Too bad he's usually talking out his ass. Tho' in this case, he's pretty much right. The industry hasn't released content to the net because they know people will just steal it. Enough theft occurs already with the relatively difficult processes (you need a number of "hacker tools" to rip DVDs, copy stuff off your DTivo, or trick your Replay 4000 into sending content to your PC) They aren't going to release anything until they have some means of keeping their fingers on their content. However, they have no incentive to develop any meaningful technology to prevent copying (which can be proven to be impossible or at minimum highly complex and very expensive) unless they can make circumvention punishable by immediate death.
The current set of laws are sufficiently laughable. It's already (and long since been) illegal to duplicate a copyrighted work weither that be photocopying a magazine or molecular duplication of a CD. Now we have a set of laws that make it legal to break the law.
Everyone seems to have long since missed the point: If you make something that's work buying, people will buy it.
Yes, mindspring has been doing this for a long time. At one point, they had the dialin users "trapped" -- port 25 always goes to their mail server no matter who you tried to connect to.
Yes, you can claim to be whomever you choose, but mindspring/eartlink will be able to tell who actually sent the message by simply looking at their logs. There they will find the IP of the origination point and thus YOU. If you connect directly to some server on korea that doesn't log anything or add any header listing where the message came from, then there's no way to tell who's responsible.
My problem is that, if this has been "known" for almost a year, why hasn't anyone fixed it yet? Cisco, who normally is very proactive in fixing things that let people crash your expensive routers and switches, only recently published an alert and started officially releasing patched software yesturday (2/12/2002)
Yes, SNMP can be very insecure. However, for many of us, it's an inescapable evil.
Actually, they aren't that bad if people actually obeyed the original charter:.com is for COMpanies,.net for NETwork providers,.org for non-profit ORGanizations.
One cannot rely completely on "first come" or trademarks or copyright. Domain disputes are complex problems. The bigest problem is all the lies and posturing associated with almost all disputes. In this case, he got there first and without incident for *12* years.
Actually, unicomsi has been around for ~20 years. They claim to have registered the trademark on UNICOM (and attached thousands of circled "R"s on their web paged) in 1989. The domain was registered in Feb of 1990. Now, I would submit, Mr. Rosenthal had no knowledge of the existance of the "international powerhouse" of unicomsi (not being a Fortune 100 company and all.)
Of course, their failure to "protect their mark" for 12 years is extremely damning evidence. I say let him keep his domain name AND nullify the trademark as well *grin*
Just a guess, but the IDE drives were in desktops that were not running under any measurable load 24/7. And the SCSI drives were old in comparison.
I've had both SCSI and IDE drives die in various inventive ways. IDE has proven to be far less reliable (tho' very much so cheaper)...
I've never had a SCSI drive be defective right out of the box. However, 4 out of 16 80G IDE drives for a 1TB array were defective right out of the retail box. Two of their factory replacements were defective.
I've never had a new SCSI drive fail within hours or days of going into service. I've got new Maxtor IDE drive to RMA that failed *4 hours* after being put in service.
Over the last decade, I've replaced seven SCSI drives -- 4 completely dead, 1 field repairable (low level format), and 2 perfectly functional tho' noisy as hell. Over the same decade, there have been dozens of defective IDE drives thrown away -- thrown heads, bad sectors in random locations, bad sectors in the partition area, drives that can no longer track, drives that no longer spin up, drives that no longer spin at a stable speed, and (my personal favorite) the drive that works for exactly three days.
There's no secret, I like SCSI. It's a perfectly clean, extensible protocol and has been from day one. IDE is a hideous, horrible stack of kludges. Over the past few years, there have been more and more crusty band-aides strapped on to give it more and more of the features and performance SCSI has enjoyed for years. IDE is cheap in more ways than just price.
They aren't building the mother of all SANs where storage integrity and thus drive lifespan are important. It might be important for the USENET archive, but their data came from somewhere so it's not that important either.
Cheap is a very powerful motivator in modern economics.
Not to start a SCSI vs. IDE war... I don't think Google is crazy (read: stupid) enough to bank their business on the cheap-ass commodity IDE drives that last a year or so. It doesn't matter that Maxtor will replace the PoS for three years; you still have to replace it and rebuild or recollect the data.
