Heh... I would say someone using a supercomputer cluster in an RF-hostile environment for now. I'm not wholly sure where Light Peak's supposed to take things outside of that, though. They're working on 40Gbit and 100Gbit interconnect for clusters, etc. right now and 10Gbit is in ATCA blade server cages right now as the fabric interconnect. Perhaps there's higher signalling rates more readily possible than with copper on this- or perhaps there's less of a distance problem with it like there is with 10G Ethernet.
Heh... Since the general user population on Windows will install damned near any sort of crap on their box because it's "nifty", "cool", etc. along with at least prior versions of IE gleefully doing it for them whether they wanted it or not- why would it be any different now?
Heh... You'll note I didn't mention what type we were ripping out in my post- it was SOFTWARE RAID1 we were ripping out. It has more inherent problems than RAID5 and RAID5 in software has it's own sets of issues. Intrinsically, "hardware" RAID is little more than software RAID on specialized hardware- but it only has to do one thing and it's typically battery backed up enough to ensure pending writes to the array get there on a power failure- if you're doing enterprise class hardware that is.
When someone mentions RAID of any kind, I question whether they understand what they're talking about. There are good reasons for RAID- but many of the people using it aren't in that domain of good reasons more often than not.
The concern in question is a legitimate one, but not one that is actually and honestly covered by the DMCA or any of the other laws on the books that I am aware of. If TI's going to be worried about that concern, they need to go about it in a completely differing manner than this one. It's not really going to fly and it's earning them enemies in the process.
Moreover, I question the need for a calc at an SAT or ACT test in the first place. My (admittedly ancient...) recollection of the sorts of questions on the test were not calculation but rather comprehension and mathematical reasoning problems. They were of the sort that once you followed through the word problem (which is what a good half or more was...) you could do the problem on a scratch pad of paper in no time flat.
Unfortunately, in the past, they've been given for any willy-nilly thing instead of handing it down for egregious conduct. I know about egregious conduct- I'm experiencing it right now in a matter that I can't discuss for legal reasons.
Fortunately or unfortunately, depending on your viewpoint, there's a cap on just how much punitive damages you can get in most of the states. Texas' is three quarters of a million after computing 2.5 times the economic damages. It's similar in other states.
So, when you say "in most sensible countries, punitive damages don't exist", it implies you know little about how it all actually works. When someone sues someone else, it's mainly for economic or actual and potential (believable potential) harm. Now, since someone can file any stupid civil cause they want to (See SCO v. IBM...) we have at least a few people out there filing all sorts of actions that waste money, court time, etc. to see if they can extort money or score big on dumb blind luck in the courtroom. Except for rare cases, there is no pursuit in punishing barritry (the promulgation of a nonexistent case...) or for penalties being brought against a party that honestly believed they had a case and didn't because they didn't do all their work. In most sensible countries, you should have penalties for bringing a case of this sort to court- but there isn't so you see "sue em" happening all the time for things that shouldn't have ever been brought to court.
RAID5 is not backup. It's resilience for bringing the whole system down with a failure.
RAID was originally developed to make what we consider small storage capacities (then massive) affordable and reasonably reliable.
You're using RAID5 in it's "intended" use- but an SSD of the same capacity will be inherently MORE reliable (by a factor of how many of those magnetic disks you remove) than your system design right now.
From personal experience with a system customer base of literally thousands of enterprise class servers spread out over many companies, RAID doesn't work QUITE the way people make it out to be. We're ripping it out of the equipment and reverting to warm backups instead- the RAID1 design they fielded made the servers unstable.
The field engineer crowd (one of my friends worked with Nortel in the field engineer group and my brother is a manager for outsource company doing a lot of the same work with the same customers...) HATES RAID.
Blow a controller? Better hope you have an identical one in stock. You can't just swap out a differing controller of the same brand or pop a different brand in- they all do things ever so slightly differently on the disks.
Blow a disk? Better hope you can get the new drive in there and integrate it properly before you lose another.
Disks don't have the reliability we once thought they had. RAID doesn't do what most people thinks it does for them.
In truth, many of them are things like a resistive digitizer pad with a cardboard overlay and a PIC or AVR replaying a speech track. Economies of scale don't QUITE explain the prices of the devices in many of these cases. Because it's got "medical device" stamped on it, they charge quite a bit more.
You'd have to meet a certain level of proof for said disability for them to provide the expensive stuff- so why would it be any different for the "fancy PDA's and Smartphones".
I've been involved with discussions on designing one of those special purpose devices on the cheap with the design and the software being open-sourced in the past. Touch-boards are heinously limited in their vocabulary and grotesquely overpriced. Take a touch-screen netbook type device and put a customized Linux distribution on it and now you have the same thing, but unlimited in vocabulary and roughly 1/3rd the price.
