You're greatly overstating the strengths of Perl's DBI. I love the system but to say "you can change backend databases trivially, merely by changing one line of code" is wrong in nearly any real-world application, in my experience.
Just one quick example, the highly common operation of obtaining the value of an autoincrement column for a row you have just inserted. That is to say, obtaining (usually) the id of the row you have just inserted. This is not abstracted across RDBMS platforms. So for MySQL it is "$statement_handle->{mysql_insertid} ". If you switch to Postgres, you have to change this in each and every case, unless you have written a custom abstraction wrapper on top of DBI.
Actually, I think the whole question of whether channels should be sold a la carte is secondary to the whether we'll need cable networks in the first place.
Someone brought up the CD analogy, which I think is apt, and we've all seen how well iTunes Music Store has done with single sales. But to get to that point with TV you'd be talking about selling individual episodes of individual shows. Which I think ultimately is the model that will prevail, either that or people subscribing to one season of a particular show.
This isn't some fantasy that needs Internet-video-stream-ready set-top boxes and complex electronic billing schemes. It's happening right now with TV shows on DVD, including HBO shows like the Sopranos and Six Feet under and cult shows that never would have found an audience otherwise.
When a better technical infrastructure does emerge, for selling shows through the set-top or aggregating them with commercials on DVRs, the role of cable networks is going to shift away from distributors and toll collectors and subsidizers and more toward branders of quality content and promoters, like a record company in the music business.
The effect will be to increase the market for paid TV programming by lowering the bar to consumer and vendor entry. How many people would start watching new episodes of the Sopranos if they didn't have to take all of HBO for $18/month, or check out Adult Swim or Stargate SGI if they didn't have to shell out $60/month for the whole digital cable package as we do in the SF Bay Area?
Re:Free software lacks usability testing
on
GNOME for Grandma
·
· Score: 1
All may not be lost, perhaps a software tool could be written to make such usability testing easier. It could record a user's desktop(perhaps using something like VNC), while also recording their audio commentary on what they are doing.
Right. So your system would test GUI usability... for people who are comfortable setting up Virtual Network Computers... and own and can hook up PC microphones... and are capable of recording to their PC... and can ship everything over the network to a usability engineer... using the very GUI you are usability testing.
They already have a product (so no money needed to front the development).
According to this page on the Opera site, the product was, in fact, self-financed until 1999, when the company borrowed $15 million, perhaps to finance the phone browser, perhaps to compete more aggressively on the desktop. The company says the money was borrowed from "financial investors," which could mean a bank, venture fund, private "angel" investors, relatives or personal credit cards.
All of these investors have one thing in common: they expect to be paid back, plus a premium for putting their money at risk. Imagine being tens or hundreds of thousands of dollars in debt and then imagine there is a way to pay off all of those debts and also avoid going to bankers for new capital for a long time, possibly forever. This is the appeal of an IPO.
Also, even if there is no debt to pay off, issuing shares has the distinct advantage over any other kind of debt in that it does not have to be paid back. If you blow the money, you may get booted personally as CEO but your company will not be foreclosed upon and sold or liquidated... excepting a hostile takeover. Of course, this "free money" does come at the cost of control of the company.
PS I am surprised no one has called Orlowski at the Reg on his misuse of the term "revenue". Revenue and profit are two different things. If Apple made no *revenue* on the store, it would be a much much bigger story.
I'm sure you're correct they told other reporters, and I can see why you wouldn't use the material in your column. But I read a lot of the coverage, with a bias toward national and SF Bay Area publications, and things up on Google News, and I never found a story focused on the business side of iTMS. In general, I think the lack of hard business-oriented coverage of the "store" was a major failure on the part of the tech news media, and I think it makes total sense that someone might be surprised that Apple is not in the black on the store.
Even at ten cents on the dollar they stand to net sales of around $13 million the first year, which is based on extrapolating the first 6 months of mac use and first week of Windows use -- a dangerous game to be sure. Anyway, their costs are probably well in excess of $13m, but by how much? After all, they had a lot of tech already in place with.Mac, perhaps.