As for speed and cost, SDRAM is about 6 orders of magnitude faster than even that fastest hard drives -- nanosecond vs. millisecond. Costwise, he's smokin' some mighty weed; SDRAM is 2 orders of magnitude more expensive than the most expensive SCSI drive in the world (the 180G Seagate Chetah.) RAM is about 30 cents per meg and hard drives range from.2 (IDE) cents to.8 (SCSI) cents per meg. In the end, one needs boatloads of both.
Funny! I've seen crazy stuff like this all over the place. The UPS has to be inside the computer to be safe. No ammount of redundency can protect against the power switch...
Micah powered down the NetApp feeding the web servers at Interpath years ago... He turned it back on immidiately *grin* Just days earlier I had been talking to Network Appliance about the hair trigger nature of that switch -- a big, black switch in the center of the front of the box (an F210, btw.) I taped an angle bracket over it the next morning.
Even logner ago at Interpath, we also had a modem bank (Microcom HDMS/fast -- huge power hungry PoS) powered down by a cleaning lady's mop once.
After Interpath moved the office, we had half the server room powered down by a short in one of the cubes. It was on a completely different circuit, however, the breakers are magneticly thrown. When that one went, it popped the breakers next to it as well. All totaled, I think there were six circuits cut off. Even the E450 powered on two different (UPS backed) circuits got cut off.
I had a network switch powered down by a vacuum cleaner when I was at Make Systems. My workstation housed the clearcase view for a patch or optimized build (I forget which.) I replaced it with a different swtich that didn't have an on/off switch.
A full USENET news feed (everything one can find) will exceed 120GB per day. It'll almost fill a DS3. (And we were receiving a "crappy" test feed from UUNet.) So, minus @alt.binaries.*, one could mirror USENET for a few years. With the binaries, it'll hold you for about a week, 2 at the most.
Two things... "off site" and "fire proof safe". Even Interpath wasn't stupid enough to keep all the data in one place. There were three sets of tapes. One set (two months old) were off site (at someone's house in a fire box.) One set (last month's full backups) were in the fire proof safe in the basement. And the current backup set was at my desk in two fire boxes. When the alarm goes off, you grab those two boxes. (At the time, someone was always there.)
While I agree, SCSI drives are simply better drives, cheap is a very powerful motivator. IDE is about one tenth the cost of SCSI. So the IDE array will last a year -- how long do a lot of companies last these days? Over time, the SCSI system may, ultimately, be cheaper -- the cost of replacing failed drives, the downtime for rebuilding and restoring the array, lost productivity of a missing database, drugs for the admin headaches...
I've built a 1.04TB array. It's an impressive hack of a system. Out of the 16 drives for the array, four (4) were defective right out of the box! And two of those replacements were suspect. After a month of handling a full news feed (120G+ per day) we've worked most of the kinks out of it (I don't recommend w2k for a drive array.)
BTW: I used a pair of 3ware Escalade (6800) controllers. They take alot of the suckiness out of IDE (tho' it's a cabling mess.)
How the hell can you blame the ISPs? Their job is to deliver packets. You aren't paying them to be a firewall, intrusion detection system, or "lameness filter". You have a packet that needs to get to some other host; it's handed to the ISP and expected to get there. ISPs that block or filter traffic receive conciderable negative feedback.
Yes, there are things ISPs can do better. There are things that can be done to reduce the impact of stupidity. However, the landscape is constantly changing and I've yet to meet an ISP employee who gives much of a damn about filling all the cracks -- and even fewer who know how. (at best a bandaid is placed over problems when they become serious.)
I have a DTivo and it's never recorded the wrong thing in a "zim" slot. It's also always had episode descriptions. I suggest you yell at TMS and Tivo about the SA having bad/stale guide data. (I highly recommend the DTivo.)
As for the sliding time slots, I used to be rather pissed at having to record two hours of bull to get a complete Anamaniacs episode. Lately they've not been so fscked up. Pinky and the Brain is a different story... butted up against the overnight Brady Bunch marathon (aka: hell) they show one skit to fill out to the normal daily schedule. The end result is up to 20mintues of crap before the actual P&B skit. (And what's with the asses plastering the damned Nick logo over everything they have? I'm surprised the logo isn't inside Pinky's head during the opening.)