These "special purpose" devices are waaay overpriced in many cases.
If that were the case, why is it that you don't see more e-Books, etc. online.
Your reasoning is flawed because you presume that the lack of DRM is the bar to bringing it online like they keep repeating.
That's not the bar, as Apple's and PayPlay and a few others have shown. DRM's not going to make the content there and cheaper. If anything it'll be priced at what the providers think the market will bear. DRM, in essence, is not about preventing "piracy"- it's about control and making you pay as often as they think they can make you do it and telling you how and where you'll play their stuff.
To be honest with you, I'd rather do without either the DRM or their stupid content if they're going to be that way about it- and they are.
Definitely. And it did it quite well, actually. What makes it more impressive is the nature of the design- they're not using the same class of tech as NASA and other space agencies have used in the past. It's capable of the same, but it's cheaper and intrinsically more robust- dual systems, etc.
Actually, the winds are minimal today in this area. I wouldn't know whether they're up over in Caddo Mills, but it's been a slow, steady, soaking rain for the last two days in the area with no appreciable winds (Caddo Mills is roughly 20 or so minutes to the North and East of Dallas along I-30 in-between Rockwall and Greenville.).
Low or no winds would be close enough to the conditions for a test for the purposes of the prizes.
The pricing in question was due to the structure of the royalty deals. It really is structured the way I described and it's a good part of why the prices are the way they are on those titles from LGP, Aspyr, etc. Unless you can afford a 25k units production run and have most of it sell, some of those prices for getting the port done up the price badly. Some of what killed Loki Games was that they'd tried to get a run like that at release for Quake3:Arena, blew the release window by all of about 3-4 weeks, and only sold like 200 units because everyone was buying the Windows SKU and "patching" it to run on Linux. They ended up owing a quarter of a mil to iD in unpaid-for royalties on the run.
Most of those publishers can only really justify a 5k unit run because most of those fees, including the per-unit royalty are up-front costs incurred either at the start of development or when you cut a run of product.
Pricing it less in many cases would be a loss of any money in profits for LGP and possibly none for the developers working on it. I'll admit that some of "poor sales" is the result of the situation in question- but a substantive part is more people claiming "predatory pricing", etc. and not buying, using the bargain bin prices for the Windows titles as a reasoning for things.
So you don't want to pay the price? Fine. Get the industry to work towards less predatory licensing practices- or to do Linux versions of titles. You won't, however, get them to sign on board without showing them real sales figures. So, you have to buy something that shows up as commercially viable- even if it's "preadtorially priced". You can discount that all you want, but unless you're talking Indies, that's how the whole lot works.
As for the flames needing a fire extinguisher... Heh... I strongly suspect you'd have needed something along those lines for the primary thrusters on the Moon mission or the Space Shuttle rockets. Just because you cut off fuel flow, doesn't mean you don't have fuel still burning in the combustion chamber- you're just not supplying it any more when you do that. It's not like a jet engine or a car engine in several ways. And, I suspect that they put the fire out because of safety concerns rather than needing to. It'd burned for probably about another minute or so and then ran out of fuel in the chamber.
With the way the thrust cone wandered all over the place as they were landing, showing them compensating for things as they brought it down- I'm suspecting a bit of Gyro stabilization coupled with thrust vectoring for directional changes.
It's the first one to accomplish the ground rules correctly that wins the purse in question. This means there's a second prize shot at Level 1 and the two Level 1 ones were up for grabs. Now only the Level 1 and Level 2 second prizes are free. They did, in fact, win first place, being first at the Level 2 pass of things. It's only not over in the sense of all the purses not being won.
Perhaps. Depends on how many nodes he had to set up.
Let's do a bit of "napkin math" on this:
I believe there's 48 unified shader cores in the 360's GPU. That's a nice amount.
There's 112 shader cores in the 9800GT. With the SLI setup, that's 224 of them at your disposal to do GPGPU thread processing with.
Now...done right (meaning not going overboard on the CPU, etc...), you can field a machine for about $600 or so that has an inexpensive SLI board, case, memory, etc. If you're doing a cluster node, you wouldn't need a disk, etc. so you could shave a bit more than you'd think off the price past the first machine bought.
$200 versus $600. The price is compelling. But, unfortunately, you're talking about a machine that's nearly 5 times more powerful (Possibly more, I'm not doing apples-to-apples comparisons on the shader cores...) at this sort of task with the PC- for only about 3 times the cost. To gain the same performance level, you would have to field 5 360's per each PC compute node. If you only need the power of two or three of the 360 nodes, then it makes some sense to do it with that, especially if you're familiar with the environment (the gent we're talking about in the threads here was that...).The power consumption will be comparable across the board, so that's not so much a consideration.
