No state is maintained from one ftp connection to the next.
No state is maintained from one telephone call to the next. No state is maintained from one quake session to the next.
The difference with HTTP is that a user "session" (browsing your web site) encompasses multiple HTTP requests. Each request stands alone and involves a single query, single response - it's not interactive. (OK I'm simplifying, I know you can do more with HTTP 1.1.) Same with NFS - the NFS 'mount' call just sets up state in the client - the server does not maintain any state between calls. Every NFS call stands alone - the server does not even maintain open files on behalf of the client.
Various mechanisms are used to get around the lack of server state - HTTP cookies, magic session URLs, HTTP auth. In the NFS world, the lock manager uses a separate protocol from the NFS transaction protocol - because locking must maintain state on the server - that being the whole point. (That is why "lockd" is usually a separate process from "nfsd" - RPC mechanisms such as ONC make life much easier if you stick to one daemon, one protocol.)
By contrast, FTP, voice-over-PBX and Quake are not stateless. A single "connection" (the FTP port 21 control channel, the phone call, the Quake data channel) corresponds to the user experience of a single "session". Note that in the case of FTP at least, some clients implement "auto-reconnect" logic in case the server disconnects - this is to maintain the illusion of an unbroken session. Such is not needed with HTTP, since the server isn't maintaining a connection across user requests anyway.
ObScore+1Interesing: SCSI was intended to be pronounced "sexy". I don't know why "scuzzy" caught on instead. Probably for the same reason "BEEP" won't take off. It tends to defy any serious discussion.
Wasn't the business version of 3.1 caleed "Windows For Workgroups"?
You have confused Windows 3.1 with Windows NT 3.1. Windows 3.1 was an upgrade from Windows 3.0 and ran on top of MS-DOS 5 or 6. Windows NT 3.1 was the first (I think) public release of a brand-new operating system inspired by VMS and spun off from OS/2 development. They called it "3.1" probably because its user interface was intended to resemble that of Windows 3.1. (You know, just like Windows NT 4.0 was intended to resemble Windows 4.0 aka Chicago aka Windows 95.)
Windows for Workgroups was a business-oriented version of Windows 3.1, later revved to 3.11 (I've never actually seen WfW 3.1 but I've heard that it existed). I don't know the differences, exactly, except that WfW had more support for Novell and NetBIOS networking.
I mean, XP (Win2000 derivative) crashes
because of problems XP has with ACPI and my CD-RW. Its definitely not my hardware, because my PC works 100% under Linux.
Uhhhhh... I distrust XP as much as the next guy, but don't be so quick to blame it. Do you have any idea how horribly complex of a mess ACPI is? Whoever defined that standard should be Rasputinised. Consider the reputation BIOS vendors have for bug-free code, and then think about how they have to write entire device drivers in a special ACPI-specific mini-language to be interpreted by the OS (you know, like Open Firmware, only different), and then you can begin to shake your head in wonder that your machine even boots.
Linux has ACPI support, but only skeletal until recent kernels, and I never enable it because I don't trust it (either the newness of the Linux implementation, or the BIOS vendors) and because Linux likes my hardware just fine without it. Does your Linux kernel use ACPI?
XP isn't extensible. What POSIX compatability?
Have you tried extending XP lately? How do you know it's so hard? (I haven't tried either, btw.) As for POSIX, well, NT4 had a POSIX layer nobody used because it was so buggy - I believe it existed solely to get DOD contracts which specified "must run on a POSIX OS" - which was a load of crock since the point of said contracts was to avoid platform lock-in, but of course they went ahead and bought "POSIX-compatible" NT and then used the Win32 subsystem instead of the useless POSIX layer, thus achieving vendor lock-in after all....
