It isn't meaningless, but it is a context-less number. Let's try this a different way.
Bob the scientist goes and gets samples from the air, soil, water, and fish at a site. His sensor can go down to 0.02 micrograms of pesticide per liter of sample. When he checks the results he finds the sensor found no pesticide in the air, 0.05 micrograms/L of pesticide in the soil, 0.02 ugrams/L in the water, and 0.15 ugrams/L in the fish.
The context provides the useful part of the data. The soil shows significant levels of pesticide, indicating it was the area directly sprayed. The air sample shows no pesticide to the limit of the sensor so the spraying was done more than a few hours ago. The water is somewhat contaminated but also at the limit of the sensor so it probably isn't that bad as long as there is rain to further dilute the compound. However the fish sample is several times higher than the water or soil sample, implying the pesticide has been used repeatedly and that the ecosystem may start suffering damage.
If the sensor was limited to detecting 10 micograms/L then you couldn't detect the pesticide prior to visible symptoms in the wildlife. By having a more sensitive sensor you have greater lead time to finding problems. For forensic-type activities, it also means it is easier to track down the point source of the pollutant.
It's a density function so it doesn't matter of what. 0.02 micograms/liter is equivalent to 0.00002 micrograms/cc
Converting to "parts per million" does require attention to "liter of what" as molecule size becomes a factor, just as "percent mass" would also require the "of what" to know the mass of the solvent.
The "of what" would depend on what the source of the sample was: ground water, soil, air, urine, blood. 0.02 micrograms of pesticide per liter of soil may be a non-issue while 0.02 micograms of pesticide per liter of water could indicate a lake is turning toxic. 0.02 micograms of pesticide per liter of blood is probably a sign that the host should call their priest of choice.
I just read the iPhone manual. Most mail requires you to set the "manual check/autcheck every 15/30/60 minutes" flag. Not too sure about Yahoo, which it states "If you have a Yahoo! email account, email is instantly transferred to iPhone as it arrives at the Yahoo! server." If the guy had a yahoo! account, it could be quite difficult to disable the email check feature. Either way, the guy had to set up the email on the machines. I thought the magic Itunes registration process configured the email.
I am a little surprised that you apparently can't disable the GSM/GPRS without also killing the WiFi. Were I on a foreign trip I might find it worthwhile to have my favorite WiFi enabled gizmo handy for websurfing in Starbucks and the like even when I didn't want to use plan minutes.
Actually, you're wrong about your Treo. Even after you put it into sleep mode it will still log itself onto the network and check email if it is configured to do so automatically. Don't believe me? Setup the included Versamail and enable the "autosynch" option.
The big difference is that Palm doesn't configure the Treo to do that automatically. With an iPhone the setup process hooks you up to the email account attached to the phone number with no user action necessary. It's "magic."
As anyone who flies on a plane with a Treo should know, to do a full power down all you need to do is hold the power-off button for ~5 seconds and it will shut down the radio completely. You can tell it is off-net by the fact there's no signal indicator anymore. Matter of fact, my Treo manual highlighted the need to shut down the radio completely when getting on a plane.
I'm not sure what the iPhone process is to go into full radio-off mode but it is hopefully as easy. Whether the manual says so or not I'm not sure. But given "Lord Jobs" rah-rah messages of "it's so easy" then I'm sure many of the faithful declined to read the manual.
About six years ago the engineering firm I work for had a public outreach program where high school students were given PDAs (Palm IIIx) to collect data on the storm and sanitary sewers along creeks. The kids would note which manholes were in the streams, where there were roof downspouts directly attached to the storm sewers, etc. Each manhole has their ID number stamped into the rim and the kids had paper maps as backup.
The data quality was spotty and the teenagers were pretty hard on the equipment but hey, they were teenagers. However the process as a whole was successful. There were virtually no complaints about the difficulty of data entry. Which is pretty encouraging since a non-programmer used a low-cost off-the-shelf PDA database to put it together. It would sync up with Access, which was good enough at the time. The kicker for most PDA data collectors is the sync process. It's worth it to pay someone for a decent data sync plug-in.
Today you should be able to do as well just as easily, if not better.
What you are describing is management as abstraction layers. Which is really as it should be, as progressive levels of management should be dealing with ever larger systems. Abstraction, aka delegation, is a common technique to deal with the greater complexity.
The biggest problem I've had with non-technical managers of a technical group is an inadequate understanding of the phrase "a good manager can manage anything." They incorrectly assume that because they have been a good manager in the past that they can easily be a good manager in something else. A good manager knows that they need to understand their goals, objectives and, most importantly, resources. A good manager who is dropped in an unfamiliar fire will put in the time and effort to learn what their resources (aka "staff") are capable of, to the limits of their need to understand.
