The problem is, you've basically screwed the whole thing up. You should be putting the most influential people in first. I'm not sure about the gamers you included -- I see at least one of the names I'd expect to see there, but don't know a huge amount about the field -- but in terms of designers you've basically missed the mark. By including multiple people from the same project (i.e. design of the Xbox) you appear to be showing bias towards that particular project. And by not picking many of the scores of designers who have real name recognition among the general gaming public (e.g. Carmack, Garriot, Braben & Bell) and instead focussing mostly on people who are actually much less well known, you give the impression that this list isn't really serious but more a sort of back-scratching thing that industry insiders might be interested in but the rest of us should ignore.
going over to AMD who can't match intel in performance per core
That depends on your workload. I have two machines here that run identical CPU-intensive processes, each of which consists of a local database server talking to a floating-point-heavy computation engine. One's a Core2 Quad processor, the other an Athlon X4, both clocked at similar speeds. The Core2 outperforms the Athlon on the database server part of the process, the Athlon on the floating-point.
That's simply not true. A 2.6 GHz Core 2 duo beats a 3.0 Pentium D in every bench mark tests. The reason why the 2.6 GHz Core 2 duo totally outperformed the faster 3.0 GHz Pentium D is due to the fact that the new Intel core architecture performs more calculations, read and write tasks per clock cycle that the older Pentium cores.
This is only true in most, not all circumstances. A Core2 chip can execute up to 3 arithmetic instructions per clock per core. In the right circumstances (and admittedly these circumstances are rare) a Pentium4 can execute *4* such instructions per clock (two ALUs, each operating for some instructions at double the chip's core clock rate). The problems with P4 were that getting the right combination of instructions was very difficult, and the penalty for branch misprediction was way too high. The P4 also has a much lower memory bandwidth than Core chips. In the rare circumstance that you had code that had all of these things right and was heavy on addition (basically the only instruction that worked at the double speed rate), and all the data fits in the cache, a P4 can theoretically outperform a similarly-clocked Core2. None of the standard benchmarks fit this definition, but there are probably applications out there that do.
That's what I was thinking too. The numbers we had when they announced it said that they needed to keep 1% of the readership for it to remain as profitable as before. Keeping 10% means they're making a lot more money than they were.
The 10% figure includes, I think, the 150,000 people who got free subscriptions. There are only 15,000 paying subscribers, which sounds a lot closer to that 1% figure (original was 1.2m unique visitors per day)...
Strong disagree. For a little site set up by you or me, yes, perhaps (although the last time I was running an ad-supported site it was doing somewhat better than that). For a mainstream media site with its own sales network supported by one of the largest media conglomerates in the world, I suspect revenues were a little higher. In fact, I'd be surprised if it represented much more than 400 ad views (i.e. £5 per 1000).
With their original viewing figures they would've had to have 200 ad views a week per user to make that kind of revenue which is a big ask.
I'd make that more like 40. Or maybe a lot less, depending on what actual revenues they got. Now, it's a while since I looked at a page on the Times site, but looking at comparable sites, they tend to have 3-5 ads per page, although some of those are text-only and therefore probably a bit cheaper than the rest. So we're starting to look at something like 20 page views per user per week, or slightly under 3 per day. That seems quite plausible to me.
My guess: they're not quite breaking even on this, but it's a close thing.
the voucher system for journalists to allow them access to the site did not work and they then had to set up paid for accounts. Depending on the numbers that would further distort the figures.
15,000 paid subscribers. You might be talking as much as 10% of the remaining subscriber base, i.e. about 1% of the 90% reduction figure.
They are saving on servers, bandwidth, etc. so who knows, it may even out.
Yeah. Best guess is they're losing thousands per day in ad revenue. A hosting plan that can handle the load (1.2m unique visitors per day -- we're not talking something on the scale of facebook here) should cost less than a tenth that.
the only users left are the ones that specifically want to insulate themselves from the Beeb (in other words, find comfort in the opinions of Murdoch/the Times and discomfort from opposing viewpoints).