There have been third-party POSIX subsystems for NT, some of which are reported to work fine. Win32 is only a subsystem on top of the microkernel; I bet this is still true in XP. I don't know if the OS/2 subsystem is still around - if I were Microsoft I probably would have dropped it with Win2k. But that doesn't mean you couldn't write your own servers. There has been talk of porting UML to Windows - can't recall if they wanted to use Win32 or the bare NT ukernel, though.
I would say pasport does a good job verifying
the creator of the nick is the user of the nick, provided you supply a good password
Maybe they do, maybe they don't. It's a moot point, since people generally pick crappy passwords. If I were working for Microsoft and it were my job to handle the security architecture of Passport, I'd do it as carefully and painstakingly as possible... yet we know their track record on these things, and it's not reassuring. And unless they can enforce good security practices on the users (picking passwords, not sharing them with others, not writing them down in obvious places, etc) it doesn't really matter all that much.
What users do with secure information like their own passwords has always been a concern, but it is orders of magnitude worse when all the security information is centralised. Yet another reason to fear MS P....
but how does this keep illegal users from
creating a passport identity to accompany a credit card and use that identity for purchase verification?
Good question. Anyone who partners (I hate that verb) with Microsoft to provide services using Passport should take any information they get from said service with a major grain of salt.
Imagine if there were several competing, proprietary protocols
implementing the basic functions of http, ftp, or smtp. Imagine if IE could only get web pages served by IIS, and Mozilla could
only get web pages served by Apache. If Outlook users couldn't exchange email with Pine users or Eudora users.
It's worth noting that email used to be this way. And some people tried to solve the problem using the Trillian method - have a single email client that spoke cc:Mail, MS Mail, SMTP, etc. Indeed, Outlook is still commonly installed with a legacy plugin for Exchange Server. (Well, "legacy" for its email functionality - I know the Exchange plugin does much more than email.)
Nowadays everyone speaks [E]SMTP except certain corporate holdouts, and even those holdouts generally offer an SMTP gateway.
Another comparison: most of those proprietary email systems were centralised by design - there was the one email server, or one set of servers, and the clients. Just like IM today. I'm convinced that to truly unify IM would require decentralisation - the user@serverhost model which SMTP and I believe Jabber use. The IRC model doesn't really scale, though it has the advantage that people don't need to know each other's server names.
yes, but much like a serial code for your warezed games, etc, i'm sure someone could crack it, and come up with a fake signature generator for whatever software you want to boot onto the xbox, right?
It's one thing to create a key generator for something like FLEXlm (whose license keys are fairly short - possibly because a once-popular way to get a license was via FAX, which forced the customer to type in a random hex string - so the shorter the better) or DVD CSS (intended for embedded devices, so they needed to minimise the hardware cost for decoding, which must be done in real time). It's quite another thing to crack a protocol which is designed to run on a 700 MHz 32-bit CPU in a non-time-critical path (only during boot, as opposed to during all DVD playback).
In other words, unlike many cases, Microsoft wasn't forced to use short keys or fast algorithms. They could quite easily have used something which is not computationally feasible to crack within the next 50 years.
Although, given Microsoft's track record with regard to security implementations, one can always hope....
If you look at the knives in the ends of the RJ45, the ones that cut through the insulation, you'll see that for each conductor, it's a single knife that cuts through the middle of the wire.
Not the RJ45 plugs I buy. They're labeled "for solid" or "for solid or stranded" and they have three knife points in a _-_ pattern. They bend around the wire core and work great.
I've had a lot more trouble with patch cables I've made from stranded wire and "for stranded" plugs. I'm willing to admit the possibility that this is due to other factors - quality of the cable, or my relative inexperience at crimping - but in any case, once I switched to solid cable I haven't looked back.
what about extracting a working "signature" from ms from a legit game and comming up with a working facsimilie (sp) to fool the xbox into letting it run?
