There have been a high number of occurances of 503's since the zero-notice updates a few weeks ago. (at the same time, all web pages started returning "no-cache" so simple brower navigation is forced to redownload every byte on every mouse click. When I logged a bug about this, it was immediately dismissed without comment.)
Excuse me for asking... exactly what the hell are they argueing about? The only units that record digital content are directv receivers. And the DTivo's will never have ethernet (HMO) enabled -- DTV explicitly stated the USB ports (and thus ethernet) are not to be enabled on DTivos. (No S1 unit will ever have the Home Media Option, and thus sharing ability, available to it, but does support an ethernet card.)
That's right, the only systems capable of sharing recorded content recorded that content off air -- standard definition, OTA... none of this FCC broadcast flag bullshit even applies.
A few other errors... no US model Tivo has digital outputs. Only DTV models record digitally broadcast data. The others digitized an analog broadcast. Sharing of recorded content will/can not be in real-time -- the item will have to be recorded completely before being transmitted at speeds 10-20x (SD) or 100x (HD) slower than the recorded bit rate, based on today's broadband technology. It is unlikely people will have home broadband connections capable of streaming HD content in realtime in the coming decades.
Except the mail isn't exactly realtime. You'll have to record the entire game, burn to DVD, and then mail it out. In the end, the DVD will not be viewed for several days after the game was aired.
Actually, you'd be surprised how closely they pay attention to these sorts of things. In order to receive sports packages, the unit must be connected to a phone line. They can (and do) verify the origin of the calls... and they will see this "LA" box calling from Oakland.
They've ruined NFL Sunday Ticket (tm) anyway... if any local station, even those DTV doesn't carry, claims to be carrying a game, it gets blacked out to force you to want the local broadcast station. In my experience over the past few years, DTV has been wrong 90% of the time. (And the local station guide data is correctly listing the right game.) And that damned "enhanced screen" is enough to get me to cancel my entire DTV.
(And it's DTV who have a case for fraud and possibly breach on contract.)
It's a DIRECTV Tivo. They, technically, never need to make any calls. The guide data comes from the dish as does a number of other things. Software updates and stats use the phone line, but really never have to happen:-)
And what makes you think the monkey loading the vending machine will not notice a plastic coke can? This is exactly why you will not find a promo can in a vending machine -- the human loading the machine will notice the abnormal "can".
And just why the hell would you be trying to carry a case of Coke through the airport? It's instantly obvious the promo coke's are not a real can of coke, so knowingly carting the thing around in places where you know better (or should know better) will get you exactly what you deserve.
To be fair, IOS just has an ELF header. IOS is a single binary loaded and executed. So, it could almost have no header and still be perfectly usable... (it'd take a boot rom update and some changes to internal Cisco processes...)
.../Cisco/12.3.9/[]:file c1700-k9o3sy7-mz.123-9.bin c1700-k9o3sy7-mz.123-9.bin: ELF 32-bit MSB executable, version 1, stripped
.../Cisco/12.3.9/[]:unzip -L c1700-k9o3sy7-mz.123-9.bin Archive: c1700-k9o3sy7-mz.123-9.bin warning [c1700-k9o3sy7-mz.123-9.bin]: extra 17208 bytes at beginning or within zipfile
(attempting to process anyway)
inflating: c1700-k9.bin
.../Cisco/12.3.9/[]:file c1700-k9.bin c1700-k9.bin: ELF 32-bit MSB executable, version 1, stripped
Now, the new "modular" IOS 12.4 may be a very different story.
Re:v6 could help solve some net problems
on
IPv6 is Here
·
· Score: 1
DHCP is a superset of BOOTP. So, DHCP can do everything BOOTP can (and more.) And that means just about everyone is running a bootp server -- most dhcp servers will answer a bootp request.
The BOINC presentation from Madrid claims folding@home is using BOINC. Unless they've recoded the clients without telling anyone, this is incorrect. Folding@home uses Cosm as it's communications framework. There is no reference to BOINC in any of the documentations I've seen from f@h.
The tracker can be run on any port. Likewise, the clients can (and do) use any port. Many (and I mean a lot) of ISPs around the world rate limit BT traffic on the "well known ports". The only reliable way to get accurate numbers on bittorrent use is to decode the traffic -- peek at the actual content of each packet look for the BT protocol.
Kazaa works much the same way.
(Without the actual.torrent, it's a little difficult to identify what's being traded without piecing together all the data.)
Moderators, please correct the lead-in... BIND and the global DNS system is what has the diversity. The problem with Akamai was their lack of diversity on top of their proprietary hacks to DNS.