I keep a shrine to spam... it's over 12M at present. One of these days, I'll hook that folder up to the web so people can see all the stupid bullshit in the world. For a while, I got no spam -- see, MAPS and RBL (and others) do work.
Dude, never believe what you hear from spammers. (They lie, you know.) Don't even trust the headers. Although, the headers are often good for a laugh.
PINE 4.33 MESSAGE INDEX spam Msg 2,130 of 2,130
2121 Jan 12 1a7info1@iol.it (7,325) Save On Your Life Insurance -FREE Qu
2122 Jan 11 j89b4m4gt@hotmail. (2,373) Settle your tax debt for pennies on
+ 2123 Jan 12 Information (4,338) Get the cash you need to pay off you
2124 Jan 12 dvdcopy102@yahoo.c (2,493) Copy DVD's with a Regular CD Writer!
2125 Jan 13 pesavento@bass.se (2,971) Burn 36% more calories, block 30% of
2126 Jan 13 1a7info1@altavista (3,042)::: NEW Universal Analog/Digital TV
2127 Jan 13 888marketing@371.n (2,074) Legal Services For Only Pennies a Da
2128 Jan 13 villaents2002@xwlf (6,546) Need Extra Money Over The Holidays?
2129 Jan 13 irene@m-ul.com (1,760) ADV: CONGRATULATIONS!! YOU WON!!
2130 Jan 13 folgermark@hotmail (2,298) ASSET & BACKGROUND CHECKS!!...WE CHE
(Note to slashcode idiots: why is there still no <pre> tag?!)
Actually, SPAM(tm) is rather expensive stuff. Might I suggest the cheaper (and much less tasty Treet(tm) meat-like product.)
PS: Seeing all the "smokers" eating cans of "Smeat" was the funniest part of WaterWorld.
PPS: A can of SPAM(tm) was seen on an episode of Outward Bound (in the rocky mtns) a few weeks ago. They didn't actually show the logo, but I know what SPAM(tm) looks like.
"free speech" That's funny. Advertising is neither "free speech" nor "free". One must pay for radio and tv spots, magazine and newspaper ads, newspaper inserts, billboards, sky writers, and all that junk that collects in your US Postal mailbox. Advertising has never been fucking free.
As for "free speech"... that's laugh-in-your-face stupid! Perhaps they should begin lobying to allow cig. and booze ads on TV. I'd love to see p0rn on interstate billboards as well while their at it:-)
Slashdot Mirror
Too bad he's usually talking out his ass. Tho' in this case, he's pretty much right. The industry hasn't released content to the net because they know people will just steal it. Enough theft occurs already with the relatively difficult processes (you need a number of "hacker tools" to rip DVDs, copy stuff off your DTivo, or trick your Replay 4000 into sending content to your PC) They aren't going to release anything until they have some means of keeping their fingers on their content. However, they have no incentive to develop any meaningful technology to prevent copying (which can be proven to be impossible or at minimum highly complex and very expensive) unless they can make circumvention punishable by immediate death.
The current set of laws are sufficiently laughable. It's already (and long since been) illegal to duplicate a copyrighted work weither that be photocopying a magazine or molecular duplication of a CD. Now we have a set of laws that make it legal to break the law.
Everyone seems to have long since missed the point: If you make something that's work buying, people will buy it.
Yes, mindspring has been doing this for a long time. At one point, they had the dialin users "trapped" -- port 25 always goes to their mail server no matter who you tried to connect to.
Yes, you can claim to be whomever you choose, but mindspring/eartlink will be able to tell who actually sent the message by simply looking at their logs. There they will find the IP of the origination point and thus YOU. If you connect directly to some server on korea that doesn't log anything or add any header listing where the message came from, then there's no way to tell who's responsible.
My problem is that, if this has been "known" for almost a year, why hasn't anyone fixed it yet? Cisco, who normally is very proactive in fixing things that let people crash your expensive routers and switches, only recently published an alert and started officially releasing patched software yesturday (2/12/2002)
Yes, SNMP can be very insecure. However, for many of us, it's an inescapable evil.
Actually, they aren't that bad if people actually obeyed the original charter: .com is for COMpanies, .net for NETwork providers, .org for non-profit ORGanizations.
One cannot rely completely on "first come" or trademarks or copyright. Domain disputes are complex problems. The bigest problem is all the lies and posturing associated with almost all disputes. In this case, he got there first and without incident for *12* years.
