
t s/table_steep_can_pop_curve.htm) and has been dropping since. You might as well argue that gay marriage causes low rates of divorce, since Massachusetts legalized gay marriage last year, and has the lowest divorce rate in the US. Of course, the real answer is that people in Massachusetts are (compared to the national average) better educated, and thus have better jobs with better incomes, and get married later in life, which leads to them being both being less likely to divorce, and also more supportive of progressive social issues. Contrast this with southern states with large "born again christian" populations, who strongly oppose gay marriage, and who get divorced far more often than the national average -- you wouldn't argue that opposing gay marriage causes high divorce rates, because the coorelation is caused by a third factor, "born again christian" social attitudes that lead to both opposition to gay marriage and a high rate of failed marriages.
I think you might have the relationship wrong -- it's well known that (as observed in dozens of countries for decades) as a country matures (and people are better educated, have more income, have longer life spans, lower infant mortality, etc.) their birthrate drops. Gay marriage is a very recent phenomenon, so it can hardly be the cause of a trend well established decades ago. I suspect that the real relationship is that as people get better educated, have better incomes, etc., they also tend to be more accepting of gays. So, for example, legalizing gay marriage last year in Canada didn't cause population growth to slow, because the rate of population growth in Canada peaked in 1951-1956 (see http://www.populationinstitute.ca/tables_and_char
"We have a separation of State and religion and that should be followed. The church has no right to impose law. Although like any other organization they try to influence the laws."
Actually, the deal is that churches get to not pay taxes to the government, in return for which they can't directly engage in politics. This is why when churches get too close to endorsing a candidate or telling people how to vote, they risk losing their tax free status.
"How do you know she wanted to die? Oh, you took the husband's word for it even though a court said there was a dispute."
Actually, as documented in the court cases, there were three other witnesses that documented her saying on six occasions things that indicated that she didn't want to have her life artificially sustained with no chance of recovery. Like most young people (she was young 15 years ago when this happened) she didn't have a written Living Will, so the court had to make a judgement based on the evidence.
The parents had no evidence other than their opinions of what she wanted. They later said that even if she'd had a signed Living Will they still would have fought it, which pretty much proves the point to me that they knew what she wanted, but didn't like it. I can understand why her parents would be in denial of the reality of her death (as a parent, I can think of nothing worse than going through 15 years of torture) but I can't understand why anyone else would support their delusions -- it seems unethical to me to encourage "hope" in desparate people given that she was quite clearly gone.
Another aspect of this that confuses me is that from what I can tell the Catholic Church disapproves of artificial means of sustaining life (because if it's God's will that you die, you shouldn't try to avoid it), but her parents claim that as a Catholic she would want to live no matter what. Does this make sense to anyone else?
"At least Republicans showed up for the vote. Your beloved Democrats didn't even appear."
Actually, 50 Democrats voted for it, too. Pretty embarassing that they'd vote for something so pointless. But as a friend put it, their calculation was that voting for it would thrill the 10% of the population that really wants it, would be ignored by the 90% that would oppose it, and didn't cost anything.
Of course, according to polls, a huge majaority of the country, and a majority of evangelicals and republicans, opposed the congressional meddling, and Bush's ratings are lower than ever. So worse than pathetic pandering to a republican fringe group, it was completely unsuccessful pathetic pandering that managed to not only not satisfy the fringe, but alienated the mainstream. There's hope for rationality to kick in yet.
"Because nobody else has developed such an API for Windows. It's not impossible for one to replace IE's API if they really tried."
If MS provided IE's underlying functionality as a DLL with a clean API, this would be true. Think of MacOS X's Safari engine, KDE's KHTML, etc. But in order to "embed" IE in the operating system, they split IE's functionality and mixed it across many DLL's, DLL's that can't be removed because they have other functions that the OS relies on. This was smart of MS, because it made it "impossible" to remove IE, and nearly impossible for any other browser to provide IE's API.
"I mean seriously, if you hate IE that much, why are you even still using Windows?"
Because I can tell the difference between a web browser and an operating system. When I use Windows, I use Firefox and Thunderbird, so I don't get bothered (as much) by pop-ups, email scripting viruses, etc.
