When I toured their previous cluster, these guys completely touted the benefits of low latency. They didn't really care about the bandwidth -- at least it wasn't one of their major concerns. In fact, one of the guys (Tim) had done major work on a Parallel port connection device that he said had extremely low latency. However, it was limited in how many nodes it could handle, and so I believe that they went with ethernet to take more nodes. Otherwise, this thing might have been connected by the old parallel ports!
In fairness to them, their Air Conditioning is so strong that very little of their exhausted heat is making it back in to the systems. They've probably got several thousand CFM of cold air pumping down on these things.
Another thing to keep in mind with their setup is that they are in a University building, and I don't think they have to pay the utilities. Why worry about power issues when the University will take care of it for "free"?
I toured the previous cluster these guys did (KLAT2) and was very impressed. However, using AMD Athlon Thunderbirds last time, it did get quite hot. I remember standing by the cluster looking at all the wiring and being bombarded by an overhead cooling vent. I'm also assuming that these cooling issues is the reason that each case has two blow-holes.
I'd also like to see these guys post in-depth specs of each machine. Being a hardware nut, I'd like to see how they got so many machines so cheap, and maybe even what vender they used. As I remember, they worked REALLY hard on their last cluster to keep costs to an absolute minimum.
Actually, you're exactly wrong in that there are NO multiple paths between any of the hosts -- just one path between each. So what has already been responded is right -- it's all about single-hop low latency results.
"The lawsuits accuse the defendants collectively of flooding Microsoft's computer systems and its customers with more than 2 billion deceptive unsolicited e-mail messages."
2,000,000,000 / 15 = 133,333,333
133 MILLION (and change) spams per spammer on average?
Alan Ralsky can get out a billion per day.... so exactly how long was Microsoft monitoring the spam? It seems like they went after some "small fries" in the Spam world.
As they say, "God is in the Details." "God" here being Bill Gates, of course.
While you may be right about the volume, I think it's a bit misleading to say "twice the size" when it's only twice the volume. (Though, thank you for pointing out that your calculation was for volume.)
Consider this:
1.3*1.3*1.3=2.197
So a 30% increase in each of the dimensions could more than double the volume of an object.
As for the increased mass, just consider that it is probably carrying around double the platters of the iPod, and an internal disk drive enclosure to hold them.
All in all, I'd still call it only "slightly larger," and I'd definately deal with the bulk for the added benefits of storage and price.
I'm not exactly the type you requested comment from -- I have yet to use Ardour -- but I have used Audacity (and looked at Ardour heavily). In my eyes, having used several audio programs, the intent seems different. Each one is produced for a "target audience", if you will.
Now Audacity is a decent multi-tracking program for beginners. However, there are many limitations to the control users innately have. For instance, the compressor in the latest release has absolutely no settings -- it just compresses according to presets written into the program. So, while your audio won't be clipping like an uncompressed signal would, but you also get zero control of the particularities of the sound that is produced. From my perspective, having mixed several multi-tracks in my day (most including vocals, drums, guitars, bass, and more), the more control I have the better.... and that's where Audacity falls short.... at least until you start downloading additional plug-ins.
Ardour, on the other hand, looks to be a much more robust design out of the box. It seems intended for more professional usage. From what I've seen, it's got excellent control of just about every processor you could normally want in a studio recording app, and an interface that provides fast and comprehensive control to them all. It's almost like sitting at a real studio mixing board -- the same layout. However, its main drawback looks to be in its learning curve. Most people who just dabble with sound now and again will probably be lost in the complexity that the program offers. However, for my money (or lack thereof), it looks to be much closer to my needs than Audacity.
So in essence, it's all about what you want to do with it and the complexity that you're willing to endure.
I've experimented with Linux and other open-source OSes for a while now, but I've never found one that can make a permanent home on my desktop. One major reason: No good multi-track recorder! Sure, there are options like Audacity, but there is nothing that can rival the mixing options of, say, Cool Edit Pro. For my mixing, I absolutely NEED features that I can control -- a good compressor, gain control, some decent effects, a reliable GUI, etc. Now, it looks like Ardour may offer that, and so I rejoice. This may be the last rail laid on my transitional track, and the golden spike is beautiful. So Ardour, I say bring it on. Lemme see what you got!