Which, to be honest, I find 15,000 a remarkably low figure for. Of course, there's still the Daily Mail offering its news for free, and it would be interesting to see how the figures shifted if the Mail also went paywall.
It would be a shame, though, not to have a free news source to take the piss out of each morning.
Imagine how the advertisers feel. It's like 90% of the former Times web readership installed adblock overnight.
It's an interesting question. Some quick arithmetic based on the figures quoted suggests they're losing at least 850,000 page views per day (probably somewhat higher), but earning around £4,500 per day in subscriptions. Now, interestingly, at a rate of £5 per thousand views (not an unrealistic figure for ad revenues, certainly), this comes out as roughly break-even.
Of course, it is likely that they're losing more page views than this -- the figures were for unique visitors, and I suspect the average visitor is looking at a lot more than one page, so it seems unlikely they're actually reaching break-even, except that the advertising is now more targetted -- every viewer is proven to be somebody who is likely to spend money. They may actually be able to raise their charges for advertising space and, perhaps with just a few more subscribers than they have at the moment, or maybe by increasing the charge just a little, might reach break-even on the project.
Of course, if they were making significantly more than £5 per thousand, or if the number of page views per day was significantly higher than 1.2 million, then this breaks down. But it's interesting to note that you can view the figures in a way that shows this entire experiment not to be completely crazy.
Which is exactly the guys point: when you're spending that much money on a game, you can't afford to be creative, because if your new idea tanks you're left with a huge bill for developing it. Much safer to make something like another game that was previously popular.
Spend a tenth as much developing a game, and you can afford to take ten times as much risk, and maybe you'll get a runaway hit. Probably not, but it doesn't really matter that much...
Assume that we use plastic, and that some small percentage (no matter how much overall mass that is) will end up somewhere dangerous. Which is best/least bad: Plastics that don't degrade, or plastics that do degrade?
I think the answer is quite simple: if discarded on land, degradable is best. If discarded at sea, nondegradable is best. One possibility is to switch to nondegradables in coastal areas, but a more interesting one would be to develop a plastic that doesn't degrade in the presence of salt (or any other chemical present in the oceans but not commonly present in large quantities on land).
4,000,000,000 kg / 10,000,000,000 m^2 = 0.4 kg/m^2
Anyone else have a problem with this?
Indeed. The idea is total bullshit. 400gsm plastic is basically a medium thickness flexible sheet, something like a laminator pouch or similar. You can't build an island out of that shit.
Their argument appears to be that ISPs don't currently think they should pay for copyright infringement... so to prevent this being a problem ISPs should be made to pay for copyright infringement, at which point their objections will go away. Does that actually make sense in any way?
I predict a novel legal theory being invented that will protect the big corporation from a seemingly straightforward contract.
You mean like the well-known fact you have to enforce a contract within 7 years of it being breached, and that this breach apparently occurred 7 years and 2 months before this case was brought to court?
Parasites in shrimp can cause them to travel toward light and swim against gravity. The parasites act as a serotonin modulator. One particular antidepressant Fluoxetine does the same thing
If it's a serotonin-moderated behaviour, and fluoxetine triggers it, I'd be pretty happy wagering that most other SSRIs trigger it too.
So we need to dump a littler over 24 million metric tons of Fluoxetine into the gulf to see this concentration?
A back-of-the-envelope calculation based on a figure of 120 million people in the US prescribed antidepressants suggests annual consumption of SSRIs (by far the most prevalent kind) is in the region of 2 million tonnes, so it isn't unreasonably to believe that this kind of concentration could eventually occur (on the assumption that these chemicals do not readily break down in the environment).
Just call the RIAA and tell them that IP downloaded a song. They seem to be able to do all the John Doe stuff through the courts to find out who it was...
Actually, you can do that stuff yourself. File a claim with the courts for recovery of your possession, send a subpoena to the ISP, get the address, then either serve papers to continue the possession claim or hand the address to the state police.