The point is, if iTMS is so important to warrant all the prominent covwerage it has received, then these are reasonable things to start asking about and estimating on. It is sad to see so many reporters just regurgitate spoon-fed sales information from Apple. It's almost enough to make me wish I was covering tech again:-)
Could you please post the URL of the article that included this information. It's good to hear you are not part of the herd of journalists who compeltely ignored the business angle of this story and made Apple talk about dollars and cents in public.
The report I linked to includes volume info, and dividing the number of CDs by the sales gave me about $15 per CD. IIRC iTMS is at about $10 for buying a whole CD. So multiply my generous $520m estimate by 1.5 and you're at $780 million, still less than a tenth of CD sales (although impressive to be sure)
I didn't even need to click. I've been doing the math to figure how big a deal this iTunes thing is (not big, at least not yet).
Here are the numbers. The U.S. record industry sold $12.6 billion worldwide in various formats (almost all CDs) in 2002. This is off a bit from the peak $14.6 billion in 1999. It's important to keep in mind that, even at those levels, we're talking about nine weeks revenue for IBM.
Assuming the Windows side of iTunes Music Store continues to sell at the initial rate of 1 million songs/$1 million revenue in the first 3.5 days, that's only about $104 million per year. The Mac side sold $13 million in tunes in the first six months, so we'll put that side at $26 million per year.
That's $130 million per year for all iTMS. Even if the store doubles its sales, and then the other stores collectively match its sales, you'd be talking about total online sales of $520 million per year, still a drop in the bucket.
The growth will need to get exponential before there is any comparison with offline music sales. I'm not saying it won't happen, but that's what we're talking about, and that's how I instantly new the hed on the posting was wrong.
You know what other game is being raised? Slashdot's masturbatory anti-Microsoft jihad posts. Just yesterday morning Slashdot had four Microsoft-borg-logoed stories, with only one other post breaking them up, all posted in the span of three and a half hours. I am glad to see the bashing has not let up today.
These threads invariably involve, at the top mod levels, derogatory comments about the quality of Microsoft code and products, conspiracy theories about the true motives behind Microsofts intentions (always), sarcastic jokes agreeing with the action in question, a sad reflection on how new users, PHBs and/or the world at large is accepting this action, and an impressively-inventive-if-completely unneccesary variety of miscellaneous other anti-Microsoft rhetoric.
I am not going to rehash the old and tired arguments about Microsoft, or even say I disagree with much of it. That is beside the point.
What is important is that open source in general and slashdot in particular should be different, and they are utterly NOT. Steve Ballmer comes out and spreads some FUD on Linux. Ya, it's FUD, and it's not true, and he's fundamentally wrong about quality and open source, and besides Microsoft just this and that and blah blah blah. So what.
I can see how the first two or three or ten times you hear this shit from Microsoft you want to scream from the mountaintops how wrong it is. What I utterly will never ever understand is how you can get off, get this big rhetorical hard on, four and five times a day week in and week out over the SAME BULLSHIT. It's FUD now just like it was FUD last year and FUD the year before that and, as far as the slashdot crowd is concerned at least, FUD in 1976 when Bill Gates wrote an Open Letter to Hobbyists.
It would seem to me that, confronted with all of this disagreeable stuff coming out of Microsoft, the slashdot crowd would eventually learn the productive and elevated response is to
A> Shrug.
B>Take the high road and acknowledge every sliver of truth in every criticism, ignoring the juvenile manner in which it may have been delivered, and use this reflection to further improve open source. Parse FUD for constructive crisiticism. If there is none to be gleaned see A>. Is there *anything* about Linux's patching model or security that could be improved? Is there the slightest kernel of truth in what Ballmer says?
But when I think about it I realize the benefit of anti-Microsoft jihad posts filled with propagandist comments isn't to convey any new information or spark new insights but to further reinforce and perpetuate the community formed around slashdot. Read Clay Shirky's brilliant A Group Is Its Own Worst Enemy. External enemy, religious veneration, it's all here. It's here to perpetuate the group, as human groups naturally want to do -- even when such patterns are against the interest of the original or stated goal of the group. A choice excerpt:
"Anyone who was around the Open Source movement in the mid-Nineties could see this all the time. If you cared about Linux on the desktop, there was a big list of jobs to do. But you could always instead get a conversation going about Microsoft and Bill Gates. And people would start bleeding from their ears, they would get so mad. "
I'm sick of it, so what, everyone seems to love it, I'll just go now and click a preference and never look at the borg crap again. I just hope in time there is enough other content to read.
they literally ALMOST KILLED HIM... One of them even pulled out a collapsable baton and hit him while he was down... they started beating her with the baton... In-bar muggings and shootings are on the rise
Hmmm, somehow I doubt we'll hear about this in Michael Moore's next documentary.