What's to stop you from doing that to a normal, non-promo, can?
I think they're being a little too paranoid. It's not like you won't notice it's not a real can of coke the instant it's in your hand. By that reasoning, it's very unlikely you'll find one of those cans in a drink machine. (if you were loading the drink machine, would you put it there? Or keep it for yourself/a friend?)
The chopper is "for illustrative purposes only". Read the small print on the screen during the commercial. Remember the Pepsi Stuff (tm) promotion showing a harrier? Pepsi didn't have any such disclaimer and some kid collected enough points...
I have no idea what's acutally inside the can. But I'm betting it's not your average nokia cellphone and an earthmate gps. It's most likely a cellular modem -- read: NO MICROPHONE -- and a dirt cheap GPS. Given the average shelf-life of a cell phone battery's charge, I really doubt it's a full cell phone.
Actually, above a few hundred peers in any torrent swarm, the performance drops off greatly. Can your BT client maintain connections and state for 38,000 connections? Most clients limit the number of peers per torrent to 100.
Actually, it tends to saturate bandwidth on all sides and right down the middle, too:-) I have seen (and about 6 seconds away from getting a coworker fired, to boot) a single warez node consume 30% of an ISPs entire aggrigate bandwidth. When you do this on the wrong side of a SONET ethernet bridge, it becomes VERY noticable. (OC3 carrying a DS3, 6 T1s, a 10M ethernet, and a 100M ethernet VLAN span... yes, that's more than 155mbps.)
Up until now, it's just people trading files because they like trading files fast.
It's a pissing a contest... who can transfer the file the fastest. They don't care what they are moving or who they may be hurting in the process. I sat and watched these idiots once... a dozen "people" all pushing the same file(s) to the same server; the first one to get the entire file there "wins". This is both stupid and wasteful.
Often it'll hit newsgroups before it gets made into a torrent somewhere
For some things, yes. But for more popular items, a torrent will ususally be available before it appears in part or whole on USENET. (I know, because I watch... and, sadly, usenet is faster than downloading via bittorrent. esp. from such places as suprnova)
Nice theory, but it falls flat when you've ordered the drives directly from Maxtor -- and they arrive with a customs form on them. Face it, quality and reliability circled the drain some time ago... Make 'em cheap and make 'em quick. It's better on the bottom line to get people to buy new drives every six months instead of using the same one for 5 years. (now we're back to one of the reasons SCSI drives are so expensive.)
Assuming a human loads them in the box, that's the only place a person handles the drives (already sealed in their static bag) before they come out of the case. I'm reasonablly sure humans are involved in loading the retail boxes. But I could be wrong -- people are expensive, even in Thailand.
I don't buy drives on ebay. The 80G drives were by the case (it takes a few of them to amass 1TB:-)) The 250G ones are from our system builders (assmeblers, whatever) who buy them by the case.
Even the 80G Maxtor in my Dell workstation (8300, 9 months old) failed. Dell replaced it with a WD drive. That speaks volumes to me.
Read the snippet about Command Queuing. That takes a good bit of memory on the drive controller to handle well. As Seagate makes some serious SCSI drives, I think they'll make good use of 16MB.
If they are both the same make and model from the same manufacturer, they are both equally statistically likely to fail. Odds are, they will not both fail at the same time. However, they are highly likely to both fail within the same time period. When they were built makes little difference as running time is a greater contributor to failure than shelf age. I had two 160G drives (same make/model) built nearly a year apart in different factories (Maylasia and Thailand) installed in the same system (same running history -- RAID0 so almost identical access even) die within two days of each other.
Therefore, when one fails, replace both of them. Or better yet, use two different drives... say a Hitachi and a Seagate.
IDE drives have the shittiest reliability. And it's getting worse, not better. Several years ago, in my experience building fileservers, Maxtor had a failure rate of 25%. Right. Out. Of. The. Box. And that was for 80G drives that they'd been making for several years at that point. Today, they have a 60% failure rate over the first week (about the same OOB.) And this is with their "top of the line, enterprise class" crap. You'd think with a 3yr warantee, they'd spin the damn thing up at least once before shoveling 20 of them in a box.
SCSI is the only way to get drives that have actually been through any testing. Each drive is individually tested; however, with IDE, only a small sample of drives are tested. This is one reason SCSI is more expensive. But demand, perception, and the money in the enterprise market place also factor into the cost... a 140G SCSI drive does NOT cost 1000$ to build and test. They use the exact same servo hardware as their IDE "white trash" cousins (in many cases -- 10k and 15k speeds aside.)