I've used both Cakewalk and Cool Edit Pro for some recording work, and I believe that both programs are a bit lacking, each in its own way. However, I think the market is getting better, as proved by ProTools. Just today, I found a new program, Ardour. While it only runs on Linux and has yet to be fully relased with a stable version, it looks like it has a lot of promise (and will be absolutely free)! It has 3 years of development under its belt, and the features list is quite extensive. So for the price of a little Linux-learning, ProTools may soon be beatable.
Having never actually read any spam logs, your post made me think. I never really considered the load on servers and/or internet connections. However, by your numbers:
3102kB of transfer was generated by 2454 messages. 3102/2454=1.264kB per message
3102kB of transfer was generated in a system that delivered 185 messages. 3102/185=16.768kB per message delivered.
16.768/1.264=13.266 times the bandwidth in a spam-full world.
So it would appear that my ISP is feeling the pain from delivering this spam almost as much as I am from getting it.
Oh well.... if they're ever taking too big a financial hit from the excess utilization, they have a lot of offers to get refinanced (the 5th lawsuit filed in AOL's recent courtroom antics, according to this article).
This is an interesting idea, but probably not feasible with 802.11b using the high recording settings. As someone else pointed out in another thread, 24bits x 192 KHz x 2 channels = 9216 Kbits/sec = 9.216 Mbps. While 802.11b can technically go to 11 Mbps, it rarely reaches it. Just like standard 10Mbps and 100Mbps ethernet, the rated speed is merely theoretical. In actuality, 802.11b often gets only 7Mbps. That would simply be too slow to support the full rate of this device. If this idea was used with a faster connection, such as 100Mbps ethernet, it could be feasible.
Now, regarding the streaming media server, the techie in me wants to applaud the "free the music" idea. However, the recording engineer in me wants to smack you for suggesting that my expensive new 24 bit at 192KHz mix be compressed (especially given the oh-so-natural sounding Real Audio history that has permiated the industry). But it does seem possible if one invested the money in a seperate streaming media server, and a fat pipe to stream it.
But my main reponse to this is these solutions it that they are expensive. Many commercial units out now, like this one, come with 20GB+ integrated hard drives, among convenient options like integrated CD burners and digital effects. If one isn't going to pay to go into a studio, it's probably cheaper to buy some mid-line recording equipment than to modify a PDA for "professional quality recording".
Got a link to support this? It's good info, but the one linke you provided doesn't seem to be working at the moment. Also, is the compression lossy? Sony quotes the frequency on their minidisc players as 20-20,000 Hz -- what is often considered the full range of human hearing. That would lead me to believe that the compression, however it is being done, is not hurting the high-end response like so many lossy compressions do.
Perhaps I should rephrase: "minidiscs can technically be CD quality." While it is true that most Minidisc players today can utilize compression technology, particularly the different versions of Sony's ATRAC compression, virtually all of them can also record without compression. The compression factor is how Sony now claims that you can fit "Up To 320 Minutes On One 80 Minute Disc." But if the discs are recorded without compression, Minidisc have a sample rate of 16 bits at 44.1 kHz -- the same as a normal CD.
For a long time, live recording has been consigned mostly to the realm of DAT recorders, Minidisc recorders, or laptop computers. On one hand you have subpar sound quality, on the other you have a bulky rig with a big 'steal me' sign attached.
Subpar sound quality? The DATs that I've worked with have better resolution than CDs (48 KHz vs 44.1 KHz sampling), minidiscs are technically CD quality, and laptop computers can be equally sensitive given the right equipment. Given what I've heard of PDA sound, there's nothing subpar about the existing recording mediums. Also, it's hard to claim that a minidisc is "a bulky rig".
By using a hardware card that allows recording via S/PDIF onto Compact Flash, you will be able to use your iPaq or Zaurus alongside a decent A/D converter to portably get field recordings at up to 24bit/192kHz.
So to record in this way, I must buy a "decent" A/D converter and a bunch of Compact Flash. And, unless they are using some compression which will lower the sound quality, this thing will suck up more MB-per-minute of audio than a CD. Good thing Hitchai (formerly IBM) makes their MicroDrive, and I have a money tree in my yard.
So, bottom line as I see it? An interesting project, but one which uses expensive hardware and media that makes it prohibitively expensive. So if you want professional digital recording, get a professional digital recorder. If you want ad-hoc "pro" sound recording from a PDA, now you've got an option.
He thinks the killer app for this one is for keeping your porn storage hidden, if you're busted by the cops.
Freudian Slip - Definition:
'An inadvertent mistake in speech or writing that is thought to reveal a person's unconscious motives, wishes, or attitudes.'