The signature most likely (i.e. "if Microsoft isn't a lot stupider
than they look") contains a cryptographic checksum. That would preclude
any attempts similar to what you describe. Cryptography discovered, and
solved, this problem perhaps 25 or 30 years ago.
but I don't recall the part of copyright law that
says you can raise prices indefinitely without consequence
And you were doing so well up until that point. What part of
copyright law do you recall? The part where the author /
artist of a work has an absolute monopoly on the right to
publish that work, allowing others to publish only small excerpts
without explicit permission?
Copyright law doesn't guarantee profits, but it does
guarantee protection from competition for a "limited" time. People
who break that law are, well, breaking the law.
It's very easy to dismiss artists that put society's collective taboos and psychoses on
display.
It's also very easy to dismiss artists who think they're being cool and/or shocking and/or tortured and/or neurotic but whose quirks are a lame, transparent stage show.
This is a variation on the Sagan quote about Bozo the Clown. Just because you're dismissed by society at large doesn't mean you have necessarily hit a sore spot society is afraid to face.
I have no particular opinion on Eminem; just thought I'd throw that out.
but Samsung accounts for better than 80% of
production, if memory serves.
Har har har.
My original point is that when I decide to buy a computer, I pick a platform, not a memory.
Good point, especially now that memory (RDRAM included!) is so cheap. But soon enough there will be good P4 motherboards that take DDR. Then you can pick a platform and a memory.
One minor quibble to this otherwise great explination
Thanks..
RDRAM actually uses
DDR technology for it's bus as well. This means that PC800 RDRAM actually runs on a 400MHz bus. End
result, 400MHz, 16-bits wide and DDR give you the 12,800Mb/s that you mentioned.
OK, +1 Informative, that's something I didn't know. I assume, since you didn't mention it, that I guessed correctly (yes it was a guess) about RDRAM interleaving access between two RIMMs to double the bandwidth.
FWIW, DDR is by no means specific to DDR SDRAM, or even memory in general.
Right - but I didn't know "DDR" was the generic term for sending signals across both clock edges. Thanks again - that makes +2 Informative (to me anyway).
Quality PC2100 is frequently marketed as PC2400. On www.pricewatch.com, the difference between PC800 and
PC2100 is $5 for 128MB.
Perhaps, but you're probably comparing single-stick to single-stick. With RDRAM you have to buy a matched pair. So the right comparison is 2x128 PC800 ($80) versus 1x256 PC2400 ($51).
Or go on up to 512MB. 2x256 PC800: $148. 1x512 PC2400: $114.
So RDRAM costs an additional 57% for 256MB, or 30% for 512MB. Nice that it's no longer double the cost, but to me that is still a significant markup. Anyone know approximately how much of that is due to
(a) economies of scale,
(b) manufacturing cost after accounting for (a), or (c) patent licenses?
When it comes to wrist pain and typing speed, money really is no object. At least, $200 is a pretty small price to pay.
Not that I'd get one, most likely - I'm Yet Another IBM Model M Fan. Not sure what I'll do when motherboards start coming without PS/2 ports. That'll suck. Oh, I know, I'll use a PS/2 port card... except that these motherboards won't have ISA slots either... aw, now I'm just getting depressed. Does anyone sell a USB-to-8042 bridge?
would someone be kind enough to
explain the differences between the different kinds of RAM mentioned in all the replies? PC2100, PC3200,
DDRxx, etc.? Just a quick primer would be great.
In the beginning there was PC100 SDRAM. Well, actually, that was mid-nineties, but that's about when most Slashkiddies were born, so moving on. Obviously everything is just a marketing label, but this one meant 100 MHz. With SDRAM, each Hz gives you 64 bits, so the bandwidth is 6400 megabits per second.
Thus PC133 and PC166 are 8500 and 10700 Mb/s.
DDR is the same tech as SDRAM, except that it uses a trick to transfer data twice per clock cycle, so you get 128 bits per Hz. Thus PC100 DDR-SDRAM would be 12800 Mb/s. But Marketing decided that was unfair, so they labeled DDR based on twice the clock speed, so we have PC266 and PC333, which of course run at 133 and 166 MHz and give you 17000 and 21000 Mb/s.