DAB receivers in cars are not that common, and DAB reception in a moving vehicle tends to be pretty poor even if they are.
Which is odd, as one of the design goals for DAB was that the receiver should be able to move freely throughout the broadcast area and always get a signal, automatically switching to the strongest available transmitter whenever it changes.
It couldn't be something to do with the fact that the cheapest DAB radio I can find right now is £35 (£60 if you want something portable), whereas you can get a portable FM receiver for under £5? Nah, it must be to do with the regulations and standards!
Don't think it is, no. Where do most people listen to radio? In their car. It probably has more to do with the fact that DAB radios haven't been standard equipment in most cars until only about a year ago, and not even an option in many of them until not long before that. Most car manufacturers seem to have waited until the digital switchover was announced (currently scheduled for 2015, I believe) before offering them.
Dunno about the US one, but the objection UK telemarketers usually have to it is that it is kind-of expensive. My company used to do telemarketing as follow-ons from our (business-to-business) mailshots, but access to the TPS database costs about £5,000 per annum, which was hard to justify for us: our turnover was only about £30,000 at the time so it would have become our single largest expense and thus sunk a huge chunk of our profits. We therefore had to stop doing marketing calls.
That argument is getting old, too. I'm older than our Prime Minister and Deputy Prime Minister, and have been working with computers since my youth and have been online (via bulletin boards) since my college days. These things go back longer than the young guns realise, and if politicians don't understand them then there has to be another reason.
There are plenty of politicians who do understand. The only MP I've met on a personal level used to run an ecommerce company, and writes web apps in JSP in his spare time.
Slashdot Mirror
The problem is, you've basically screwed the whole thing up. You should be putting the most influential people in first. I'm not sure about the gamers you included -- I see at least one of the names I'd expect to see there, but don't know a huge amount about the field -- but in terms of designers you've basically missed the mark. By including multiple people from the same project (i.e. design of the Xbox) you appear to be showing bias towards that particular project. And by not picking many of the scores of designers who have real name recognition among the general gaming public (e.g. Carmack, Garriot, Braben & Bell) and instead focussing mostly on people who are actually much less well known, you give the impression that this list isn't really serious but more a sort of back-scratching thing that industry insiders might be interested in but the rest of us should ignore.
Although to be fair, Arkansas recently repealed their ban on "idiots" voting
It is quite hard to run a decent election when there are only two legal voters.
going over to AMD who can't match intel in performance per core
That depends on your workload. I have two machines here that run identical CPU-intensive processes, each of which consists of a local database server talking to a floating-point-heavy computation engine. One's a Core2 Quad processor, the other an Athlon X4, both clocked at similar speeds. The Core2 outperforms the Athlon on the database server part of the process, the Athlon on the floating-point.
That's simply not true. A 2.6 GHz Core 2 duo beats a 3.0 Pentium D in every bench mark tests. The reason why the 2.6 GHz Core 2 duo totally outperformed the faster 3.0 GHz Pentium D is due to the fact that the new Intel core architecture performs more calculations, read and write tasks per clock cycle that the older Pentium cores.
This is only true in most, not all circumstances. A Core2 chip can execute up to 3 arithmetic instructions per clock per core. In the right circumstances (and admittedly these circumstances are rare) a Pentium4 can execute *4* such instructions per clock (two ALUs, each operating for some instructions at double the chip's core clock rate). The problems with P4 were that getting the right combination of instructions was very difficult, and the penalty for branch misprediction was way too high. The P4 also has a much lower memory bandwidth than Core chips. In the rare circumstance that you had code that had all of these things right and was heavy on addition (basically the only instruction that worked at the double speed rate), and all the data fits in the cache, a P4 can theoretically outperform a similarly-clocked Core2. None of the standard benchmarks fit this definition, but there are probably applications out there that do.
That's what I was thinking too. The numbers we had when they announced it said that they needed to keep 1% of the readership for it to remain as profitable as before. Keeping 10% means they're making a lot more money than they were.