Where it really hits the wall is with the cluster fabric itself. Using PS3's and 360's is "cool" but it's actually not overly practical past about 10 or so machines for most performance computing applications because of the limitation of the cluster interconnect you have at your disposal. With those machines you will be limited to 1Gb Ethernet which limits your interconnect performance to about 750Mbits per node. When you go to match the performance of the PC box, you will find that you can do it, but it'll take 5 or so 360's to do it because of the overhead, lower performing hardware, and all. You'll have difficulty matching a cluster of the same numbers of PC's- and we won't get into using Myrinet, Infiniband, or iWarp channel adapters for 10Gb interconnects on the PC's which will make it be basically a huge SMP machine for all intents and purposes until you scale it to about 32 or so machines.
I think the assessment that it's familiarity and "cool" factor that drove this decision- not price or actual usefulness.
Heh.. Brave soul, you are, Jeff, for posting in the discussion threads on this. I have to commend you on that much.:-D
However, Spiderweb's been in business for a LONG time and you really didn't have much DRM in the earlier versions of your games. I can assure you that piracy was nearly as rampant then as it is now, it was just lower-key than it is nowadays. Even of your older shareware titles.
If you're making similar money now as on the old ones with your pricing, piracy's not your problem. Seriously.
If you're making more, it still didn't fix things to put DRM in place. And CD/Registration keys don't really count as DRM so much, especially if you're not online authenticating them.
Heh... The guy's been around quite a while in the Shareware space. I didn't think his stuff was worth what he was asking for it then. His Avernum series has gotten to the level where it might be worth $10-15 per game, but it's certainly not worth what he's charging now for it- and he's been that way for a long time now.
Slashdot Mirror
Heh... I would say someone using a supercomputer cluster in an RF-hostile environment for now. I'm not wholly sure where Light Peak's supposed to take things outside of that, though. They're working on 40Gbit and 100Gbit interconnect for clusters, etc. right now and 10Gbit is in ATCA blade server cages right now as the fabric interconnect. Perhaps there's higher signalling rates more readily possible than with copper on this- or perhaps there's less of a distance problem with it like there is with 10G Ethernet.
It's called "sarcasm"...
Humor: (Noun)
1. a comic, absurd, or incongruous quality causing amusement: the humor of a situation.
2. the faculty of perceiving what is amusing or comical: He is completely without humor. (Something you seem to lack yourself...)
Yep...I'm betting that they realize this, but are hoping the unwashed masses won't twig onto what they just said there.
Heh... Since the general user population on Windows will install damned near any sort of crap on their box because it's "nifty", "cool", etc. along with at least prior versions of IE gleefully doing it for them whether they wanted it or not- why would it be any different now?
Heh... You'll note I didn't mention what type we were ripping out in my post- it was SOFTWARE RAID1 we were ripping out. It has more inherent problems than RAID5 and RAID5 in software has it's own sets of issues. Intrinsically, "hardware" RAID is little more than software RAID on specialized hardware- but it only has to do one thing and it's typically battery backed up enough to ensure pending writes to the array get there on a power failure- if you're doing enterprise class hardware that is.
When someone mentions RAID of any kind, I question whether they understand what they're talking about. There are good reasons for RAID- but many of the people using it aren't in that domain of good reasons more often than not.
The concern in question is a legitimate one, but not one that is actually and honestly covered by the DMCA or any of the other laws on the books that I am aware of. If TI's going to be worried about that concern, they need to go about it in a completely differing manner than this one. It's not really going to fly and it's earning them enemies in the process.
Moreover, I question the need for a calc at an SAT or ACT test in the first place. My (admittedly ancient...) recollection of the sorts of questions on the test were not calculation but rather comprehension and mathematical reasoning problems. They were of the sort that once you followed through the word problem (which is what a good half or more was...) you could do the problem on a scratch pad of paper in no time flat.
Depends on your version of "sensible".
They exist to hammer home wrongs done.
Unfortunately, in the past, they've been given for any willy-nilly thing instead of handing it down for egregious conduct. I know about egregious conduct- I'm experiencing it right now in a matter that I can't discuss for legal reasons.
Fortunately or unfortunately, depending on your viewpoint, there's a cap on just how much punitive damages you can get in most of the states. Texas' is three quarters of a million after computing 2.5 times the economic damages. It's similar in other states.