RDRAM is based on a new tech that gives you only 16 bits per clock cycle instead of 64 for SDRAM and 128 for DDR-SDRAM. The difference is that you can clock it way up. So there was PC600, PC700 and PC800 RDRAM, again based on MHz, so that gave you 9600, 11200, and 12800 Mb/s bandwidth. Basically you divide the number in four to compare with SDRAM speeds, since you only get 1/4 as many bits per cycle. Actually I believe modern Rambus controllers double this by interleaving two sticks, so now you divide by 2 - PC800 has four times the bandwidth of PC100, but requires a matched pair of sticks.
Then the DDR people decided to start talking direct bandwidth, rather than megahertz. But unlike me, they mean megabytes, rather than megabits, per second. PC1600 is DDR-SDRAM at 100 MHz, since DDR gives you 128 bits or 16 bytes per cycle. PC2100 is DDR at 133 MHz, formerly known as PC266. PC2700 is DDR at 166 MHz, and PC3200 is DDR at 200 MHz.
With interleaving, Rambus gives you 32 bits or 4 bytes per cycle. PC800 has the same bandwidth as PC3200 DDR, and the relatively new PC1066 has more - 4266 megabytes per second.
Bandwidth is a good baseline for comparison, but RDRAM has a higher latency than SDRAM or DDR-SDRAM. That's why DDR, with its lower maximum bandwidth, is still speed-competitive with RDRAM (for a lot less money).
Why can't they just do interleaving (call it stripping/RAID-0 for memory)? No need to crank up those
Mhz's, but spread the load over a couple of DIMM's. Most large systems (at least Sun I know off) still
use 100Mhz or so DIMM's but do 8-way interleaving (maybe even higher) to get their high memory
bandwidths.
We used to do that. SIMMs were 32 bits wide and the Pentium architecture wanted 64 bits at a time, so you had to pair them up. People found that annoying, so DIMMs have 64 bits of bandwidth. I think consumers would balk at having to buy DIMMs in sets of 2 or 4 or 8, though server class people don't seem to mind so much, since at the server end you don't really expect a commodity market for upgrades anyway.
You'll note that RDRAM does continue to use this trick. I don't know if individual RIMMs are only 8 bits wide or what, but you do have to have matched pairs. (It would surprise me if a RIMM were only 8 bits wide, since I've seen some with 16/18 chips.)
Perhaps the other answer to your question is that with SDRAM, you have 64 bits on each DIMM, so interleaving means your memory controller has to be able to handle 128 or 256 or 512 bits at a time, which is a lot of traces and a lot of logic and perhaps a lot of expense for the motherboard and specifically the controller / L2 cache.
There's certainly something to be said for proprietary memory technology. Sure, it's expensive, and
Rambus does all kinds of dishonest lawyer tricks with the patent system, but you probably won't find
that level of integration between the processor and the memory on a standards-based SDRAM system.
Beg to differ. The AMD Hammer will actually have a DDR memory controller on-chip. That's what I call integration between the processor and the memory. I don't see that lawyers and patents enter into this at all. I know it is currently vapor, but I'm very much looking forward to seeing the Hammer in action (and probably buying one).
It's my
understanding that if a patent is not defended by the holder, the protection may be ruled unenforceable
later.
If that were true we wouldn't have PNG. Unisys decided to start making threats about LZW compression only after GIFs became popular - i.e. very late in the game.
As noted elsewhere, you're thinking of trademarks (and probably, by extension, service marks).
It would have been better to call it the MX-EVA2 (Asuka's)
Ah yes, the one that went down into the volcano. (Well, Unit 01 did too.) I guess Unit 02 must have had a pretty decent cooling system - actually probably more than $500 worth.
Admitedly a week uptime is not alot but often there are new virus updates within the week so I have to reboot anyways.