In all seriousness, thank you for the word, but I'm not sure it is yours any more, any more than Unix can be defined by AT&T or SCO. Also, nearly any serious document on the history of weblogs includes both the sites I linked to as among the earliest examples of the format, so arguing against diarism seems a bit revisionist.
Wallowing in narcissism has nothing to do with weblogs, although the mass media have been propagating that slur since the earliest days.
Does having a popular weblog somehow give _you_ the right to define what weblogging is or should be, what is included and excluded? Or are you basing this on some survey of weblogs out there?
I certainly don't consider your non-personal blog any more authentic than things like this that were exploring personal topics eight years ago. Dave Winer has been posting psuedo-diary entries on Scripting News and DaveNet since the mid-1990s.
How dare you try to define weblogging for the rest of us.
I found SQLZoo.net an extremeley useful tutorial. One of the cool things is you can pick the SQL engine used for the exercises, including MySQL, Postgres, Oracle and SQLServer.
Before going hands-on, Philip Greenspun's SQL for Web Nerds offers a relatively speedy but deep overview.
... and the legal scholars I talked to found plenty for Hummer Winblad to worry about. Of course, this was before Bertelsmann became involved.
My article, from Upside.com: Hummer Winblad could answer for Napster's sins Legal experts say there is a good chance the flush venture capital firm Hummer Winblad stands to lose more than its $13 million investment in Napster Inc. if the music-swapping firm is fined for music piracy.
Ya really. That stat is totally useless. Because there are so many first world countries less than 1/55th the size of America, boasting populations of under 5 million people.
Moore does state in the movie that the per capita figures are equally harrowing, and while he doesn't give stats, as he should, the viewer tends to believe him, because he throws out insanely low numbers (by comparison) for countries like Japan, Germany, Britain and Canada, which have decent-sized populations.
I went into the documentary expecting Michael Moore to have already decided how he felt on the issue and to rail blindly and shrill-ly for liberal causes like gun control, but I was surprised. The movie was nuanced, funny and really, really good.
Right, a STOVL strike fighter is a waste in the new era of low intensity conflicts and rapid deployment.
Because air power was not an absolutely critical factor in winning the war against the Taliban with fewer than *200* U.S. troops on the ground in Afghanistan at any one time.
You should really check out the Frontline special on that war, in particular the part where one of the spec-ops soldiers says his team didn't really win over the warlords they were working with until they pointed lasers at some Taliban strongholds and made them go away (with Paveways or somesuch).
Give me a break. The JSF will be able to operate from very, very short landing strips in the middle of nowhere and deliver the sort of smart munitions that have become a critical part of modern LICs -- and more (eg the laser).
The JSF hasn't reached its final form, much less deployed. How can you say it won't work? Did anyone predit the F-4 Phantom II would be so useful to all branches, or that the UH-1 would, or the A-4 (navy/mc/israel), or the FA-18 (navy/mc/australia/canada/etc), or that the British would retake the Falklands using only Harriers?
if your machine crashes half way through executing a DB update, all the locks in the world won't help you to guarantee a consistent state after restart.
The classic example cited by Philip Greenspun is an online bank. If someone transfers $1000 from checking to savings, you need to be absolutely certain that the transfer is ONLY completed if savings is credited AND checking is debited. If someone pulls the power cord halfway through, neither you nor your customer want to be out $1000.
So locking absolutely doesn't equal transactions. To "fake" transactions you would need to record what you are about to do, then delete that record only after you do it. That way, the transaction can be "rolled back."
Slashdot Mirror
You're greatly overstating the strengths of Perl's DBI. I love the system but to say "you can change backend databases trivially, merely by changing one line of code" is wrong in nearly any real-world application, in my experience.
