Now there's finally a basis for development of proprietary closed-source derivatives of CVS.
GPL'd software sucks, because there's no way for Microsoft to lock consumers into proprietary derivatives.
you wouldn't want to build an exact copy.
Agreed. But I was just trying to refute the notion that building a second would be cheaper.
you might launch to an L2 orbit instead
AFAIK, that's exactly the plan for the Webb space telescope. But what will they do when its gyros need to be replaced?
but it's probably cheaper to build a new one from scratch
I don't buy it. A shuttle mission to the Hubble will cost about two billion. I don't think they can build another and launch it for that. Even though the technology has improved, most of the costs of designing, building, and launching have gone UP, not down.
Building a new space telescope is a good idea, and they're planning to do it anyhow. But if they're going to spend another two billion dollars on the Hubble, they're better off with a manned mission with a better than 75% chance of success (and perhaps 2% chance of catastrophic mission failure) than a two billion dollar robotic mission with a sub-10% chance of success. Of course, it's easy for me to say that since I won't be risking my life on it, but I suspect that they can find astronauts willing to volunteer for a manned repair mission.
That might be true if NASA built exactly the same thing again. But they generally don't.
And even if they tried to build exactly the same item again twenty years later, much of the knowledge of how to build it is lost. Unfortunately not all the important knowledge gained from the designing and building a complex system gets recorded in a form useful for posterity. And twenty years later, many of the people that worked on the first one and have the necessary knowledge and skills are long since moved on to greener pastures. So building it again involves a lot more research than you might think.
I doubt that they even have all of the correct engineering drawings for the Hubble. A report last year indicated that NASA doesn't even have all of the correct engineering drawings for ISS!
When they built the fifth Space Shuttle, even though they used a lot of leftover parts from the construction of the first four, and didn't need to do any research, it still cost more than any of the first four did.
30% more expensive, after 15 years of R&D, than when invented by NASA? Of course not.
You're assuming that things get less expensive over time. That's only true of things that are mass-produced. If you build one of something, then twenty years later you want to build a second one, you definitely should expect the second to cost more.
For the most part Dell can't dictate Server OSes, the customer choses which one they want from a dropdown or just gets it bare.
I have no idea what your point is. Dell expects to sell some number of systems with RHEL installed, so they can negotiate on the basis of that number. That doesn't mean that any given Dell customer is forced to buy any particular OS.
(Dell also must ship way more than 5K units of RedHat a year.)
Sure, I was just picking 5K as an example. If they bought 50K licenses from Red Hat, they should be able to get an even better deal.
It probably wouldn't be 5000 different customers. And anyhow, I didn't say they'd give it away for free. I just said they'd offer a discount. I don't know of too many things that you don't get a discount when you buy 5000.
If they got a guaranteed sale of 5000 units, it probably would be worthwhile for them to offer a fairly steep discount.
And you probably thought that today's mass transit has too much vandalism and graffiti. Just wait until the losers who don't have anything better to do than damage or deface public property get their own private car for each trip.
My theory of graffiti is that it's from people who have low self esteem and don't think that they can leave a mark on the world in any but the most literal sense.
I agree. Between 1999 and 2002 at Pluris (now out of business), we routinely used 24-layer and 26-layer boards that were about 24x36 inches. These boards were packed very densely with BGAs and other fine-pitch surface mount parts, so the trace geometry was also very fine. This was pushing the limits of PCB fabrication; some PCB houses that claimed they could do it in fact failed to produce a single good board for us. Even the PCB houses that could do it did not get very good yields. Every bare board underwent a 100% electrical test before we'd even think about sending them to the assembly house.
I've still got a bare board in my office as a souvenir.
Seiko Epson may well have come up with a new technique, but it's certainly not the world's first twenty layer board.
"Gate equivalents", as I understand them, are the number of logic gates that you could get if you used every single resource on the FPGA (which is impossible of course)
The gate counts they specify are for fairly high utilization, but not 100%. Also, it is possible to get 100% utilization, if you have a highly regular design and use manual floorplanning. In this case the "gate count" output in the report from the Xilinx ISE software can be substantially higher than the rated gate count of the part.