- Courtesy of The Random House College Dictionary, 1980
The site www.tonerrefillkits.com isn't working for me, though the WHOIS info says that it's been registered since 1999 (through 2004). Are you sure that this wonder-website isn't another dot-com bust?
You CANNOT fit a 100 foot dish to a satellite and orbit it around Mars or Jupiter, etc, to pick up signals from further out and relay them to earth. Its simply not possible.
Perhaps this is only not possible by conventional thinking, if even then. But picture the Space Elevator hoisting up a large dish, or several small pieces of it. Beyond that, if we can launch an International Space Station that weighs 393,733 lbs, then even a dish factory in space is fathomable. ("Break your satellite? Need new parts? Just order on the universe wide web at uww.spaceparts.com!")
I'm glad this came out on Slashdot: I never would have known otherwise. I have one of the afflicted units, and I'd rather not have a melted puddle of lead-acid powered backup get lose near my PC.
A funny side note: I also found my serial number on my Quality Assurance Test receipt. Some quality assurance, I guess.
Show me some research! While it is an interesting idea, I don't see it holding much truth. The author's principle is that waveforms are not uniform to the source, rather approximations. Therefore, the author believes that we may begin to expect such approximated waveforms rather than more exact waveforms. While this may be true, it is not physiological damage. Think of it in another medium. Just because computer monitors use an approximation of images -- that is, images are composed of relatively lossy blocks rather than exact lines and shading -- does not mean that one who uses a computer for extended periods becomes unable to notice the exact sharpness of real life images. The author's idea is also further disproved by the idea that people have been listening to altered signals for years now (think distorted guitar or compressed vocals) without documented hazard.
Personally, I don't buy what this author is dishing out. It has the stench of "anti-MP3." I'd start looking for the author's alterior motives.
After what Kazaa did to MusicCity Morphius (eg. forced them off their joint P2P network without warning), it serves them right! In a market that is based entirely on appeal (generally from reputation) and customer satisfaction, how can any P2P service try to play "hardball" with its "competition" and expect to come out ahead? I'm all for the Napster-esque services, open source, and variety, but I can't see how any P2P services can survive an RIAA onslaught without a glowing reputation, a strong userbase, and the respect of the P2P community. Kazaa, obviously, did not have all of that.
Slashdot Mirror
When I toured their previous cluster, these guys completely touted the benefits of low latency. They didn't really care about the bandwidth -- at least it wasn't one of their major concerns. In fact, one of the guys (Tim) had done major work on a Parallel port connection device that he said had extremely low latency. However, it was limited in how many nodes it could handle, and so I believe that they went with ethernet to take more nodes. Otherwise, this thing might have been connected by the old parallel ports!
In fairness to them, their Air Conditioning is so strong that very little of their exhausted heat is making it back in to the systems. They've probably got several thousand CFM of cold air pumping down on these things. Another thing to keep in mind with their setup is that they are in a University building, and I don't think they have to pay the utilities. Why worry about power issues when the University will take care of it for "free"?
I toured the previous cluster these guys did (KLAT2) and was very impressed. However, using AMD Athlon Thunderbirds last time, it did get quite hot. I remember standing by the cluster looking at all the wiring and being bombarded by an overhead cooling vent. I'm also assuming that these cooling issues is the reason that each case has two blow-holes. I'd also like to see these guys post in-depth specs of each machine. Being a hardware nut, I'd like to see how they got so many machines so cheap, and maybe even what vender they used. As I remember, they worked REALLY hard on their last cluster to keep costs to an absolute minimum.
Actually, you're exactly wrong in that there are NO multiple paths between any of the hosts -- just one path between each. So what has already been responded is right -- it's all about single-hop low latency results.
"The lawsuits accuse the defendants collectively of flooding Microsoft's computer systems and its customers with more than 2 billion deceptive unsolicited e-mail messages."
2,000,000,000 / 15 = 133,333,333
133 MILLION (and change) spams per spammer on average?
Alan Ralsky can get out a billion per day.... so exactly how long was Microsoft monitoring the spam? It seems like they went after some "small fries" in the Spam world.
As they say, "God is in the Details." "God" here being Bill Gates, of course.
While you may be right about the volume, I think it's a bit misleading to say "twice the size" when it's only twice the volume. (Though, thank you for pointing out that your calculation was for volume.)
Consider this:
1.3*1.3*1.3=2.197
So a 30% increase in each of the dimensions could more than double the volume of an object.
As for the increased mass, just consider that it is probably carrying around double the platters of the iPod, and an internal disk drive enclosure to hold them.