Just one quick example, the highly common operation of obtaining the value of an autoincrement column for a row you have just inserted. That is to say, obtaining (usually) the id of the row you have just inserted. This is not abstracted across RDBMS platforms. So for MySQL it is "$statement_handle->{mysql_insertid} ". If you switch to Postgres, you have to change this in each and every case, unless you have written a custom abstraction wrapper on top of DBI.
Actually, I think the whole question of whether channels should be sold a la carte is secondary to the whether we'll need cable networks in the first place.
Someone brought up the CD analogy, which I think is apt, and we've all seen how well iTunes Music Store has done with single sales. But to get to that point with TV you'd be talking about selling individual episodes of individual shows. Which I think ultimately is the model that will prevail, either that or people subscribing to one season of a particular show.
This isn't some fantasy that needs Internet-video-stream-ready set-top boxes and complex electronic billing schemes. It's happening right now with TV shows on DVD, including HBO shows like the Sopranos and Six Feet under and cult shows that never would have found an audience otherwise.
When a better technical infrastructure does emerge, for selling shows through the set-top or aggregating them with commercials on DVRs, the role of cable networks is going to shift away from distributors and toll collectors and subsidizers and more toward branders of quality content and promoters, like a record company in the music business.
The effect will be to increase the market for paid TV programming by lowering the bar to consumer and vendor entry. How many people would start watching new episodes of the Sopranos if they didn't have to take all of HBO for $18/month, or check out Adult Swim or Stargate SGI if they didn't have to shell out $60/month for the whole digital cable package as we do in the SF Bay Area?
All may not be lost, perhaps a software tool could be written to make such usability testing easier. It could record a user's desktop(perhaps using something like VNC), while also recording their audio commentary on what they are doing.
... for people who are comfortable setting up Virtual Network Computers ... and own and can hook up PC microphones ... and are capable of recording to their PC ... and can ship everything over the network to a usability engineer... using the very GUI you are usability testing.
Right. So your system would test GUI usability
Ya, that's grandma alright!!
They already have a product (so no money needed to front the development).
... excepting a hostile takeover. Of course, this "free money" does come at the cost of control of the company.
According to this page on the Opera site, the product was, in fact, self-financed until 1999, when the company borrowed $15 million, perhaps to finance the phone browser, perhaps to compete more aggressively on the desktop. The company says the money was borrowed from "financial investors," which could mean a bank, venture fund, private "angel" investors, relatives or personal credit cards.
All of these investors have one thing in common: they expect to be paid back, plus a premium for putting their money at risk. Imagine being tens or hundreds of thousands of dollars in debt and then imagine there is a way to pay off all of those debts and also avoid going to bankers for new capital for a long time, possibly forever. This is the appeal of an IPO.
Also, even if there is no debt to pay off, issuing shares has the distinct advantage over any other kind of debt in that it does not have to be paid back. If you blow the money, you may get booted personally as CEO but your company will not be foreclosed upon and sold or liquidated
PS I am surprised no one has called Orlowski at the Reg on his misuse of the term "revenue". Revenue and profit are two different things. If Apple made no *revenue* on the store, it would be a much much bigger story.
I'm sure you're correct they told other reporters, and I can see why you wouldn't use the material in your column. But I read a lot of the coverage, with a bias toward national and SF Bay Area publications, and things up on Google News, and I never found a story focused on the business side of iTMS. In general, I think the lack of hard business-oriented coverage of the "store" was a major failure on the part of the tech news media, and I think it makes total sense that someone might be surprised that Apple is not in the black on the store.
.Mac, perhaps.
:-)
Even at ten cents on the dollar they stand to net sales of around $13 million the first year, which is based on extrapolating the first 6 months of mac use and first week of Windows use -- a dangerous game to be sure. Anyway, their costs are probably well in excess of $13m, but by how much? After all, they had a lot of tech already in place with
The point is, if iTMS is so important to warrant all the prominent covwerage it has received, then these are reasonable things to start asking about and estimating on. It is sad to see so many reporters just regurgitate spoon-fed sales information from Apple. It's almost enough to make me wish I was covering tech again
Cheers
r
Glenn:
Could you please post the URL of the article that included this information. It's good to hear you are not part of the herd of journalists who compeltely ignored the business angle of this story and made Apple talk about dollars and cents in public.