And it's not that I disagree with you. My best IT boss ever was an accountant. He was willing to admit when he didn't understand our jargon and forced us to translate into english, which went a long way towards ferreting out the wastes of flesh that could spout buzzwords but didn't produce drek. He was smart enough to differentiate between "don't want to" and "can not" as part of filtering out the wastes. His lack of IT skills was actually a benefit because if we said "we can't do it" he really couldn't come up with some out-of-date process that was utterly unrelated to current conditions. Since his mental approach was based on hard numbers, he could grok quality assurance requirements in the contracts as well as the staffing performance metrics, which meant he understood the math that said when we needed more staff and didn't just say "do better" but went and fought with the executives. Plus, when we could demonstrate that a capital cost (more detailed user manuals) would reduce operating cost (tech support calls) he would take that to the suits and make them eat it.
But I stand my statement.. anyone who says they can take a serial application and run it in parallel is full of sh*t and they know it. In certain, limited circumstances, yes... but in general. NO.
Then you miss the point of the PRAM concept. It is designed to evaluate concurrency and to identify parallel-friendly routines. Part of this project is funded by the NSF and includes developing compiler modifications and an API.
But the PRAM model only requires that the parallel memory accesses occur at constant time - that constant could still be huge and it would run PRAM programs.
Right, which is why I said that it would be significant even if it wasn't fast. It's a prototype, first of its kind. It could run like a dog and still be "news for geeks." Kind of like the first quantum computer is/will be even if it doesn't really go faster.
If they have got those times down to something useful that is clearly a step forward, but why then are they claiming superlinear speedup (64 CPUs performing as 100) and "desktop applications"?
I'm honestly not sure what metric they are using. I'd imagine they could run custom software and I'd be surprised if they didn't have a PRAM algorithm interpreter, so they probably ran a suite of functions that followed some hypothetical "desktop software" equivalence to get some performance indexing.
I know they had the hardware available for public access (public meaning the attendees at the ACM International Conference on Supercomputing) so it probably isn't a load of hooey, but it could still be market speak for "does some stuff horrifically fast and could be on the market in 5 years." I'd really expect this to show up as a coprocessor or series of subunits on a traditional CPU at first. I figure it would be like the GPU-based protein folding software or the physics processor, taking tasks flagged as "PRAM-friendly" that throws in some overdrive.
Ironically, gaming would get a huge boost from A good parallel processing CPU and API. Many aspects of games could be parallelized (each bot/unit gets their own process, for instance) although I really have no idea if a PRAM processor cluster would be better than using a multicore x86 cpu.
Up 'til now, Parallel Random Access Model (PRAM) computing has been a theory of parallel processing that was a thought model. It hadn't been built. Some people had written programs to emulate a PRAM computer but they were not complete versions.
It could work at a snail's pace and still be a technological accomplishment as it is the very first, complete, working, hardware PRAM computer. It's on par with the Z3, Colossus and Eniac, the first programmable computers (German, English, American, in historical order).
Fortunately, they made the algorithms work well, or at least, if the press release it to be believed, work so that 64 75Mhz computers could produce 100x the performance of a current desktop on at least one particular function. Which is pretty impressive in first-time hardware even if it turns out to be an obscurely used math function known only to about a dozen coders.
Both. I've got an old Series1 from my inlaws that I don't pay for service and a Series 2 that I do. The Series1 does not get any program guide information due to our lack of a landline.
What we need is a simple tapeless video recorder that lets you watch a show that has already started while it finishes recording. That's all I want, and I want it in a 19" appliance and without any subscription to anything. I haven't found it yet.
Get a cheap TiVo, possibly a used one, and don't pay for service. We program the Series 1 like a regular VCR, by setting the time date and recurrence options. You can still pause live TV and start to watch a show being recorded from the beginning.
You can do that with TiVo as well. The series 2 has a PC server program that will let you copy your files off your TiVo onto a PC. Yeah, they have DRM (to keep the MPAA off their backs) but it's hella weak and there are programs out there that strip it off easily*. Windows Media Player asks for a password and then you can play the video.
Multiple TiVos can actually copy videos back and forth. I've only got an 802.11b wifi adapter on my TiVo and it provides slightly better than real-time data transfer rates. I've been archiving off "Good Eats" to my PC since my wife doesn't appreciate having ~40 episodes stored.
The server program also lets you play MP3s and pictures on your TiVo.