All in all, I'd still call it only "slightly larger," and I'd definately deal with the bulk for the added benefits of storage and price.
I'm not exactly the type you requested comment from -- I have yet to use Ardour -- but I have used Audacity (and looked at Ardour heavily). In my eyes, having used several audio programs, the intent seems different. Each one is produced for a "target audience", if you will.
Now Audacity is a decent multi-tracking program for beginners. However, there are many limitations to the control users innately have. For instance, the compressor in the latest release has absolutely no settings -- it just compresses according to presets written into the program. So, while your audio won't be clipping like an uncompressed signal would, but you also get zero control of the particularities of the sound that is produced. From my perspective, having mixed several multi-tracks in my day (most including vocals, drums, guitars, bass, and more), the more control I have the better.... and that's where Audacity falls short.... at least until you start downloading additional plug-ins.
Ardour, on the other hand, looks to be a much more robust design out of the box. It seems intended for more professional usage. From what I've seen, it's got excellent control of just about every processor you could normally want in a studio recording app, and an interface that provides fast and comprehensive control to them all. It's almost like sitting at a real studio mixing board -- the same layout. However, its main drawback looks to be in its learning curve. Most people who just dabble with sound now and again will probably be lost in the complexity that the program offers. However, for my money (or lack thereof), it looks to be much closer to my needs than Audacity.
So in essence, it's all about what you want to do with it and the complexity that you're willing to endure.
I've experimented with Linux and other open-source OSes for a while now, but I've never found one that can make a permanent home on my desktop. One major reason: No good multi-track recorder! Sure, there are options like Audacity, but there is nothing that can rival the mixing options of, say, Cool Edit Pro. For my mixing, I absolutely NEED features that I can control -- a good compressor, gain control, some decent effects, a reliable GUI, etc. Now, it looks like Ardour may offer that, and so I rejoice. This may be the last rail laid on my transitional track, and the golden spike is beautiful. So Ardour, I say bring it on. Lemme see what you got!
I've used both Cakewalk and Cool Edit Pro for some recording work, and I believe that both programs are a bit lacking, each in its own way. However, I think the market is getting better, as proved by ProTools. Just today, I found a new program, Ardour. While it only runs on Linux and has yet to be fully relased with a stable version, it looks like it has a lot of promise (and will be absolutely free)! It has 3 years of development under its belt, and the features list is quite extensive. So for the price of a little Linux-learning, ProTools may soon be beatable.
Having never actually read any spam logs, your post made me think. I never really considered the load on servers and/or internet connections. However, by your numbers:
3102kB of transfer was generated by 2454 messages. 3102/2454=1.264kB per message
3102kB of transfer was generated in a system that delivered 185 messages. 3102/185=16.768kB per message delivered.
16.768/1.264=13.266 times the bandwidth in a spam-full world.
So it would appear that my ISP is feeling the pain from delivering this spam almost as much as I am from getting it.
Oh well.... if they're ever taking too big a financial hit from the excess utilization, they have a lot of offers to get refinanced (the 5th lawsuit filed in AOL's recent courtroom antics, according to this article).
dupe, dope.
This is an interesting idea, but probably not feasible with 802.11b using the high recording settings. As someone else pointed out in another thread, 24bits x 192 KHz x 2 channels = 9216 Kbits/sec = 9.216 Mbps. While 802.11b can technically go to 11 Mbps, it rarely reaches it. Just like standard 10Mbps and 100Mbps ethernet, the rated speed is merely theoretical. In actuality, 802.11b often gets only 7Mbps. That would simply be too slow to support the full rate of this device. If this idea was used with a faster connection, such as 100Mbps ethernet, it could be feasible.
Now, regarding the streaming media server, the techie in me wants to applaud the "free the music" idea. However, the recording engineer in me wants to smack you for suggesting that my expensive new 24 bit at 192KHz mix be compressed (especially given the oh-so-natural sounding Real Audio history that has permiated the industry). But it does seem possible if one invested the money in a seperate streaming media server, and a fat pipe to stream it.
But my main reponse to this is these solutions it that they are expensive. Many commercial units out now, like this one, come with 20GB+ integrated hard drives, among convenient options like integrated CD burners and digital effects. If one isn't going to pay to go into a studio, it's probably cheaper to buy some mid-line recording equipment than to modify a PDA for "professional quality recording".
