Wait, so, they're saying that in 1999 (times 2 million) BC, a runaway nuclear reaction blew the moon INTO orbit? Someone call Gerry Anderson, I smell a prequel.
Yea, when I saw the line "and could be used by post-apocalyptic people to feed a hungry planet", I repeated back from the article "Yea, if they can get to place that's built 'into the side of a mountain. On a remote island. Near the North pole.'"
I would comment on this article, but slashdot is sadly behind the times and doesn't support my Microsoft Slashdot (tm) Reader's (tm) Article Post (pat. pend. microsoft) Internet (tm) Extension (tm), or even the patentend "Click-to-post" button... I'm afraid until Slashdot straightens up and supports modern (microsoft) interoperable (microsoft) standards, and joins the (microsoft) internet that they'll be sadly left behind.
(tm)
Well, you don't see a use because you live in a country with essentially unmonitored and easy access to information...
I see this as being of great interest to dissident groups. You disseminate information from the backpack cell. Members just need a laptop, and to be in the vicinity. They don't even have to really know each other, or who the guy with the backback is. The gov't would have to quickly pick up on the ap, and zero in on the signal.. And they wearer can be walking through the street market, as are the people with the laptops busily downloaded the censored information...
Drawing from today's headlines, say the Taiwanese gov't cracks down on the KMT; they could walk through the nightmarket and exchange info. bring the AP to an internet cafe, and not even use the cafe's network, but still have an online exchange.
The racks were layed out (and the numbering scheme designed) to yank out the three across towers and rack in 5 down xserves. And the power and cooling budget allows for it.
Actually, yea, it is, because we used up all the spare capacity on the pdu's. We need to install a new one if we except to keep buying computers for the rest of the machine room.
No, because the entire complex was designed to account for the arrival of xserves eventually. The space was even layed out for it; you pull out the three across towers and rack in 5 down xserves, and the numbering scheme for the nodes already allows for it. Although I believe it was laid out to allow for up to 2u boxes, so more 1u units would fit, and still be in the heat and power budget (although the numbering scheme would have to be redone. Which I think took longer than installing the processors)
That was mentioned back when the project was first talked about.
The question becomes, assuming a shipment of xserves to tech is announced, how long will it be before people start yelling again about not getting their machines because of bigmac?
Uh, no. That vast farm of already-cabled empty racks in the machine room are waiting on G5s to be dropped into them. Much thought has gone into airflow and cooling. If a as yet unannounced G5 xserve becomes available in the future, the racks are well positioned to swapping them in. Assuming the same form factor as the current xserve, they can handle 5 xserves in the space where three g5 are now.
No, but what you're probably going to see is an evolution of the way they do ads. Instead of breaks every N minutes for 2-3 minutes of non-program commercials, you'll probably have fewer seperae commercials, and a lot of on-screen crap. In addition to the ever-present bug in the corner, you'll have the bottom 1/8th of the screen constantly filled with product ads and commercials. Some of the cable channels have been doing full-blown animations and eye catchs for other programs that are frickin HUGE, and very *not* transparent. Now imagine those as product ads. A pepsi can dances onto the bottom of your screen, and "pepsi: choice of an ad-loving generation" appears. Heck, map them to product tie ins. Someone drinks a pepsi on screen, and a pepsi ad drops into the bottom.
I'm waiting for them to realize tha once a large number of homes have wide screen tvs that they can continue producing and airing 4:3 programming, but to those higher income "we got a $3000 tv" send out the 16:9 signal, with a 4:3 image in the middle, and commercials running non stop down either edge.
There's just too much of a lack of quality radio programming these days for me to ever consider buying one of these.
Yea, they could loose the radio tuner, and substitute a subscription from clearchannel that just once a month sends you the radio shows you would have heard if you were using the radio anyway. Regardless of where in the country you are...
Yea. The hubble site has it in multiple formats. The 738K "large print" format would make a decent bit of wall paper, but if you're a masochist, or just really into astronomy, you can tweak your own from the 127Meg raw tiff...
No problem:-). The sad thing is I made a, well, typo isn't quite right.. That last "Soyuz" should have been "Salyut".. The T's were the Soyuz intended specifically to transport folk to the Salyut stations.