Anyhow, if you were building graphics cards, you'd be buying newer FPGAs like the Spartan 3, which gets you a lot more bang for the buck than the older Spartan II and IIe families. Digikey is slow to get new Xilinx parts, but Spartan 3 parts are available from other Xilinx distributors.
I don't know much about FPGAs -- what will $30 get you?
In quantity, that will buy you a XilinxSpartan 3 XC3S1500, which is claimed to be equivalent to 1.5 million gates. In more concrete terms, it contains over 26,000 logic elements each of which has a flip-flop and a four-LUT (programmable look-up-table that can generate an arbitrary function of four boolean inputs). It also has 32 parallel multipliers (18-bit w/ 36-bit product) and 32 BlockRAMs of 18 Kbits each.
That together with some DDR SDRAM and suitable video DACs should be sufficient for a fairly decently performing 2D card.
No, the "line-of-sight" from the phone to my testicles is through more than six inches of flesh and bone. Not much 1.9 GHz RF is going to get through that.
Digital Equipment Corporation (the remains of which are now part of HP) had an online customer ordering system in the early 1990s that did everything described in claim 1 of the patent.
I'm pretty sure there were other systems in operation more than a year before the patent filing that did this as well.
I keep my GSM phone in my pocket. Bluetooth uses much lower transmit power than the phone. I suppose I have less risk of acoustic neuroma than if I had the phone to my ear, but maybe more risk of something else near my thigh.
It will run any operating system and any software, but it may not be Trusted
You are correct that this is the currently stated position of Microsoft and their partners. However, I think it would be extremely naive to believe that Microsoft will not provide substantial incentives to hardware manufacturers to build machines that will not run unapproved operating systems at all.
Traditionally Microsoft only was willing to offer high-volume OEM discounts based on the total computer sales, even for machines not shipped with Windows. They got in trouble for that, so they're probably not doing it any longer. However, they may well be able to offer the OEM a better discount if the hardware and BIOS will only run Windows. We're not there yet, but it's just around the corner.
Richard Stallman refers to it not as "Trusted Computing", but as "Treacherous Computing". The phrase "Trusted Computing" was deliberately chosen by the TCPA because sounds like a wonderful thing. Everyone wants to trust their computer. And trusted computing does provide a little of that. But what it really does is ensure that other people trust your computer. Specifically, that other people trust it not to do what you want it to, but only what they are willing to allow.
The story said:
trusted chips will eventually be used by software manufacturers to make sure the computer's owner does not do anything with the software which the manufacturer does not want to permit.
It should be noted that what we're really talking about is preventing the computer's owner from doing things that Microsoft and their allies (such as the MPAA) don't want to permit.
The computer manufacturer, such as IBM, is largely irrelevant, except to the extent that they may eventually offer hardware that will refuse to run operating systems they don't approve of. Since IBM supports Linux, it doesn't seem likely that they will build machines that can't run Linux, but many other vendors have hitched their wagons more firmly to Microsoft.
Within minutes of posting this request, an employee of the company let me know that there will be an SDK for software developers, probably available within the next week. He also said that the device is easy to control using HID and Audio class drivers, so that a specific Linux driver is probably not necessary. Just an application. I'll order a RadioSHARK and try to write one.
Um, what exactly is wrong with asking about Linux drivers? It seems like a perfectly reasonable request. I was not in any way criticizing the product or the company.
For a fictional example of serious problems resulting from exactly this sort of speech recognition ambiguity, read the book "Sewer, Gas and Electric" by Matt Ruff.
Slashdot Mirror
Now there's finally a basis for development of proprietary closed-source derivatives of CVS. GPL'd software sucks, because there's no way for Microsoft to lock consumers into proprietary derivatives.
Sorry, that was an incredibly dumb movie, but I just couldn't resist.