Got a link to support this? It's good info, but the one linke you provided doesn't seem to be working at the moment. Also, is the compression lossy? Sony quotes the frequency on their minidisc players as 20-20,000 Hz -- what is often considered the full range of human hearing. That would lead me to believe that the compression, however it is being done, is not hurting the high-end response like so many lossy compressions do.
Perhaps I should rephrase: "minidiscs can technically be CD quality." While it is true that most Minidisc players today can utilize compression technology, particularly the different versions of Sony's ATRAC compression, virtually all of them can also record without compression. The compression factor is how Sony now claims that you can fit "Up To 320 Minutes On One 80 Minute Disc." But if the discs are recorded without compression, Minidisc have a sample rate of 16 bits at 44.1 kHz -- the same as a normal CD.
For a long time, live recording has been consigned mostly to the realm of DAT recorders, Minidisc recorders, or laptop computers. On one hand you have subpar sound quality, on the other you have a bulky rig with a big 'steal me' sign attached.
Subpar sound quality? The DATs that I've worked with have better resolution than CDs (48 KHz vs 44.1 KHz sampling), minidiscs are technically CD quality, and laptop computers can be equally sensitive given the right equipment. Given what I've heard of PDA sound, there's nothing subpar about the existing recording mediums. Also, it's hard to claim that a minidisc is "a bulky rig".
By using a hardware card that allows recording via S/PDIF onto Compact Flash, you will be able to use your iPaq or Zaurus alongside a decent A/D converter to portably get field recordings at up to 24bit/192kHz.
So to record in this way, I must buy a "decent" A/D converter and a bunch of Compact Flash. And, unless they are using some compression which will lower the sound quality, this thing will suck up more MB-per-minute of audio than a CD. Good thing Hitchai (formerly IBM) makes their MicroDrive, and I have a money tree in my yard.
So, bottom line as I see it? An interesting project, but one which uses expensive hardware and media that makes it prohibitively expensive. So if you want professional digital recording, get a professional digital recorder. If you want ad-hoc "pro" sound recording from a PDA, now you've got an option.
Freudian Slip!
He thinks the killer app for this one is for keeping your porn storage hidden, if you're busted by the cops.
Freudian Slip - Definition: 'An inadvertent mistake in speech or writing that is thought to reveal a person's unconscious motives, wishes, or attitudes.' - Courtesy of The Random House College Dictionary, 1980
The site www.tonerrefillkits.com isn't working for me, though the WHOIS info says that it's been registered since 1999 (through 2004). Are you sure that this wonder-website isn't another dot-com bust?
You CANNOT fit a 100 foot dish to a satellite and orbit it around Mars or Jupiter, etc, to pick up signals from further out and relay them to earth. Its simply not possible.
Perhaps this is only not possible by conventional thinking, if even then. But picture the Space Elevator hoisting up a large dish, or several small pieces of it. Beyond that, if we can launch an International Space Station that weighs 393,733 lbs, then even a dish factory in space is fathomable. ("Break your satellite? Need new parts? Just order on the universe wide web at uww.spaceparts.com!")
I'm glad this came out on Slashdot: I never would have known otherwise. I have one of the afflicted units, and I'd rather not have a melted puddle of lead-acid powered backup get lose near my PC. A funny side note: I also found my serial number on my Quality Assurance Test receipt. Some quality assurance, I guess.
Show me some research! While it is an interesting idea, I don't see it holding much truth. The author's principle is that waveforms are not uniform to the source, rather approximations. Therefore, the author believes that we may begin to expect such approximated waveforms rather than more exact waveforms. While this may be true, it is not physiological damage. Think of it in another medium. Just because computer monitors use an approximation of images -- that is, images are composed of relatively lossy blocks rather than exact lines and shading -- does not mean that one who uses a computer for extended periods becomes unable to notice the exact sharpness of real life images. The author's idea is also further disproved by the idea that people have been listening to altered signals for years now (think distorted guitar or compressed vocals) without documented hazard.
Personally, I don't buy what this author is dishing out. It has the stench of "anti-MP3." I'd start looking for the author's alterior motives.
After what Kazaa did to MusicCity Morphius (eg. forced them off their joint P2P network without warning), it serves them right! In a market that is based entirely on appeal (generally from reputation) and customer satisfaction, how can any P2P service try to play "hardball" with its "competition" and expect to come out ahead? I'm all for the Napster-esque services, open source, and variety, but I can't see how any P2P services can survive an RIAA onslaught without a glowing reputation, a strong userbase, and the respect of the P2P community. Kazaa, obviously, did not have all of that.