"I work retail, and our store received 60 PSP's and sold 10 the entire day."
I am a gadget fiend, and I looked at PSP's at two stores today and didn't buy one.
EB had the PSP in stock for $250, but had no demo units, though there were three employee units running movies against the wall behind the sales counter. Apparently they'd had a demo unit for people to test, and the unit got badly scratched so they stopped letting people see the PSP. As cool as the PSP looks from a distance, I'm not buying a $250 piece of electronics that I can't even hold first.
Then I went to KB Toys. All of their units were boxed, so I couldn't even see one one. Even more insulting, they were so convinced that people would go crazy over the PSP that they refused to sell it except as a "bundle" with two games for $350. To add insult to injury, you had to pick one game from list 'A' and one from list 'B', so you were forced to buy a crappy game!
If someone had let me hold a PSP for a minute, I'm sure that the unit would have sold itself. But with retailers this stupid, no wonder it's not selling.
It kinda reminds me of a visit to an electronic retailer long ago when the 3DO was hot. The store had signage and a demo kiosk set up, but off. When I asked the salesman in the nearly empty videogame area, he said that he unplugged it because it attracted such large crowds that he couldn't do his job (which apparently didn't involve selling things to people). His stupidity lives on.
So I can't say whether Sony did a good job launching the PSP (aside from NYC being blanketed in nice PSP posters), because the retailers were so braindead that the PSP never had a chance.
"The non-traceability of the Open Source process leaves any given product open to contamination from copyrighted/patented IP."
This claim gets the real world situation backwards.
All major open source projects retain a complete and precise development history through use of a a source code repository (e.g. CVS). This source code repository is open to public inspection, so anyone who wishes can determine the exact time and submitter of every line in that project. This has the effect of discouraging cheating, because the cheating is easy to detect, and the perpetrator is easily identified.
Proprietary software, on the other hand, may not have such a record o contributions, and even if one exists, it's certainly not open to public inspection (short of a lawsuit). So if you question the origin of some aspect of a proprietary system, you have to ask the company (i.e. sue them) for the information that you want. This is a bit of a catch-22, since you can't provide evidence of cheating until after you sue them in order to reveal the evidence.
It's certainly true that someone could illegally submit code that they don't own into an open source project, but the same is true for a proprietary project. And if someone thinks that their IP has been incorporated into an open source project, they can easily inspect the project's source code repository, and determine where the code came from and when, which should clarify the situation (and if someone submitted code illegally, smack them and remove the code from the project).
The only case where there's a problem is with proprietary code bases, where it's very difficult to determine whether IP has been illegally used, and if so it's extremely difficult to determine the source of the code.
Note that despite the theoretical risk of commercial IP getting into Open Sourrce projects, in practice I can't think of any cases where that's been shown to have occurred (even SCO gave up making all such charges against IBM), perhaps because the open source projects are open to inspection so such cheating is discouraged, while proprietary products are revealed fairly regularly to include open source software (or illegally used proprietary software), perhaps because the perpetrators thought that nobody outside their company would see the source code, so the cheating was "safe".
"no major distribution, and definitely no kernel, should leave the foundry without knowing who touched it."
You should be happy, because that's already the case for open source software. If the same were true of proprietary software, then we could put the whole issue to bed.
"A similar thing which happened to the story submitter happened to me once. My ISP got a call from the RIAA that someone was sharing loads of music online. So my ISP just looked for the biggest bandwidth user and shut them off. Well, that someone they shut off was me. But I wasn't the one sharing music. I was seeding Knoppix. A phone call to my ISP quickly resolved the situation."
Wow, what a lame ISP! The RIAA complaint would certainly have included an IP address, and they nail you instead of looking up who had that IP address? Amazing!
"Last I checked, apple pays 60c per song and resells them for 99c. That is approximately a 40% margin.... You can hardly say iTMS is next to nothing in margin. ...
You were pretty harsh for someone who is so totally wrong. "
To back up the analysis, let me provide some numbers.