What?!? You have to reboot for software updates? You must be running either HP/UX or something from Microsoft! (: Oh, you mentioned viruses - never heard of viruses for HP/UX, so....
(I still fondly remember the time, must have been 1997 or 1998, when someone on Usenet came up with an idea for how init could restart itself without losing state. The sysvinit maintainers promptly implemented it, thereby eliminating the last reason to have to reboot a Linux machine to update any software short of the kernel image itself.)
Re:Something interesting about Moz on Windows XP
on
Mozilla RC3 Released
·
· Score: 1
and after the.9, Mozilla become my master browser.
Would that be your local master browser, or your domain master browser? (:
That means pressing a lot of
different CDs simultaneously on different machines instead of cranking out the same one over and over
again. Semi-fixed costs, such as studio time, cover art expenses, music video production expenses, etc.
that become a smaller and smaller percentage of the cost of pressing each CD as the number of CDs
pressed goes up don't become a smaller and smaller percentage if you're pressing a lot of different CDs
instead of a lot of copies of the same CD. A million CDs from just one act means more profit than 10
acts each selling only one-hundred thousand CDs or one hundred different acts each selling only ten
thousand CDs.
I don't think it works out that way, because the label passes these costs on to the artist. That's why if you are a moderately successful artist, you go broke, because all your royalties go to paying back the label for all those expenses you just listed. I believe most record contracts work this way, with the label only lending money for up-front costs.
Somehow VH1 always forgets to mention this part on Behind the Music. They always talk about the hotshot band finally making it big time and living large, and then a couple years later going bankrupt because they "took some bad business advice" or "were too generous to their relatives" or "got careless and didn't keep track of their money". I figure most of the time it's really "got screwed by the label contract" and for some odd reason VH1 just doesn't want to put it that way.
Slashdot Mirror
D00d! An ASCII representation of Pascal strings?!? What will djb think of next? Hash tables? (:
The difference with HTTP is that a user "session" (browsing your web site) encompasses multiple HTTP requests. Each request stands alone and involves a single query, single response - it's not interactive. (OK I'm simplifying, I know you can do more with HTTP 1.1.) Same with NFS - the NFS 'mount' call just sets up state in the client - the server does not maintain any state between calls. Every NFS call stands alone - the server does not even maintain open files on behalf of the client.
Various mechanisms are used to get around the lack of server state - HTTP cookies, magic session URLs, HTTP auth. In the NFS world, the lock manager uses a separate protocol from the NFS transaction protocol - because locking must maintain state on the server - that being the whole point. (That is why "lockd" is usually a separate process from "nfsd" - RPC mechanisms such as ONC make life much easier if you stick to one daemon, one protocol.)
By contrast, FTP, voice-over-PBX and Quake are not stateless. A single "connection" (the FTP port 21 control channel, the phone call, the Quake data channel) corresponds to the user experience of a single "session". Note that in the case of FTP at least, some clients implement "auto-reconnect" logic in case the server disconnects - this is to maintain the illusion of an unbroken session. Such is not needed with HTTP, since the server isn't maintaining a connection across user requests anyway.
See the difference?
ObScore+1Interesing: SCSI was intended to be pronounced "sexy". I don't know why "scuzzy" caught on instead. Probably for the same reason "BEEP" won't take off. It tends to defy any serious discussion.
OK, you're wrong, and you're excused.
You have confused Windows 3.1 with Windows NT 3.1. Windows 3.1 was an upgrade from Windows 3.0 and ran on top of MS-DOS 5 or 6. Windows NT 3.1 was the first (I think) public release of a brand-new operating system inspired by VMS and spun off from OS/2 development. They called it "3.1" probably because its user interface was intended to resemble that of Windows 3.1. (You know, just like Windows NT 4.0 was intended to resemble Windows 4.0 aka Chicago aka Windows 95.)