Building a new space telescope is a good idea, and they're planning to do it anyhow. But if they're going to spend another two billion dollars on the Hubble, they're better off with a manned mission with a better than 75% chance of success (and perhaps 2% chance of catastrophic mission failure) than a two billion dollar robotic mission with a sub-10% chance of success. Of course, it's easy for me to say that since I won't be risking my life on it, but I suspect that they can find astronauts willing to volunteer for a manned repair mission.
And even if they tried to build exactly the same item again twenty years later, much of the knowledge of how to build it is lost. Unfortunately not all the important knowledge gained from the designing and building a complex system gets recorded in a form useful for posterity. And twenty years later, many of the people that worked on the first one and have the necessary knowledge and skills are long since moved on to greener pastures. So building it again involves a lot more research than you might think.
I doubt that they even have all of the correct engineering drawings for the Hubble. A report last year indicated that NASA doesn't even have all of the correct engineering drawings for ISS!
When they built the fifth Space Shuttle, even though they used a lot of leftover parts from the construction of the first four, and didn't need to do any research, it still cost more than any of the first four did.
If they got a guaranteed sale of 5000 units, it probably would be worthwhile for them to offer a fairly steep discount.
But otherwise their statement is just so much empty posturing, not unlike how Microsoft says that hardware should be free (as in free beer).
I'm more concerned about the FBR, CEA, and especially TPC.
My theory of graffiti is that it's from people who have low self esteem and don't think that they can leave a mark on the world in any but the most literal sense.
I've still got a bare board in my office as a souvenir.
Seiko Epson may well have come up with a new technique, but it's certainly not the world's first twenty layer board.
Anyhow, if you were building graphics cards, you'd be buying newer FPGAs like the Spartan 3, which gets you a lot more bang for the buck than the older Spartan II and IIe families. Digikey is slow to get new Xilinx parts, but Spartan 3 parts are available from other Xilinx distributors.
That together with some DDR SDRAM and suitable video DACs should be sufficient for a fairly decently performing 2D card.
... you insensitive clod!
No, the "line-of-sight" from the phone to my testicles is through more than six inches of flesh and bone. Not much 1.9 GHz RF is going to get through that.
Digital Equipment Corporation (the remains of which are now part of HP) had an online customer ordering system in the early 1990s that did everything described in claim 1 of the patent. I'm pretty sure there were other systems in operation more than a year before the patent filing that did this as well.
I keep my GSM phone in my pocket. Bluetooth uses much lower transmit power than the phone. I suppose I have less risk of acoustic neuroma than if I had the phone to my ear, but maybe more risk of something else near my thigh.
Traditionally Microsoft only was willing to offer high-volume OEM discounts based on the total computer sales, even for machines not shipped with Windows. They got in trouble for that, so they're probably not doing it any longer. However, they may well be able to offer the OEM a better discount if the hardware and BIOS will only run Windows. We're not there yet, but it's just around the corner.
The story said:
It should be noted that what we're really talking about is preventing the computer's owner from doing things that Microsoft and their allies (such as the MPAA) don't want to permit.The computer manufacturer, such as IBM, is largely irrelevant, except to the extent that they may eventually offer hardware that will refuse to run operating systems they don't approve of. Since IBM supports Linux, it doesn't seem likely that they will build machines that can't run Linux, but many other vendors have hitched their wagons more firmly to Microsoft.
Within minutes of posting this request, an employee of the company let me know that there will be an SDK for software developers, probably available within the next week. He also said that the device is easy to control using HID and Audio class drivers, so that a specific Linux driver is probably not necessary. Just an application. I'll order a RadioSHARK and try to write one.
Um, what exactly is wrong with asking about Linux drivers? It seems like a perfectly reasonable request. I was not in any way criticizing the product or the company.
Are there Linux drivers? If not, is the company willing to provide a spec for the USB commands to control it, so that a Linux driver can be written?
For a fictional example of serious problems resulting from exactly this sort of speech recognition ambiguity, read the book "Sewer, Gas and Electric" by Matt Ruff.
Has anyone noticed that the OpenXM project has vanished from SourceForge? Did XM or the RIAA send SourceForge a DMCA notice?