If a track sells for 99 cents, the record company gets perhaps 70 cents (which then pays royalties to artist, composer, label, etc.). Of the remaining 29 cents, the credit card company charges perhaps 15 cents (depending on how well Apple negotiated them down from the 25 cents that's their starting position). So now you've got 14 cents left, which has to pay for implementing a large scale e-Commerce system that can sell music, plus the hardware and bandwidth costs, plus of course the cost to have the engineers build and run the system, the customer support representatives answer phone calls, the cost of marketing, etc.
Given how iTunes' sales volume has ramped up, I'd hope that iTMS is profitable, but it's a rough business, and for a company that can make high margins on hardware, it wouldn't make sense to spend a lot of money to enter a low-margin business. Buf if they can brek even selling music, and by doing so enable the iPod business to grow (and somehat lock anyone who bought music into buying iPods forever), that's a very smart move.
Yes, there are a few stores selling music slightly cheaper than iTMS, but they do so in somewhat misleading ways. For example, Walmart sells music as a loss leader to generate traffic into the rest of their store. That's also how they sell CD's cheaper than everyone else -- they are willing to sell one heavily promoted CD at a loss, but they place the music section as far from the front door as possible, and sell everyone who goes to the store to buy the music $100 worth of other stuff. And the company advertising 79 cent tracks only sells _some_ tracks for 79 cents -- their tracks range up to $1.49 (last time I looked) so their average prices was about the same as Apple's, and the usage rules were all over the place (e.g. you can burn some music that you buy to CD, other music you can't, etc.). Then there was a company selling music at "reverse auction" prices, where music started cheap and got more expensive the more popular it was, which is cool, but the reality is that they paid the same for their music as everyone else, and decided that they could take the loss in order to get some PR for being a cool place to buy music. Napster is at least interesting, since they sell someting completely different -- essentially they're not sellig music, but are selling subscription access to a library on demand for as long as you keep subscribing, but once you stop subscribing all of the music goes away. Since they're not really selling music that you own, their costs are lower (or at least different). Their problem, though, is that what people want to do is buy music and own it, not rent it. In particular, while in theory the "infinite jukebox in the sky" sounds great, and accountants certainly like the idea of a permanent revenue stream, but in reality teenagers hate the idea of committing to a permanent $15/month subscription fee forever.
"Maybe all the great Windows developers all have jobs"
.Net programmers that were "great", but we don't want to require users to install .Net runtime (which doesn't support many older machines, and in any case is a very large download/install), so we're coding in C/C++. This filtered out developers that only want to use the latest cool stuff from MS. Similarly, we're not using Java on the desktop for the same reason (and which would have been _way_ easier to hire for and build than C). .Net. So in terms of what's being taught, C/C++ and Windows aren't on the radar, so that filtered out hiring anyone straight out of school.
This could explain some of it, but I also went to universities and talked with students, and at least in the NY area CS students aren't getting trained in writing Windows software.
So, to be more precise, I'll say that we found:
- a fair number of "ok" windows developers. But we're shipping an app that has to be "great" consumer software, so it has to look great, be very efficient, take advantage of all of the latest cool Windows API's to do things nobody else is doing, etc., which filtered out all of the corporate/VB developers, etc.
- A few
- We went to schools and talked with their best CS students. They're being taught Java and Python, and play with Ruby. I think one mentioned
So we did find a great Windows developer, but it was like pulling teeth. Perhaps if we were in Seattle instead of NYC it would have been different, but in NYC banks are all doing J2EE/web app's, web shops are using DHTML and Flash, etc., so nobody's doing Windows development here.
"what percentage of the market is Linux going to have to take before hackers switch over? I think it'll be well over 50%."
It's easy to show that market share alone doesn't cause security problems. Historically, there's a market where non-Windows software has clearly dominate Windows software -- web servers. The numbers vary over time, but Apache usually is about 2x more popular than IIS as a web server. And if you look at the data on web site exploits, despite the fact that most sites (and most high-profile sites) run Apache/UNIX, 2/3rds of the sites compromised run IIS/NT, and only 1/3rd ran Apache/UNIX (including Linux, BSD, etc.). So for that data (a few million web sites, over a decade), NT/IIS appears to be about 4x as vulnerable to compromise as Apache/UNIX relative to market share.