Windows for Workgroups was a business-oriented version of Windows 3.1, later revved to 3.11 (I've never actually seen WfW 3.1 but I've heard that it existed). I don't know the differences, exactly, except that WfW had more support for Novell and NetBIOS networking.
Uhhhhh ... I distrust XP as much as the next guy, but don't be so quick to blame it. Do you have any idea how horribly complex of a mess ACPI is? Whoever defined that standard should be Rasputinised. Consider the reputation BIOS vendors have for bug-free code, and then think about how they have to write entire device drivers in a special ACPI-specific mini-language to be interpreted by the OS (you know, like Open Firmware, only different), and then you can begin to shake your head in wonder that your machine even boots.
Linux has ACPI support, but only skeletal until recent kernels, and I never enable it because I don't trust it (either the newness of the Linux implementation, or the BIOS vendors) and because Linux likes my hardware just fine without it. Does your Linux kernel use ACPI?
Have you tried extending XP lately? How do you know it's so hard? (I haven't tried either, btw.) As for POSIX, well, NT4 had a POSIX layer nobody used because it was so buggy - I believe it existed solely to get DOD contracts which specified "must run on a POSIX OS" - which was a load of crock since the point of said contracts was to avoid platform lock-in, but of course they went ahead and bought "POSIX-compatible" NT and then used the Win32 subsystem instead of the useless POSIX layer, thus achieving vendor lock-in after all....
There have been third-party POSIX subsystems for NT, some of which are reported to work fine. Win32 is only a subsystem on top of the microkernel; I bet this is still true in XP. I don't know if the OS/2 subsystem is still around - if I were Microsoft I probably would have dropped it with Win2k. But that doesn't mean you couldn't write your own servers. There has been talk of porting UML to Windows - can't recall if they wanted to use Win32 or the bare NT ukernel, though.
Maybe they do, maybe they don't. It's a moot point, since people generally pick crappy passwords. If I were working for Microsoft and it were my job to handle the security architecture of Passport, I'd do it as carefully and painstakingly as possible ... yet we know their track record on these things, and it's not reassuring. And unless they can enforce good security practices on the users (picking passwords, not sharing them with others, not writing them down in obvious places, etc) it doesn't really matter all that much.
What users do with secure information like their own passwords has always been a concern, but it is orders of magnitude worse when all the security information is centralised. Yet another reason to fear MS P....
Good question. Anyone who partners (I hate that verb) with Microsoft to provide services using Passport should take any information they get from said service with a major grain of salt.
It's worth noting that email used to be this way. And some people tried to solve the problem using the Trillian method - have a single email client that spoke cc:Mail, MS Mail, SMTP, etc. Indeed, Outlook is still commonly installed with a legacy plugin for Exchange Server. (Well, "legacy" for its email functionality - I know the Exchange plugin does much more than email.)
Nowadays everyone speaks [E]SMTP except certain corporate holdouts, and even those holdouts generally offer an SMTP gateway.
Another comparison: most of those proprietary email systems were centralised by design - there was the one email server, or one set of servers, and the clients. Just like IM today. I'm convinced that to truly unify IM would require decentralisation - the user@serverhost model which SMTP and I believe Jabber use. The IRC model doesn't really scale, though it has the advantage that people don't need to know each other's server names.
It's one thing to create a key generator for something like FLEXlm (whose license keys are fairly short - possibly because a once-popular way to get a license was via FAX, which forced the customer to type in a random hex string - so the shorter the better) or DVD CSS (intended for embedded devices, so they needed to minimise the hardware cost for decoding, which must be done in real time). It's quite another thing to crack a protocol which is designed to run on a 700 MHz 32-bit CPU in a non-time-critical path (only during boot, as opposed to during all DVD playback).
In other words, unlike many cases, Microsoft wasn't forced to use short keys or fast algorithms. They could quite easily have used something which is not computationally feasible to crack within the next 50 years.
Although, given Microsoft's track record with regard to security implementations, one can always hope....