Actually, unicomsi has been around for ~20 years. They claim to have registered the trademark on UNICOM (and attached thousands of circled "R"s on their web paged) in 1989. The domain was registered in Feb of 1990. Now, I would submit, Mr. Rosenthal had no knowledge of the existance of the "international powerhouse" of unicomsi (not being a Fortune 100 company and all.)
Of course, their failure to "protect their mark" for 12 years is extremely damning evidence. I say let him keep his domain name AND nullify the trademark as well *grin*
I've had both SCSI and IDE drives die in various inventive ways. IDE has proven to be far less reliable (tho' very much so cheaper)
I've never had a SCSI drive be defective right out of the box. However, 4 out of 16 80G IDE drives for a 1TB array were defective right out of the retail box. Two of their factory replacements were defective.
I've never had a new SCSI drive fail within hours or days of going into service. I've got new Maxtor IDE drive to RMA that failed *4 hours* after being put in service.
Over the last decade, I've replaced seven SCSI drives -- 4 completely dead, 1 field repairable (low level format), and 2 perfectly functional tho' noisy as hell. Over the same decade, there have been dozens of defective IDE drives thrown away -- thrown heads, bad sectors in random locations, bad sectors in the partition area, drives that can no longer track, drives that no longer spin up, drives that no longer spin at a stable speed, and (my personal favorite) the drive that works for exactly three days.
There's no secret, I like SCSI. It's a perfectly clean, extensible protocol and has been from day one. IDE is a hideous, horrible stack of kludges. Over the past few years, there have been more and more crusty band-aides strapped on to give it more and more of the features and performance SCSI has enjoyed for years. IDE is cheap in more ways than just price.
Translation: they are banking on cheap NODES.
They aren't building the mother of all SANs where storage integrity and thus drive lifespan are important. It might be important for the USENET archive, but their data came from somewhere so it's not that important either.
Cheap is a very powerful motivator in modern economics.
s/Win2k/NT/ (That option has been around for a long time.)
The DisablePagingExecutive key doesn't do a whole lot. Most of the kernel is still paged out.
HD seek time: 4.9 MILLIseconds
SDRAM "seek" time: 5 NANOseconds
Therefore RAM is approx. 1 MILLION times faster. (DDR SDRAM is even faster.)
Not to start a SCSI vs. IDE war... I don't think Google is crazy (read: stupid) enough to bank their business on the cheap-ass commodity IDE drives that last a year or so. It doesn't matter that Maxtor will replace the PoS for three years; you still have to replace it and rebuild or recollect the data.
.2 (IDE) cents to .8 (SCSI) cents per meg. In the end, one needs boatloads of both.
As for speed and cost, SDRAM is about 6 orders of magnitude faster than even that fastest hard drives -- nanosecond vs. millisecond. Costwise, he's smokin' some mighty weed; SDRAM is 2 orders of magnitude more expensive than the most expensive SCSI drive in the world (the 180G Seagate Chetah.) RAM is about 30 cents per meg and hard drives range from
Funny! I've seen crazy stuff like this all over the place. The UPS has to be inside the computer to be safe. No ammount of redundency can protect against the power switch...
Micah powered down the NetApp feeding the web servers at Interpath years ago... He turned it back on immidiately *grin* Just days earlier I had been talking to Network Appliance about the hair trigger nature of that switch -- a big, black switch in the center of the front of the box (an F210, btw.) I taped an angle bracket over it the next morning.
Even logner ago at Interpath, we also had a modem bank (Microcom HDMS/fast -- huge power hungry PoS) powered down by a cleaning lady's mop once.
After Interpath moved the office, we had half the server room powered down by a short in one of the cubes. It was on a completely different circuit, however, the breakers are magneticly thrown. When that one went, it popped the breakers next to it as well. All totaled, I think there were six circuits cut off. Even the E450 powered on two different (UPS backed) circuits got cut off.
I had a network switch powered down by a vacuum cleaner when I was at Make Systems. My workstation housed the clearcase view for a patch or optimized build (I forget which.) I replaced it with a different swtich that didn't have an on/off switch.
Actually, accorinding to RealPlayer, it's being streamed via http.