The 10% figure includes, I think, the 150,000 people who got free subscriptions. There are only 15,000 paying subscribers, which sounds a lot closer to that 1% figure (original was 1.2m unique visitors per day)...
£2 probably represents 2000 ad views.
Strong disagree. For a little site set up by you or me, yes, perhaps (although the last time I was running an ad-supported site it was doing somewhat better than that). For a mainstream media site with its own sales network supported by one of the largest media conglomerates in the world, I suspect revenues were a little higher. In fact, I'd be surprised if it represented much more than 400 ad views (i.e. £5 per 1000).
With their original viewing figures they would've had to have 200 ad views a week per user to make that kind of revenue which is a big ask.
I'd make that more like 40. Or maybe a lot less, depending on what actual revenues they got. Now, it's a while since I looked at a page on the Times site, but looking at comparable sites, they tend to have 3-5 ads per page, although some of those are text-only and therefore probably a bit cheaper than the rest. So we're starting to look at something like 20 page views per user per week, or slightly under 3 per day. That seems quite plausible to me.
My guess: they're not quite breaking even on this, but it's a close thing.
the voucher system for journalists to allow them access to the site did not work and they then had to set up paid for accounts. Depending on the numbers that would further distort the figures.
15,000 paid subscribers. You might be talking as much as 10% of the remaining subscriber base, i.e. about 1% of the 90% reduction figure.
They are saving on servers, bandwidth, etc. so who knows, it may even out.
Yeah. Best guess is they're losing thousands per day in ad revenue. A hosting plan that can handle the load (1.2m unique visitors per day -- we're not talking something on the scale of facebook here) should cost less than a tenth that.
the only users left are the ones that specifically want to insulate themselves from the Beeb (in other words, find comfort in the opinions of Murdoch/the Times and discomfort from opposing viewpoints).
Which, to be honest, I find 15,000 a remarkably low figure for. Of course, there's still the Daily Mail offering its news for free, and it would be interesting to see how the figures shifted if the Mail also went paywall.
It would be a shame, though, not to have a free news source to take the piss out of each morning.
Imagine how the advertisers feel. It's like 90% of the former Times web readership installed adblock overnight.
It's an interesting question. Some quick arithmetic based on the figures quoted suggests they're losing at least 850,000 page views per day (probably somewhat higher), but earning around £4,500 per day in subscriptions. Now, interestingly, at a rate of £5 per thousand views (not an unrealistic figure for ad revenues, certainly), this comes out as roughly break-even.
Of course, it is likely that they're losing more page views than this -- the figures were for unique visitors, and I suspect the average visitor is looking at a lot more than one page, so it seems unlikely they're actually reaching break-even, except that the advertising is now more targetted -- every viewer is proven to be somebody who is likely to spend money. They may actually be able to raise their charges for advertising space and, perhaps with just a few more subscribers than they have at the moment, or maybe by increasing the charge just a little, might reach break-even on the project.
Of course, if they were making significantly more than £5 per thousand, or if the number of page views per day was significantly higher than 1.2 million, then this breaks down. But it's interesting to note that you can view the figures in a way that shows this entire experiment not to be completely crazy.
This is the NYT; [...] Would anyone PAY for this leftist crap?
Erm, no, this is the Times, a somewhat-right-wing paper published in England. You might have heard of it?
The game app doesn't even have to work! It will get released on schedule none the less.
Hell, at least it means you'll always get at least one customer...
Sure there is: make software so crappy that nobody wants to pirate it.
Don't think it works. I found a Vista torrent the other day. There's even a Windows ME one, too!
What was the breach event? I thought it was selling 20% to Microsoft.
Don't think so. I'd say it's not transferring it when delivery became late.
Which is exactly the guys point: when you're spending that much money on a game, you can't afford to be creative, because if your new idea tanks you're left with a huge bill for developing it. Much safer to make something like another game that was previously popular.