Yes, it was. In this case, it's not Tycho Magnetic Anomaly, but rather Transport Mir Anthropometric. The TMA's are the "large astronaut" retrofit (The US allows taller astronauts than the russians do; surprisingly few of ours fit the older soyuz, which means they could never be station crew) of the TM model, which, itself, was the unit customized to be the ferry craft for Mir from the T class transport which was supporting Soyuz...
Actually, if you hear your own traffic coming back, something is very wrong.
The limit comes from the Collision Detect portion of CSMA/CD. A cable run in a collision domain is controlled by the the amount of time it takes for the minimum size packet to make it to the most distant station, be detected, and a collision signal to make it back, before said minimum sized packet has finished transmitting.
Collisions happen because station a started transmitting, not hearing anything on the cable. (i.e. the cable voltage showed no other signal.) Meanwhile station a's injection of voltage, hurtling along at the speed of light in copper hasn't reached station b, who also starts transmitting, thinking the cable is unused. When they detect that the cable voltage has jumped to roughtly twice normal strength or more, that means there's a collision going on, and we need to jam the cable momentarily to indicate that, and start the backoff and retransmit process. If the cables too long, and the stations are too far apart, it's possible for the packets to cause a collision, but the station not be aware of it, because it finished transmitting before the collision arrived; that's a late collision, and it means you're going to be mysteriously dropping data.
It's better and worse in the xBaseT world, because you only have two stations to worry about, not a coax full of them, but the stations are going to be at the extreme ends of the cable, so you can go from "working fine" to "screwed" very quickly as you add length...
Minimum packet sizes and cable lengths in ethernet were set by doing a sort of dance between the two, balancing "reasonable size" vs. "reasonable length". If you make the minimum packet be, say, 20K, you can make the cables much much longer and still have no problems. But now you have all that other overhead of sending 20bytes of data requires 19980 bytes of useless crap wrapped around it, waiting for a packet gap takes longer, etc, etc.
I read the specs a while back. They're interesting reading for anyone who's done a little network programming - you can get the pdf off their web site.
Basically, it's their own protocol with ethernet framing, which will allow them to use switches and such. 3com's customized some of the asic's.
When you plug a device in it has a specific sequence of events that happen, that get the network master to issue it a magic id, and allow it to introduce itself to its neighbors, and find out what functions they have, and tell them what functions it has, and that properly gets passed up the chain so that the appropriate devices find each other. That allows you to plug a magic guitar into a magic amp, for instance, and have the guitar automatically map its volume control knob, for instance, to control the amp. On a more complex level, it'd allow "Bob's Guitar" to pop up as a channel label on the mixing boad automatically, with the board recognizing it, and firing off a set of presets to a known state automatically. The master node (which is selected through an unambiguos set of rules, which handle networks dividing and recombing well) is responsible for assigning numbers to everyone, which can regularly and easliy change - it just walks the network tree and and everyone assigns themselves the next number, and is also the master timing source.
All you need is a huge rack of cd-duplicator machines... You can buy boxes that you just drop the cd into the top, and a stack of 5, 10, even 20 drives immediately clone the disk. I've even seen some that clone the disk, then drop it into a disk printer. Since concerts generally have the same order, you predo the jewel case inserts, and pre silk-screen the blanks. Record off the mixing board onto a digital source, and immediately burn to a master cd, and drop that into the first duplicator. Then burn another, and drop that into the next duplicator.
If you bring a truck to the concert with all that stuff pre-racked and powered, you could easily start churning out a couple of hundred cds every 15 minutes or so.
Honestly, I'd support a retinal scan based i.d. card, but strongly oppose a fingerprint one. I don't so much mind my retina being on file, because I don't tend to innocently leave it lying around on things. When the only fingerprint they can lift at the crime scene is mine from when I was innocently there three weeks ago (as opposed to the crook who wore gloves) I object to them being able to just come hunt me down to make me prove it wasn't me.
On the other hand, if they find my eyeball, I'm perfectly happy for them to be able to quickly figure out that I'm the one missing one:-).
Yea, I love my call intercept. But did you notice in the proposal the requirement that all telemarketers will have to provide caller-id information? There goes call intercept. That's what it *keys* on. It'll let through all the calls it's been blocking (my telemarketing calls went from 50-60 month to about 2) because they'll all be "known" callers.