C:\>netstat
Active Connections
Proto Local Address Foreign Address State
TCP moose:1272 210.59.224.72:http ESTABLISHED
A full USENET news feed (everything one can find) will exceed 120GB per day. It'll almost fill a DS3. (And we were receiving a "crappy" test feed from UUNet.) So, minus @alt.binaries.*, one could mirror USENET for a few years. With the binaries, it'll hold you for about a week, 2 at the most.
Two things... "off site" and "fire proof safe". Even Interpath wasn't stupid enough to keep all the data in one place. There were three sets of tapes. One set (two months old) were off site (at someone's house in a fire box.) One set (last month's full backups) were in the fire proof safe in the basement. And the current backup set was at my desk in two fire boxes. When the alarm goes off, you grab those two boxes. (At the time, someone was always there.)
While I agree, SCSI drives are simply better drives, cheap is a very powerful motivator. IDE is about one tenth the cost of SCSI. So the IDE array will last a year -- how long do a lot of companies last these days? Over time, the SCSI system may, ultimately, be cheaper -- the cost of replacing failed drives, the downtime for rebuilding and restoring the array, lost productivity of a missing database, drugs for the admin headaches...
I've built a 1.04TB array. It's an impressive hack of a system. Out of the 16 drives for the array, four (4) were defective right out of the box! And two of those replacements were suspect. After a month of handling a full news feed (120G+ per day) we've worked most of the kinks out of it (I don't recommend w2k for a drive array.)
BTW: I used a pair of 3ware Escalade (6800) controllers. They take alot of the suckiness out of IDE (tho' it's a cabling mess.)
Wow, that's so funny it had me in tears.
WHAT!?
How the hell can you blame the ISPs? Their job is to deliver packets. You aren't paying them to be a firewall, intrusion detection system, or "lameness filter". You have a packet that needs to get to some other host; it's handed to the ISP and expected to get there. ISPs that block or filter traffic receive conciderable negative feedback.
Yes, there are things ISPs can do better. There are things that can be done to reduce the impact of stupidity. However, the landscape is constantly changing and I've yet to meet an ISP employee who gives much of a damn about filling all the cracks -- and even fewer who know how. (at best a bandaid is placed over problems when they become serious.)
On a related note, VH1 is showing Ren and Stimpy!
I have a DTivo and it's never recorded the wrong thing in a "zim" slot. It's also always had episode descriptions. I suggest you yell at TMS and Tivo about the SA having bad/stale guide data. (I highly recommend the DTivo.)
As for the sliding time slots, I used to be rather pissed at having to record two hours of bull to get a complete Anamaniacs episode. Lately they've not been so fscked up. Pinky and the Brain is a different story... butted up against the overnight Brady Bunch marathon (aka: hell) they show one skit to fill out to the normal daily schedule. The end result is up to 20mintues of crap before the actual P&B skit. (And what's with the asses plastering the damned Nick logo over everything they have? I'm surprised the logo isn't inside Pinky's head during the opening.)
I think Sprint has a fully meshed multicast network. Any Sprint customer can request to join the multicast "grid".
As for the broadcasting of source code, you'd have to use some form of forward error correction (FEC) similiar to what is used in sat. broadcasting.
Set the ftp server to serve the tarball at 16k/s and that patch at unrestictive speeds. Problem solved.
Dude, never believe what you hear from spammers. (They lie, you know.) Don't even trust the headers. Although, the headers are often good for a laugh.
(Note to slashcode idiots: why is there still no <pre> tag?!)
And then what? Put 'em back and do it again?
Actually, SPAM(tm) is rather expensive stuff. Might I suggest the cheaper (and much less tasty Treet(tm) meat-like product.)
PS: Seeing all the "smokers" eating cans of "Smeat" was the funniest part of WaterWorld.
PPS: A can of SPAM(tm) was seen on an episode of Outward Bound (in the rocky mtns) a few weeks ago. They didn't actually show the logo, but I know what SPAM(tm) looks like.
AMEN!
:-)
"free speech" That's funny. Advertising is neither "free speech" nor "free". One must pay for radio and tv spots, magazine and newspaper ads, newspaper inserts, billboards, sky writers, and all that junk that collects in your US Postal mailbox. Advertising has never been fucking free.
As for "free speech"... that's laugh-in-your-face stupid! Perhaps they should begin lobying to allow cig. and booze ads on TV. I'd love to see p0rn on interstate billboards as well while their at it