So, when you say "in most sensible countries, punitive damages don't exist", it implies you know little about how it all actually works. When someone sues someone else, it's mainly for economic or actual and potential (believable potential) harm. Now, since someone can file any stupid civil cause they want to (See SCO v. IBM...) we have at least a few people out there filing all sorts of actions that waste money, court time, etc. to see if they can extort money or score big on dumb blind luck in the courtroom. Except for rare cases, there is no pursuit in punishing barritry (the promulgation of a nonexistent case...) or for penalties being brought against a party that honestly believed they had a case and didn't because they didn't do all their work. In most sensible countries, you should have penalties for bringing a case of this sort to court- but there isn't so you see "sue em" happening all the time for things that shouldn't have ever been brought to court.
RAID5 is not backup. It's resilience for bringing the whole system down with a failure.
RAID was originally developed to make what we consider small storage capacities (then massive) affordable and reasonably reliable.
You're using RAID5 in it's "intended" use- but an SSD of the same capacity will be inherently MORE reliable (by a factor of how many of those magnetic disks you remove) than your system design right now.
From personal experience with a system customer base of literally thousands of enterprise class servers spread out over many companies, RAID doesn't work QUITE the way people make it out to be. We're ripping it out of the equipment and reverting to warm backups instead- the RAID1 design they fielded made the servers unstable.
The field engineer crowd (one of my friends worked with Nortel in the field engineer group and my brother is a manager for outsource company doing a lot of the same work with the same customers...) HATES RAID.
Blow a controller? Better hope you have an identical one in stock. You can't just swap out a differing controller of the same brand or pop a different brand in- they all do things ever so slightly differently on the disks.
Blow a disk? Better hope you can get the new drive in there and integrate it properly before you lose another.
Disks don't have the reliability we once thought they had.
RAID doesn't do what most people thinks it does for them.
It's also easy to provide bullshit remarks to try to avoid answering a legitimate question.
It's a legit request- either answer the poster or spare us.
In truth, many of them are things like a resistive digitizer pad with a cardboard overlay and a PIC or AVR replaying a speech track. Economies of scale don't QUITE explain the prices of the devices in many of these cases. Because it's got "medical device" stamped on it, they charge quite a bit more.
You'd have to meet a certain level of proof for said disability for them to provide the expensive stuff- so why would it be any different for the "fancy PDA's and Smartphones".
I've been involved with discussions on designing one of those special purpose devices on the cheap with the design and the software being open-sourced in the past. Touch-boards are heinously limited in their vocabulary and grotesquely overpriced. Take a touch-screen netbook type device and put a customized Linux distribution on it and now you have the same thing, but unlimited in vocabulary and roughly 1/3rd the price.
These "special purpose" devices are waaay overpriced in many cases.
If that were the case, why is it that you don't see more e-Books, etc. online.
Your reasoning is flawed because you presume that the lack of DRM is the bar to bringing it online like they keep repeating.
That's not the bar, as Apple's and PayPlay and a few others have shown. DRM's not going to make the content there and cheaper. If anything it'll be priced at what the providers think the market will bear. DRM, in essence, is not about preventing "piracy"- it's about control and making you pay as often as they think they can make you do it and telling you how and where you'll play their stuff.
To be honest with you, I'd rather do without either the DRM or their stupid content if they're going to be that way about it- and they are.
Unfortunately, Vista's got BSoDs again.
It's not obsolete as you thought it was. You should be careful about making remarks about IE8 for the same reasons.
Definitely. And it did it quite well, actually. What makes it more impressive is the nature of the design- they're not using the same class of tech as NASA and other space agencies have used in the past. It's capable of the same, but it's cheaper and intrinsically more robust- dual systems, etc.
Actually, the winds are minimal today in this area. I wouldn't know whether they're up over in Caddo Mills, but it's been a slow, steady, soaking rain for the last two days in the area with no appreciable winds (Caddo Mills is roughly 20 or so minutes to the North and East of Dallas along I-30 in-between Rockwall and Greenville.).
Low or no winds would be close enough to the conditions for a test for the purposes of the prizes.
The pricing in question was due to the structure of the royalty deals. It really is structured the way I described and it's a good part of why the prices are the way they are on those titles from LGP, Aspyr, etc. Unless you can afford a 25k units production run and have most of it sell, some of those prices for getting the port done up the price badly. Some of what killed Loki Games was that they'd tried to get a run like that at release for Quake3:Arena, blew the release window by all of about 3-4 weeks, and only sold like 200 units because everyone was buying the Windows SKU and "patching" it to run on Linux. They ended up owing a quarter of a mil to iD in unpaid-for royalties on the run.
Most of those publishers can only really justify a 5k unit run because most of those fees, including the per-unit royalty are up-front costs incurred either at the start of development or when you cut a run of product.