Not the RJ45 plugs I buy. They're labeled "for solid" or "for solid or stranded" and they have three knife points in a _-_ pattern. They bend around the wire core and work great.
I've had a lot more trouble with patch cables I've made from stranded wire and "for stranded" plugs. I'm willing to admit the possibility that this is due to other factors - quality of the cable, or my relative inexperience at crimping - but in any case, once I switched to solid cable I haven't looked back.
The signature most likely (i.e. "if Microsoft isn't a lot stupider than they look") contains a cryptographic checksum. That would preclude any attempts similar to what you describe. Cryptography discovered, and solved, this problem perhaps 25 or 30 years ago.
And you were doing so well up until that point. What part of copyright law do you recall? The part where the author / artist of a work has an absolute monopoly on the right to publish that work, allowing others to publish only small excerpts without explicit permission?
Copyright law doesn't guarantee profits, but it does guarantee protection from competition for a "limited" time. People who break that law are, well, breaking the law.
It's also very easy to dismiss artists who think they're being cool and/or shocking and/or tortured and/or neurotic but whose quirks are a lame, transparent stage show.
This is a variation on the Sagan quote about Bozo the Clown. Just because you're dismissed by society at large doesn't mean you have necessarily hit a sore spot society is afraid to face.
I have no particular opinion on Eminem; just thought I'd throw that out.
Har har har.
Good point, especially now that memory (RDRAM included!) is so cheap. But soon enough there will be good P4 motherboards that take DDR. Then you can pick a platform and a memory.
Thanks..
OK, +1 Informative, that's something I didn't know. I assume, since you didn't mention it, that I guessed correctly (yes it was a guess) about RDRAM interleaving access between two RIMMs to double the bandwidth.
Right - but I didn't know "DDR" was the generic term for sending signals across both clock edges. Thanks again - that makes +2 Informative (to me anyway).
Perhaps, but you're probably comparing single-stick to single-stick. With RDRAM you have to buy a matched pair. So the right comparison is 2x128 PC800 ($80) versus 1x256 PC2400 ($51).
Or go on up to 512MB. 2x256 PC800: $148. 1x512 PC2400: $114.
So RDRAM costs an additional 57% for 256MB, or 30% for 512MB. Nice that it's no longer double the cost, but to me that is still a significant markup. Anyone know approximately how much of that is due to
(b) manufacturing cost after accounting for (a), or
(c) patent licenses?
RTFG. First hit lists it for $189.
When it comes to wrist pain and typing speed, money really is no object. At least, $200 is a pretty small price to pay.
Not that I'd get one, most likely - I'm Yet Another IBM Model M Fan. Not sure what I'll do when motherboards start coming without PS/2 ports. That'll suck. Oh, I know, I'll use a PS/2 port card ... except that these motherboards won't have ISA slots either ... aw, now I'm just getting depressed. Does anyone sell a USB-to-8042 bridge?
In the beginning there was PC100 SDRAM. Well, actually, that was mid-nineties, but that's about when most Slashkiddies were born, so moving on. Obviously everything is just a marketing label, but this one meant 100 MHz. With SDRAM, each Hz gives you 64 bits, so the bandwidth is 6400 megabits per second.
Thus PC133 and PC166 are 8500 and 10700 Mb/s.
DDR is the same tech as SDRAM, except that it uses a trick to transfer data twice per clock cycle, so you get 128 bits per Hz. Thus PC100 DDR-SDRAM would be 12800 Mb/s. But Marketing decided that was unfair, so they labeled DDR based on twice the clock speed, so we have PC266 and PC333, which of course run at 133 and 166 MHz and give you 17000 and 21000 Mb/s.