It would be better to require that they all have something identifiable in their caller id that indicates "telemarketing".. That way services and cheap gadgets could see "incoming marketer" and decide for you if you wanted the call.
Huh. Maybe there are some companies that are misusing it, cause I've seen (S) next to some things. Similarly there's a (P) that goes with the (C) involving recorded music.
Sorta, yes and no... The (c) may well *be* a perfectly legal symbol - the problem is no one has taken someone to court on a copyright case that uses it, and had the (c) not being a legitimate mark be challenged, and had the court rule that it was indeed legit. The law only specifically mentioned the copyright symbol and the word. However, thanks to a court decision just like I was just mentioning, C in hexagon is perfectly valid. The court ruled that it was obvious from the context what the person intended to convey, and thus it was a perfectly legit display of copyright. Given that precedent, there's every probability that (c) will be upheld, given the widespread use of it, should the issue ever come up.
Of course, if you want to be *sure*, and it's important to you, use the real logo, or spell out the word, and you won't be spending money later on being the test case...
Um, that's what all that paperwork is for that goes to the library of congress. To centrally register your copyright.
You don't *have* to do it, but if you do, you get extra legal benefits if you sue.
Similarly, there's no need to centrally register a TM. You can put that on anything you feel is one of your trademarks. You're thinking of the R in a circle (R), that's a *registered* trademark. (There's also an (S) for registered Service Mark).
*Now*, yes. After the last major set up changes to the copyright laws, in the late 80s, the convention was that copyright existed as soon as something was put into fixed form. So, you write a story, even if you didn't mark it as copyrighted, it was. Now, if you mark it as such, and file paperwork with the library of congress to register it as having been copyrighted, then you can claim significantly higher damages if someone rips you off. But you can always get them to just stop ripping you off if you don't, so long as you can prove you wrote it before they did.
But before that last revision copyright didn't exist unless you specifically marked *every* copy as being copyrighted. Let something get out into the public without a copyright notice, and you'd just accidentally released it into the public domain.
That's why there's a version of the hobbit that is fair game for anyone; they didn't copyright it when it hit the US. Minor changes were made, and the next edition was.
Sun ONE is a particular combination of service software that sun has, and is principally a marketting thing. So, things like the Iplanet stuff, various ecommerce bits, etc, all fall under SunONE. It incorporates a bunch of stuff that's all largely based around open standards.
The Liberty Alliance is a group, of which Sun is a founding member, which is producing a standard for a particular service.
So, at some point the SunONE offering, if they've haven't moved on from that name, will likely implement the Liberty Alliance authentication standard as one of the features in the appropriate products, and might include the softare to manage the server side of it as a SunONE product.
Slashdot Mirror
Wait, so, they're saying that in 1999 (times 2 million) BC, a runaway nuclear reaction blew the moon INTO orbit? Someone call Gerry Anderson, I smell a prequel.
Yea, when I saw the line "and could be used by post-apocalyptic people to feed a hungry planet", I repeated back from the article "Yea, if they can get to place that's built 'into the side of a mountain. On a remote island. Near the North pole.'"
I would comment on this article, but slashdot is sadly behind the times and doesn't support my Microsoft Slashdot (tm) Reader's (tm) Article Post (pat. pend. microsoft) Internet (tm) Extension (tm), or even the patentend "Click-to-post" button... I'm afraid until Slashdot straightens up and supports modern (microsoft) interoperable (microsoft) standards, and joins the (microsoft) internet that they'll be sadly left behind. (tm)
Yes, Debian sounds amazingly like the mandarin word for "shit"...
I see this as being of great interest to dissident groups. You disseminate information from the backpack cell. Members just need a laptop, and to be in the vicinity. They don't even have to really know each other, or who the guy with the backback is. The gov't would have to quickly pick up on the ap, and zero in on the signal.. And they wearer can be walking through the street market, as are the people with the laptops busily downloaded the censored information...
Drawing from today's headlines, say the Taiwanese gov't cracks down on the KMT; they could walk through the nightmarket and exchange info. bring the AP to an internet cafe, and not even use the cafe's network, but still have an online exchange.
There's all sorts of subversive uses.
Given they're 1U and not 2U, even more would fit.