Pricing it less in many cases would be a loss of any money in profits for LGP and possibly none for the developers working on it. I'll admit that some of "poor sales" is the result of the situation in question- but a substantive part is more people claiming "predatory pricing", etc. and not buying, using the bargain bin prices for the Windows titles as a reasoning for things.
So you don't want to pay the price? Fine. Get the industry to work towards less predatory licensing practices- or to do Linux versions of titles. You won't, however, get them to sign on board without showing them real sales figures. So, you have to buy something that shows up as commercially viable- even if it's "preadtorially priced". You can discount that all you want, but unless you're talking Indies, that's how the whole lot works.
As for the flames needing a fire extinguisher... Heh... I strongly suspect you'd have needed something along those lines for the primary thrusters on the Moon mission or the Space Shuttle rockets. Just because you cut off fuel flow, doesn't mean you don't have fuel still burning in the combustion chamber- you're just not supplying it any more when you do that. It's not like a jet engine or a car engine in several ways. And, I suspect that they put the fire out because of safety concerns rather than needing to. It'd burned for probably about another minute or so and then ran out of fuel in the chamber.
With the way the thrust cone wandered all over the place as they were landing, showing them compensating for things as they brought it down- I'm suspecting a bit of Gyro stabilization coupled with thrust vectoring for directional changes.
Actually... The rules are a bit interesting...
It's the first one to accomplish the ground rules correctly that wins the purse in question. This means there's a second prize shot at Level 1 and the two Level 1 ones were up for grabs. Now only the Level 1 and Level 2 second prizes are free. They did, in fact, win first place, being first at the Level 2 pass of things. It's only not over in the sense of all the purses not being won.
Perhaps. Depends on how many nodes he had to set up.
Let's do a bit of "napkin math" on this:
I believe there's 48 unified shader cores in the 360's GPU. That's a nice amount.
There's 112 shader cores in the 9800GT. With the SLI setup, that's 224 of them at your disposal to do GPGPU thread processing with.
Now...done right (meaning not going overboard on the CPU, etc...), you can field a machine for about $600 or so that has an inexpensive SLI board, case, memory, etc. If you're doing a cluster node, you wouldn't need a disk, etc. so you could shave a bit more than you'd think off the price past the first machine bought.
$200 versus $600. The price is compelling. But, unfortunately, you're talking about a machine that's nearly 5 times more powerful (Possibly more, I'm not doing apples-to-apples comparisons on the shader cores...) at this sort of task with the PC- for only about 3 times the cost. To gain the same performance level, you would have to field 5 360's per each PC compute node. If you only need the power of two or three of the 360 nodes, then it makes some sense to do it with that, especially if you're familiar with the environment (the gent we're talking about in the threads here was that...).The power consumption will be comparable across the board, so that's not so much a consideration.
Where it really hits the wall is with the cluster fabric itself. Using PS3's and 360's is "cool" but it's actually not overly practical past about 10 or so machines for most performance computing applications because of the limitation of the cluster interconnect you have at your disposal. With those machines you will be limited to 1Gb Ethernet which limits your interconnect performance to about 750Mbits per node. When you go to match the performance of the PC box, you will find that you can do it, but it'll take 5 or so 360's to do it because of the overhead, lower performing hardware, and all. You'll have difficulty matching a cluster of the same numbers of PC's- and we won't get into using Myrinet, Infiniband, or iWarp channel adapters for 10Gb interconnects on the PC's which will make it be basically a huge SMP machine for all intents and purposes until you scale it to about 32 or so machines.
I think the assessment that it's familiarity and "cool" factor that drove this decision- not price or actual usefulness.
Heh.. Brave soul, you are, Jeff, for posting in the discussion threads on this. I have to commend you on that much. :-D
However, Spiderweb's been in business for a LONG time and you really didn't have much DRM in the earlier versions of your games. I can assure you that piracy was nearly as rampant then as it is now, it was just lower-key than it is nowadays. Even of your older shareware titles.
If you're making similar money now as on the old ones with your pricing, piracy's not your problem. Seriously.
If you're making more, it still didn't fix things to put DRM in place. And CD/Registration keys don't really count as DRM so much, especially if you're not online authenticating them.
Definitely. You can expect something on the order of 3-ish per title being added in for paying the royalties to someone like SecuROM on this stuff.
While the Senator might have been right- what we did in response wasn't.
Heh... The guy's been around quite a while in the Shareware space. I didn't think his stuff was worth what he was asking for it then. His Avernum series has gotten to the level where it might be worth $10-15 per game, but it's certainly not worth what he's charging now for it- and he's been that way for a long time now.