RDRAM is based on a new tech that gives you only 16 bits per clock cycle instead of 64 for SDRAM and 128 for DDR-SDRAM. The difference is that you can clock it way up. So there was PC600, PC700 and PC800 RDRAM, again based on MHz, so that gave you 9600, 11200, and 12800 Mb/s bandwidth. Basically you divide the number in four to compare with SDRAM speeds, since you only get 1/4 as many bits per cycle. Actually I believe modern Rambus controllers double this by interleaving two sticks, so now you divide by 2 - PC800 has four times the bandwidth of PC100, but requires a matched pair of sticks.
Then the DDR people decided to start talking direct bandwidth, rather than megahertz. But unlike me, they mean megabytes, rather than megabits, per second. PC1600 is DDR-SDRAM at 100 MHz, since DDR gives you 128 bits or 16 bytes per cycle. PC2100 is DDR at 133 MHz, formerly known as PC266. PC2700 is DDR at 166 MHz, and PC3200 is DDR at 200 MHz.
With interleaving, Rambus gives you 32 bits or 4 bytes per cycle. PC800 has the same bandwidth as PC3200 DDR, and the relatively new PC1066 has more - 4266 megabytes per second.
Bandwidth is a good baseline for comparison, but RDRAM has a higher latency than SDRAM or DDR-SDRAM. That's why DDR, with its lower maximum bandwidth, is still speed-competitive with RDRAM (for a lot less money).
We used to do that. SIMMs were 32 bits wide and the Pentium architecture wanted 64 bits at a time, so you had to pair them up. People found that annoying, so DIMMs have 64 bits of bandwidth. I think consumers would balk at having to buy DIMMs in sets of 2 or 4 or 8, though server class people don't seem to mind so much, since at the server end you don't really expect a commodity market for upgrades anyway.
You'll note that RDRAM does continue to use this trick. I don't know if individual RIMMs are only 8 bits wide or what, but you do have to have matched pairs. (It would surprise me if a RIMM were only 8 bits wide, since I've seen some with 16/18 chips.)
Perhaps the other answer to your question is that with SDRAM, you have 64 bits on each DIMM, so interleaving means your memory controller has to be able to handle 128 or 256 or 512 bits at a time, which is a lot of traces and a lot of logic and perhaps a lot of expense for the motherboard and specifically the controller / L2 cache.
Beg to differ. The AMD Hammer will actually have a DDR memory controller on-chip. That's what I call integration between the processor and the memory. I don't see that lawyers and patents enter into this at all. I know it is currently vapor, but I'm very much looking forward to seeing the Hammer in action (and probably buying one).
If that were true we wouldn't have PNG. Unisys decided to start making threats about LZW compression only after GIFs became popular - i.e. very late in the game.
As noted elsewhere, you're thinking of trademarks (and probably, by extension, service marks).
I can think of half a dozen ways to break chroot if you are root - how do you 'get out of jail free' if you aren't?
Ah yes, the one that went down into the volcano. (Well, Unit 01 did too.) I guess Unit 02 must have had a pretty decent cooling system - actually probably more than $500 worth.
What?!? You have to reboot for software updates? You must be running either HP/UX or something from Microsoft! (: Oh, you mentioned viruses - never heard of viruses for HP/UX, so....
(I still fondly remember the time, must have been 1997 or 1998, when someone on Usenet came up with an idea for how init could restart itself without losing state. The sysvinit maintainers promptly implemented it, thereby eliminating the last reason to have to reboot a Linux machine to update any software short of the kernel image itself.)
Would that be your local master browser, or your domain master browser? (:
I don't think it works out that way, because the label passes these costs on to the artist. That's why if you are a moderately successful artist, you go broke, because all your royalties go to paying back the label for all those expenses you just listed. I believe most record contracts work this way, with the label only lending money for up-front costs.
Somehow VH1 always forgets to mention this part on Behind the Music. They always talk about the hotshot band finally making it big time and living large, and then a couple years later going bankrupt because they "took some bad business advice" or "were too generous to their relatives" or "got careless and didn't keep track of their money". I figure most of the time it's really "got screwed by the label contract" and for some odd reason VH1 just doesn't want to put it that way.