My conclusion from this data is that security problems are not caused by popularity (in which case, Apache/UNIX would have been less secure than NT/IIS), but by meaningful differences in the design and implementation of the different operating systems.
I started a company a few months ago that's building consumer software that runs on MacOS X and Windows (and Linux, etc., eventually). Our strategy is to build the core in tight C code, and then build platform-specific applications in the appropriate language, so the result is a great ObjC Mac app, a great C++ Windows app, etc. While I like Java, Ruby, etc., our goal is to make the app small and efficient, so asking people to install 30 MB runtimes is out. Interestingly, it was easy to recruit first-class Mac and Java (server) developers, and nearly impossible to recruit a really great Windows developer. It turns out that the best CS students are _all_ working in modern cross-platform environments (e.g. Java, Python, Ruby), most use Mac's, almost none are using C++, and nobody even _considers_ writing Windows applications any more. While this is kinda neat in one respect, it's a bit surreal that the vast majority of great developers won't write software that runs natively for the platform on 95% of desktops. Weird.
I thought about using VMware, but I wanted to make sure that I saw the actual behavior on a normal PC, and I suspected that low level networking running within VMware might not match.
I've got a four digit ID. Neener, neener.
On a more serious note, I think that this is a fantastic piece of analysis. I did a simple version of this last year (nothing formal enough to publish, but interesting) and it took days, because KMD, etc., so thoroughly destabilize a PC on installation that you have to spend hours cleaning/reinstalling/etc., each time.
">> One suggestion is that labels want to introduce variable pricing - so they can charge more for top selling tracks.
> You know what? I'm all for it... if it means they are willing to discount the less-popular stuff in exchange.
So, you think that the record companies, who are already in a frothy panic because they think they're losing all their profits to those "p2p pirates" are going to suddenly become magnanimous and cut you a break by lowering their profit margin?"
Record companies don't discount back catalog sales because they're generous -- they do so because they have a huge inventory of older music that people won't buy for as much as the "hot new hits", so they can either sell it discounted or not sell it. So the best way to maximize their income is to dicount catalog sales, either directly (e.g. lower track and album prices) or indirectly (e.g. selling boxed sets, "best of" albums).
"Lets see, it is using an over hyped keyword based on an Apple product when just about any mp3 playing device works. And all it is is using RSS feeds to find and pull down new music/shows."
I think you're missing the point of why PodCasting is so cool. It's not interesting because it's hard technology, it's interesting because it's trivial technology that is trivial to implement, thus has been widely implemented, and has (and will have) a huge impact on people's lives.
Remember when the Enterprise crew list was leaked, and everyone thought that it had to be a fake because it was so derivative of the previous Trek shows? I wish that analysis had been correct -- the cast feels like the cast of previous Trek's rotated slightly (this time the southerner is the Weapons Officer instead of the Doctor, and we'll have alien instead of a hologram). It's a shame that the writing is so bad -- the acting is first rate, the effects are great, etc., but until the last few episodes I was _hoping_ that they'd cancel Enterprise (and fire B&B) and try again in a year or two.
Me, I'm annoyed that they stripped out the dock and cable because it lets them drop the price of the 60 GB iPod Photo by $150 but call it a (trivially) new product by stripping out $60 (list) of cable and dock, so price protection won't apply to the 60 GB iPod Photo that I bought on Monday.
Damn.
So far (AFAIK) in every one of these cases where the RIAA was "suing the wrong person" it was not because they were wrong about what was going on -- the documentation of copyrighted files being shared over p2p file sharing networks from an IP address is pretty straightforward -- it was due to "the wrong person" being the person that the ISP billed, and the person who was illegally sharing copyrighted music was someone else using the internet connection. As a legal defense, "my mom didn't do it, I did" is pretty weak. :-)
Does anyone know if the owner of an internet connection is legally responsible for the actions of people who use that connection (i.e. the way the owner of a car is liable for speeding tickets when the polics look up his license plate number, even if he claims he wasn't driving at the time)? It seems to me that if the owner's liable, then liabilityt issues will force people not to run public wireless access points. And if the owner's not liable, then enforcement for illegal online activity will become nearly impossible. Messy either way.