Actually, yea, it is, because we used up all the spare capacity on the pdu's. We need to install a new one if we except to keep buying computers for the rest of the machine room.
That was mentioned back when the project was first talked about.
The question becomes, assuming a shipment of xserves to tech is announced, how long will it be before people start yelling again about not getting their machines because of bigmac?
Uh, no. That vast farm of already-cabled empty racks in the machine room are waiting on G5s to be dropped into them. Much thought has gone into airflow and cooling. If a as yet unannounced G5 xserve becomes available in the future, the racks are well positioned to swapping them in. Assuming the same form factor as the current xserve, they can handle 5 xserves in the space where three g5 are now.
I'm waiting for them to realize tha once a large number of homes have wide screen tvs that they can continue producing and airing 4:3 programming, but to those higher income "we got a $3000 tv" send out the 16:9 signal, with a 4:3 image in the middle, and commercials running non stop down either edge.
Yea, they could loose the radio tuner, and substitute a subscription from clearchannel that just once a month sends you the radio shows you would have heard if you were using the radio anyway. Regardless of where in the country you are...
Yea. The hubble site has it in multiple formats. The 738K "large print" format would make a decent bit of wall paper, but if you're a masochist, or just really into astronomy, you can tweak your own from the 127Meg raw tiff...
No problem :-). The sad thing is I made a, well, typo isn't quite right.. That last "Soyuz" should have been "Salyut".. The T's were the Soyuz intended specifically to transport folk to the Salyut stations.
Yes, it was. In this case, it's not Tycho Magnetic Anomaly, but rather Transport Mir Anthropometric. The TMA's are the "large astronaut" retrofit (The US allows taller astronauts than the russians do; surprisingly few of ours fit the older soyuz, which means they could never be station crew) of the TM model, which, itself, was the unit customized to be the ferry craft for Mir from the T class transport which was supporting Soyuz...
Actually, if you hear your own traffic coming back, something is very wrong.
The limit comes from the Collision Detect portion of CSMA/CD. A cable run in a collision domain is controlled by the the amount of time it takes for the minimum size packet to make it to the most distant station, be detected, and a collision signal to make it back, before said minimum sized packet has finished transmitting.
Collisions happen because station a started transmitting, not hearing anything on the cable. (i.e. the cable voltage showed no other signal.) Meanwhile station a's injection of voltage, hurtling along at the speed of light in copper hasn't reached station b, who also starts transmitting, thinking the cable is unused. When they detect that the cable voltage has jumped to roughtly twice normal strength or more, that means there's a collision going on, and we need to jam the cable momentarily to indicate that, and start the backoff and retransmit process. If the cables too long, and the stations are too far apart, it's possible for the packets to cause a collision, but the station not be aware of it, because it finished transmitting before the collision arrived; that's a late collision, and it means you're going to be mysteriously dropping data.
It's better and worse in the xBaseT world, because you only have two stations to worry about, not a coax full of them, but the stations are going to be at the extreme ends of the cable, so you can go from "working fine" to "screwed" very quickly as you add length...
Minimum packet sizes and cable lengths in ethernet were set by doing a sort of dance between the two, balancing "reasonable size" vs. "reasonable length". If you make the minimum packet be, say, 20K, you can make the cables much much longer and still have no problems. But now you have all that other overhead of sending 20bytes of data requires 19980 bytes of useless crap wrapped around it, waiting for a packet gap takes longer, etc, etc.
I read the specs a while back. They're interesting reading for anyone who's done a little network programming - you can get the pdf off their web site.
Basically, it's their own protocol with ethernet framing, which will allow them to use switches and such. 3com's customized some of the asic's.
When you plug a device in it has a specific sequence of events that happen, that get the network master to issue it a magic id, and allow it to introduce itself to its neighbors, and find out what functions they have, and tell them what functions it has, and that properly gets passed up the chain so that the appropriate devices find each other. That allows you to plug a magic guitar into a magic amp, for instance, and have the guitar automatically map its volume control knob, for instance, to control the amp. On a more complex level, it'd allow "Bob's Guitar" to pop up as a channel label on the mixing boad automatically, with the board recognizing it, and firing off a set of presets to a known state automatically. The master node (which is selected through an unambiguos set of rules, which handle networks dividing and recombing well) is responsible for assigning numbers to everyone, which can regularly and easliy change - it just walks the network tree and and everyone assigns themselves the next number, and is also the master timing source.