[Disclaimer: I don't have as much experience with Seagate's IDE (PATA or SATA) lineup. But I can say, I've never had any Seagate drive, SCSI or IDE, fail right out of the box or shortly there after. Of the few that have failed, 2 overheated and melted their logic boards (temp. swapped with another drive to fetch the data:-)) which Seagate replaced. Another half dozen developed "stiction" problems after several years and needed a little help to get spun back up.]
Slashdot Mirror
And this is why I don't pay for /. (I will never get my money's worth.)
There have been a high number of occurances of 503's since the zero-notice updates a few weeks ago. (at the same time, all web pages started returning "no-cache" so simple brower navigation is forced to redownload every byte on every mouse click. When I logged a bug about this, it was immediately dismissed without comment.)
Excuse me for asking... exactly what the hell are they argueing about? The only units that record digital content are directv receivers. And the DTivo's will never have ethernet (HMO) enabled -- DTV explicitly stated the USB ports (and thus ethernet) are not to be enabled on DTivos. (No S1 unit will ever have the Home Media Option, and thus sharing ability, available to it, but does support an ethernet card.)
That's right, the only systems capable of sharing recorded content recorded that content off air -- standard definition, OTA... none of this FCC broadcast flag bullshit even applies.
A few other errors... no US model Tivo has digital outputs. Only DTV models record digitally broadcast data. The others digitized an analog broadcast. Sharing of recorded content will/can not be in real-time -- the item will have to be recorded completely before being transmitted at speeds 10-20x (SD) or 100x (HD) slower than the recorded bit rate, based on today's broadband technology. It is unlikely people will have home broadband connections capable of streaming HD content in realtime in the coming decades.
Except the mail isn't exactly realtime. You'll have to record the entire game, burn to DVD, and then mail it out. In the end, the DVD will not be viewed for several days after the game was aired.
Actually, you'd be surprised how closely they pay attention to these sorts of things. In order to receive sports packages, the unit must be connected to a phone line. They can (and do) verify the origin of the calls... and they will see this "LA" box calling from Oakland.
They've ruined NFL Sunday Ticket (tm) anyway... if any local station, even those DTV doesn't carry, claims to be carrying a game, it gets blacked out to force you to want the local broadcast station. In my experience over the past few years, DTV has been wrong 90% of the time. (And the local station guide data is correctly listing the right game.) And that damned "enhanced screen" is enough to get me to cancel my entire DTV.
(And it's DTV who have a case for fraud and possibly breach on contract.)
It's a DIRECTV Tivo. They, technically, never need to make any calls. The guide data comes from the dish as does a number of other things. Software updates and stats use the phone line, but really never have to happen :-)
And what makes you think the monkey loading the vending machine will not notice a plastic coke can? This is exactly why you will not find a promo can in a vending machine -- the human loading the machine will notice the abnormal "can".
And just why the hell would you be trying to carry a case of Coke through the airport? It's instantly obvious the promo coke's are not a real can of coke, so knowingly carting the thing around in places where you know better (or should know better) will get you exactly what you deserve.
DHCP is a superset of BOOTP. So, DHCP can do everything BOOTP can (and more.) And that means just about everyone is running a bootp server -- most dhcp servers will answer a bootp request.
The BOINC presentation from Madrid claims folding@home is using BOINC. Unless they've recoded the clients without telling anyone, this is incorrect. Folding@home uses Cosm as it's communications framework. There is no reference to BOINC in any of the documentations I've seen from f@h.
The tracker can be run on any port. Likewise, the clients can (and do) use any port. Many (and I mean a lot) of ISPs around the world rate limit BT traffic on the "well known ports". The only reliable way to get accurate numbers on bittorrent use is to decode the traffic -- peek at the actual content of each packet look for the BT protocol.
.torrent, it's a little difficult to identify what's being traded without piecing together all the data.)
Kazaa works much the same way.
(Without the actual
I had a "wtf" expression as soon as I read it... knowing just how "diverse" tje Akamai network is. (distributed? yes, but not diverse.)
Moderators, please correct the lead-in... BIND and the global DNS system is what has the diversity. The problem with Akamai was their lack of diversity on top of their proprietary hacks to DNS.
What's to stop you from doing that to a normal, non-promo, can?
I think they're being a little too paranoid. It's not like you won't notice it's not a real can of coke the instant it's in your hand. By that reasoning, it's very unlikely you'll find one of those cans in a drink machine. (if you were loading the drink machine, would you put it there? Or keep it for yourself/a friend?)