Great point. Between this and the other post maybe we should say $520m on iTunes = $1.5b in lost CD sales.
The report I linked to includes volume info, and dividing the number of CDs by the sales gave me about $15 per CD. IIRC iTMS is at about $10 for buying a whole CD. So multiply my generous $520m estimate by 1.5 and you're at $780 million, still less than a tenth of CD sales (although impressive to be sure)
Sorry, I misspoke, those were U.S. shipments. RIAA claims $40b globally. Good point to raise.
I didn't even need to click. I've been doing the math to figure how big a deal this iTunes thing is (not big, at least not yet).
Here are the numbers. The U.S. record industry sold $12.6 billion worldwide in various formats (almost all CDs) in 2002. This is off a bit from the peak $14.6 billion in 1999. It's important to keep in mind that, even at those levels, we're talking about nine weeks revenue for IBM.
Assuming the Windows side of iTunes Music Store continues to sell at the initial rate of 1 million songs/$1 million revenue in the first 3.5 days, that's only about $104 million per year. The Mac side sold $13 million in tunes in the first six months, so we'll put that side at $26 million per year.
That's $130 million per year for all iTMS. Even if the store doubles its sales, and then the other stores collectively match its sales, you'd be talking about total online sales of $520 million per year, still a drop in the bucket.
The growth will need to get exponential before there is any comparison with offline music sales. I'm not saying it won't happen, but that's what we're talking about, and that's how I instantly new the hed on the posting was wrong.
Down modded? Looks like more narrow minded slashdot jihadism.
These threads invariably involve, at the top mod levels, derogatory comments about the quality of Microsoft code and products, conspiracy theories about the true motives behind Microsofts intentions (always), sarcastic jokes agreeing with the action in question, a sad reflection on how new users, PHBs and/or the world at large is accepting this action, and an impressively-inventive-if-completely unneccesary variety of miscellaneous other anti-Microsoft rhetoric.
I am not going to rehash the old and tired arguments about Microsoft, or even say I disagree with much of it. That is beside the point.
What is important is that open source in general and slashdot in particular should be different, and they are utterly NOT. Steve Ballmer comes out and spreads some FUD on Linux. Ya, it's FUD, and it's not true, and he's fundamentally wrong about quality and open source, and besides Microsoft just this and that and blah blah blah. So what.
I can see how the first two or three or ten times you hear this shit from Microsoft you want to scream from the mountaintops how wrong it is. What I utterly will never ever understand is how you can get off, get this big rhetorical hard on, four and five times a day week in and week out over the SAME BULLSHIT. It's FUD now just like it was FUD last year and FUD the year before that and, as far as the slashdot crowd is concerned at least, FUD in 1976 when Bill Gates wrote an Open Letter to Hobbyists.
It would seem to me that, confronted with all of this disagreeable stuff coming out of Microsoft, the slashdot crowd would eventually learn the productive and elevated response is to
A> Shrug.
B>Take the high road and acknowledge every sliver of truth in every criticism, ignoring the juvenile manner in which it may have been delivered, and use this reflection to further improve open source. Parse FUD for constructive crisiticism. If there is none to be gleaned see A>. Is there *anything* about Linux's patching model or security that could be improved? Is there the slightest kernel of truth in what Ballmer says?
But when I think about it I realize the benefit of anti-Microsoft jihad posts filled with propagandist comments isn't to convey any new information or spark new insights but to further reinforce and perpetuate the community formed around slashdot. Read Clay Shirky's brilliant A Group Is Its Own Worst Enemy. External enemy, religious veneration, it's all here. It's here to perpetuate the group, as human groups naturally want to do -- even when such patterns are against the interest of the original or stated goal of the group. A choice excerpt:
I'm sick of it, so what, everyone seems to love it, I'll just go now and click a preference and never look at the borg crap again. I just hope in time there is enough other content to read.
they literally ALMOST KILLED HIM ... One of them even pulled out a collapsable baton and hit him while he was down ... they started beating her with the baton ... In-bar muggings and shootings are on the rise
Hmmm, somehow I doubt we'll hear about this in Michael Moore's next documentary.