All you need is a huge rack of cd-duplicator machines... You can buy boxes that you just drop the cd into the top, and a stack of 5, 10, even 20 drives immediately clone the disk. I've even seen some that clone the disk, then drop it into a disk printer. Since concerts generally have the same order, you predo the jewel case inserts, and pre silk-screen the blanks. Record off the mixing board onto a digital source, and immediately burn to a master cd, and drop that into the first duplicator. Then burn another, and drop that into the next duplicator. If you bring a truck to the concert with all that stuff pre-racked and powered, you could easily start churning out a couple of hundred cds every 15 minutes or so.
Honestly, I'd support a retinal scan based i.d. card, but strongly oppose a fingerprint one. I don't so much mind my retina being on file, because I don't tend to innocently leave it lying around on things. When the only fingerprint they can lift at the crime scene is mine from when I was innocently there three weeks ago (as opposed to the crook who wore gloves) I object to them being able to just come hunt me down to make me prove it wasn't me. On the other hand, if they find my eyeball, I'm perfectly happy for them to be able to quickly figure out that I'm the one missing one :-).
Yea, I love my call intercept. But did you notice in the proposal the requirement that all telemarketers will have to provide caller-id information? There goes call intercept. That's what it *keys* on. It'll let through all the calls it's been blocking (my telemarketing calls went from 50-60 month to about 2) because they'll all be "known" callers.
It would be better to require that they all have something identifiable in their caller id that indicates "telemarketing".. That way services and cheap gadgets could see "incoming marketer" and decide for you if you wanted the call.
Only if it's got 8 sides instead of six.
Huh. Maybe there are some companies that are misusing it, cause I've seen (S) next to some things. Similarly there's a (P) that goes with the (C) involving recorded music.
Sorta, yes and no... The (c) may well *be* a perfectly legal symbol - the problem is no one has taken someone to court on a copyright case that uses it, and had the (c) not being a legitimate mark be challenged, and had the court rule that it was indeed legit. The law only specifically mentioned the copyright symbol and the word. However, thanks to a court decision just like I was just mentioning, C in hexagon is perfectly valid. The court ruled that it was obvious from the context what the person intended to convey, and thus it was a perfectly legit display of copyright. Given that precedent, there's every probability that (c) will be upheld, given the widespread use of it, should the issue ever come up.
Of course, if you want to be *sure*, and it's important to you, use the real logo, or spell out the word, and you won't be spending money later on being the test case...
Um, that's what all that paperwork is for that goes to the library of congress. To centrally register your copyright. You don't *have* to do it, but if you do, you get extra legal benefits if you sue. Similarly, there's no need to centrally register a TM. You can put that on anything you feel is one of your trademarks. You're thinking of the R in a circle (R), that's a *registered* trademark. (There's also an (S) for registered Service Mark).
*Now*, yes. After the last major set up changes to the copyright laws, in the late 80s, the convention was that copyright existed as soon as something was put into fixed form. So, you write a story, even if you didn't mark it as copyrighted, it was. Now, if you mark it as such, and file paperwork with the library of congress to register it as having been copyrighted, then you can claim significantly higher damages if someone rips you off. But you can always get them to just stop ripping you off if you don't, so long as you can prove you wrote it before they did.
But before that last revision copyright didn't exist unless you specifically marked *every* copy as being copyrighted. Let something get out into the public without a copyright notice, and you'd just accidentally released it into the public domain.
That's why there's a version of the hobbit that is fair game for anyone; they didn't copyright it when it hit the US. Minor changes were made, and the next edition was.
Sun ONE is a particular combination of service software that sun has, and is principally a marketting thing. So, things like the Iplanet stuff, various ecommerce bits, etc, all fall under SunONE. It incorporates a bunch of stuff that's all largely based around open standards. The Liberty Alliance is a group, of which Sun is a founding member, which is producing a standard for a particular service. So, at some point the SunONE offering, if they've haven't moved on from that name, will likely implement the Liberty Alliance authentication standard as one of the features in the appropriate products, and might include the softare to manage the server side of it as a SunONE product.