Spend a tenth as much developing a game, and you can afford to take ten times as much risk, and maybe you'll get a runaway hit. Probably not, but it doesn't really matter that much...
Assume that we use plastic, and that some small percentage (no matter how much overall mass that is) will end up somewhere dangerous. Which is best/least bad: Plastics that don't degrade, or plastics that do degrade?
I think the answer is quite simple: if discarded on land, degradable is best. If discarded at sea, nondegradable is best. One possibility is to switch to nondegradables in coastal areas, but a more interesting one would be to develop a plastic that doesn't degrade in the presence of salt (or any other chemical present in the oceans but not commonly present in large quantities on land).
4,000,000,000 kg / 10,000,000,000 m^2 = 0.4 kg/m^2
Anyone else have a problem with this?
Indeed. The idea is total bullshit. 400gsm plastic is basically a medium thickness flexible sheet, something like a laminator pouch or similar. You can't build an island out of that shit.
Their argument appears to be that ISPs don't currently think they should pay for copyright infringement... so to prevent this being a problem ISPs should be made to pay for copyright infringement, at which point their objections will go away. Does that actually make sense in any way?
I predict a novel legal theory being invented that will protect the big corporation from a seemingly straightforward contract.
You mean like the well-known fact you have to enforce a contract within 7 years of it being breached, and that this breach apparently occurred 7 years and 2 months before this case was brought to court?
Parasites in shrimp can cause them to travel toward light and swim against gravity. The parasites act as a serotonin modulator. One particular antidepressant Fluoxetine does the same thing
If it's a serotonin-moderated behaviour, and fluoxetine triggers it, I'd be pretty happy wagering that most other SSRIs trigger it too.
So we need to dump a littler over 24 million metric tons of Fluoxetine into the gulf to see this concentration?
A back-of-the-envelope calculation based on a figure of 120 million people in the US prescribed antidepressants suggests annual consumption of SSRIs (by far the most prevalent kind) is in the region of 2 million tonnes, so it isn't unreasonably to believe that this kind of concentration could eventually occur (on the assumption that these chemicals do not readily break down in the environment).
Just call the RIAA and tell them that IP downloaded a song. They seem to be able to do all the John Doe stuff through the courts to find out who it was...
Actually, you can do that stuff yourself. File a claim with the courts for recovery of your possession, send a subpoena to the ISP, get the address, then either serve papers to continue the possession claim or hand the address to the state police.
DAB receivers in cars are not that common, and DAB reception in a moving vehicle tends to be pretty poor even if they are.
Which is odd, as one of the design goals for DAB was that the receiver should be able to move freely throughout the broadcast area and always get a signal, automatically switching to the strongest available transmitter whenever it changes.
It couldn't be something to do with the fact that the cheapest DAB radio I can find right now is £35 (£60 if you want something portable), whereas you can get a portable FM receiver for under £5? Nah, it must be to do with the regulations and standards!
Don't think it is, no. Where do most people listen to radio? In their car. It probably has more to do with the fact that DAB radios haven't been standard equipment in most cars until only about a year ago, and not even an option in many of them until not long before that. Most car manufacturers seem to have waited until the digital switchover was announced (currently scheduled for 2015, I believe) before offering them.
Dunno about the US one, but the objection UK telemarketers usually have to it is that it is kind-of expensive. My company used to do telemarketing as follow-ons from our (business-to-business) mailshots, but access to the TPS database costs about £5,000 per annum, which was hard to justify for us: our turnover was only about £30,000 at the time so it would have become our single largest expense and thus sunk a huge chunk of our profits. We therefore had to stop doing marketing calls.
That argument is getting old, too. I'm older than our Prime Minister and Deputy Prime Minister, and have been working with computers since my youth and have been online (via bulletin boards) since my college days. These things go back longer than the young guns realise, and if politicians don't understand them then there has to be another reason.
There are plenty of politicians who do understand. The only MP I've met on a personal level used to run an ecommerce company, and writes web apps in JSP in his spare time.