The chopper is "for illustrative purposes only". Read the small print on the screen during the commercial. Remember the Pepsi Stuff (tm) promotion showing a harrier? Pepsi didn't have any such disclaimer and some kid collected enough points...
I have no idea what's acutally inside the can. But I'm betting it's not your average nokia cellphone and an earthmate gps. It's most likely a cellular modem -- read: NO MICROPHONE -- and a dirt cheap GPS. Given the average shelf-life of a cell phone battery's charge, I really doubt it's a full cell phone.
Actually, above a few hundred peers in any torrent swarm, the performance drops off greatly. Can your BT client maintain connections and state for 38,000 connections? Most clients limit the number of peers per torrent to 100.
- saturate bandwidth on the receiving sides
Actually, it tends to saturate bandwidth on all sides and right down the middle, too- Up until now, it's just people trading files because they like trading files fast.
It's a pissing a contest... who can transfer the file the fastest. They don't care what they are moving or who they may be hurting in the process. I sat and watched these idiots once... a dozen "people" all pushing the same file(s) to the same server; the first one to get the entire file there "wins". This is both stupid and wasteful.- Often it'll hit newsgroups before it gets made into a torrent somewhere
For some things, yes. But for more popular items, a torrent will ususally be available before it appears in part or whole on USENET. (I know, because I watch... and, sadly, usenet is faster than downloading via bittorrent. esp. from such places as suprnova)Nice theory, but it falls flat when you've ordered the drives directly from Maxtor -- and they arrive with a customs form on them. Face it, quality and reliability circled the drain some time ago... Make 'em cheap and make 'em quick. It's better on the bottom line to get people to buy new drives every six months instead of using the same one for 5 years. (now we're back to one of the reasons SCSI drives are so expensive.)
Assuming a human loads them in the box, that's the only place a person handles the drives (already sealed in their static bag) before they come out of the case. I'm reasonablly sure humans are involved in loading the retail boxes. But I could be wrong -- people are expensive, even in Thailand.
Yeah... Seagate instead of Maxtor. ('tho I'm gonna give WD a go. They're good enough for Tivo...)
I don't buy drives on ebay. The 80G drives were by the case (it takes a few of them to amass 1TB :-)) The 250G ones are from our system builders (assmeblers, whatever) who buy them by the case.
Even the 80G Maxtor in my Dell workstation (8300, 9 months old) failed. Dell replaced it with a WD drive. That speaks volumes to me.
Read the snippet about Command Queuing. That takes a good bit of memory on the drive controller to handle well. As Seagate makes some serious SCSI drives, I think they'll make good use of 16MB.
See Also: eBay
I don't know if you can find LTO gear on the cheap (haven't looked), but DDS and AIT gear is readily available.
If they are both the same make and model from the same manufacturer, they are both equally statistically likely to fail. Odds are, they will not both fail at the same time. However, they are highly likely to both fail within the same time period. When they were built makes little difference as running time is a greater contributor to failure than shelf age. I had two 160G drives (same make/model) built nearly a year apart in different factories (Maylasia and Thailand) installed in the same system (same running history -- RAID0 so almost identical access even) die within two days of each other.
Therefore, when one fails, replace both of them. Or better yet, use two different drives... say a Hitachi and a Seagate.
IDE drives have the shittiest reliability. And it's getting worse, not better. Several years ago, in my experience building fileservers, Maxtor had a failure rate of 25%. Right. Out. Of. The. Box. And that was for 80G drives that they'd been making for several years at that point. Today, they have a 60% failure rate over the first week (about the same OOB.) And this is with their "top of the line, enterprise class" crap. You'd think with a 3yr warantee, they'd spin the damn thing up at least once before shoveling 20 of them in a box.
:-)) which Seagate replaced. Another half dozen developed "stiction" problems after several years and needed a little help to get spun back up.]
SCSI is the only way to get drives that have actually been through any testing. Each drive is individually tested; however, with IDE, only a small sample of drives are tested. This is one reason SCSI is more expensive. But demand, perception, and the money in the enterprise market place also factor into the cost... a 140G SCSI drive does NOT cost 1000$ to build and test. They use the exact same servo hardware as their IDE "white trash" cousins (in many cases -- 10k and 15k speeds aside.)
[Disclaimer: I don't have as much experience with Seagate's IDE (PATA or SATA) lineup. But I can say, I've never had any Seagate drive, SCSI or IDE, fail right out of the box or shortly there after. Of the few that have failed, 2 overheated and melted their logic boards (temp. swapped with another drive to fetch the data