Wow, there are only five comments posted at this moment and already the link is dead.
Well, it's 404, not slashdot effect, so I'll save the snide comments about Netscape Enterprise Server.
Anyway, here's a working link. Should be good for at least a few minutes.
Mod parent up. :--)
In all seriousness, thank you for the word, but I'm not sure it is yours any more, any more than Unix can be defined by AT&T or SCO. Also, nearly any serious document on the history of weblogs includes both the sites I linked to as among the earliest examples of the format, so arguing against diarism seems a bit revisionist.
Cheers
r
Wallowing in narcissism has nothing to do with weblogs, although the mass media have been propagating that slur since the earliest days.
Does having a popular weblog somehow give _you_ the right to define what weblogging is or should be, what is included and excluded? Or are you basing this on some survey of weblogs out there?
I certainly don't consider your non-personal blog any more authentic than things like this that were exploring personal topics eight years ago. Dave Winer has been posting psuedo-diary entries on Scripting News and DaveNet since the mid-1990s.
How dare you try to define weblogging for the rest of us.
I found SQLZoo.net an extremeley useful tutorial. One of the cool things is you can pick the SQL engine used for the exercises, including MySQL, Postgres, Oracle and SQLServer.
Before going hands-on, Philip Greenspun's SQL for Web Nerds offers a relatively speedy but deep overview.
Totally sharp question. I had similar thoughts.
The article ran on the website, UpsideToday, now defunct.
Cheers, r
... and the legal scholars I talked to found plenty for Hummer Winblad to worry about. Of course, this was before Bertelsmann became involved.
My article, from Upside.com:
Hummer Winblad could answer for Napster's sins
Legal experts say there is a good chance the flush venture capital firm Hummer Winblad stands to lose more than its $13 million investment in Napster Inc. if the music-swapping firm is fined for music piracy.
then the Economist did a story:
Hummer's Napster bummer: Napster's backers under attack
It would be like Compaq/DEC
...
Or Sun/Cobalt
please don't propagate incomplete facts.
Ya really. That stat is totally useless. Because there are so many first world countries less than 1/55th the size of America, boasting populations of under 5 million people.
Moore does state in the movie that the per capita figures are equally harrowing, and while he doesn't give stats, as he should, the viewer tends to believe him, because he throws out insanely low numbers (by comparison) for countries like Japan, Germany, Britain and Canada, which have decent-sized populations.
I went into the documentary expecting Michael Moore to have already decided how he felt on the issue and to rail blindly and shrill-ly for liberal causes like gun control, but I was surprised. The movie was nuanced, funny and really, really good.
Right, a STOVL strike fighter is a waste in the new era of low intensity conflicts and rapid deployment.
Because air power was not an absolutely critical factor in winning the war against the Taliban with fewer than *200* U.S. troops on the ground in Afghanistan at any one time.
You should really check out the Frontline special on that war, in particular the part where one of the spec-ops soldiers says his team didn't really win over the warlords they were working with until they pointed lasers at some Taliban strongholds and made them go away (with Paveways or somesuch).
Give me a break. The JSF will be able to operate from very, very short landing strips in the middle of nowhere and deliver the sort of smart munitions that have become a critical part of modern LICs -- and more (eg the laser).
The JSF hasn't reached its final form, much less deployed. How can you say it won't work? Did anyone predit the F-4 Phantom II would be so useful to all branches, or that the UH-1 would, or the A-4 (navy/mc/israel), or the FA-18 (navy/mc/australia/canada/etc), or that the British would retake the Falklands using only Harriers?
if your machine crashes half way through executing a DB update, all the locks in the world won't help you to guarantee a consistent state after restart.
The classic example cited by Philip Greenspun is an online bank. If someone transfers $1000 from checking to savings, you need to be absolutely certain that the transfer is ONLY completed if savings is credited AND checking is debited. If someone pulls the power cord halfway through, neither you nor your customer want to be out $1000.
So locking absolutely doesn't equal transactions. To "fake" transactions you would need to record what you are about to do, then delete that record only after you do it. That way, the transaction can be "rolled back."