*I've managed to strip the DRM off and copy the videos to DVD but there are frame synch issues that irritate me. The audio is just barely off and the video stream flickers; I think it's discarding some frames. I'm not sure if it is a result of the DRM removal or a setting on the DVD recorder.
There is a service called "Tivo Basic" that is a lifetime service attached to some products. It does not have the wishlist or the online TiVo component but it does provide the guide service and program search which appears to be on par with Zap2It's listing system.
It comes on the Toshiba SDH-400 TiVo/DVD player (which I own) and the Pioneer DVR-810H and DVR-57H DVD recorders. You can pick up an 810 for about $250. There are probably more but I stopped at the first Google hit.
Well, back in the 18th century when the original system was set up using physical objects was about the best they could do. The meter was based on a percentage of the earth's meridian and the kilogram is derived from volume of water at maximum density. However that turned out to be a variable number, even using the same container, because of issues with pressure.
In the late 19th century the iridium-platinum prototype was created that were accepted to be equal to the mass of a "kilogram" of water under the conditions expected at the time of definition. Actually, three of them were created and stored in different locations to provide a check against any mass drift in the prototypes. Multiple copies have been made over the years delivered to nations across the planet to provide their own base references. There is apparently some change in mass but no one can explain why, and this is important.
The problem is that in the interim the other units have been changed to something that can be derived. Time is based on a certain number of cycles in a cesium atom at absolute zero. Length is defined as the speed of light in a vacuum in one second. The kicker here is that as improved measuring devices appear, the accuracy of the meter and second only improve as the new measurements add extra digits off the decimal place. Anybody can do them, anywhere. Build an experimental chamber that gets closer to absolute zero and has more sensitive cesium detector = more accurate measurement of a second. Build an experimental chamber that gets closer to absolute vacuum and a more accurate measurement of a second = more accurate measurement of a meter.
To develop a more accurate measurement of a kilogram and...you need the lump of iridium-platinum. And that measurement will be limited by the tiny change in mass that appears to be happening with no explanation.
Once a kilogram is defined as X atoms of silicon, anyone with a silicon atom counter can make their own base unit for equipment calibration.
Sony is teh suxxorz where drivers are concerned. I've had to deal with multiple Sony laptops and Clies. They are beautiful pieces of hardware while they work but once they freak out it becomes a total nightmare. Freaky custom installers, lack of downloadable files ($25 for a freaking driver CD? Bite me!), and sometimes custom cabling that they have no supplies for (I'm looking at you, external CDRom drive that needed a weird two-port USB cable for power!)
Apple can be picky about drivers, no doubt. But with the exception of a blue G3 (G4?) tower with a defective mobo, I've never had an Apple degrade the same way every Sony computer device has. (I only speak of their computers. I've got a Playstation 1 that refuses to die, along with an old receiver.)
I guess that depends on the expected usage of the unit over its operating life. If you assume it will be used twice a year plus an annual maintenance startup then you have a nominal life of 33 years, which is on par or greater than traditional internal combustion generators. If, on the other hand, you were in someplace with relatively frequent power outages, say monthly, then it would only last a decade.
The flip side is that if you have need for continuous generation power, particularly in remote areas where the cost of transporting the fuel is significant, the increased efficiency over the life of the unit could be a major cost savings.
I wonder what the startup time is on the cells. The lack of moving parts and high efficiency sounds like it would be ideal for a backup generator since you could get twice the duration for the same fuel tank. The big question is how long it will take to reach nominal load. If you need an excessive amount of batteries to make the transition it could still be unfeasible.
One would think that you could get racks of the things to get generation capacity in excess of 5KW since the units already consist of multiple tubes. It would simply mean removing the individual DC/AC converters and using one big one.
Anyone have any idea what the maintenance cycles are on fuel cells and how long you can let one sit idle?
I really, really long for a big palm pilot, somewhere around the size of a tablet PC. PalmOS supports XGA resolutions and it would be great for web browsing on WiFi/BT DUN, doing light data entry (Docs to go is pretty good, IMO), VNC/RDP to my PCs, run my handy PalmOS apps, support BT stereo audio for my MP3s, and have at least 2 flash memory slots to ensure plenty of storage.
This....is like they took that idea and threw away the good parts. They are NOT running a PalmOS emulator on Linux or else they wouldn't need to port Opera or Docs to Go. They are NOT making it tablet-like since it has no touch screen and the display won't fold all the way over.
Unless Palm hacked up the OS or used special libraries, there's nothing that the Foleo does that couldn't be done with a Nokia N800.
And if I could get a PalmOS emulator for the N800 so I could keep my palm apps, I might be satisfied with that, even if it is a little smaller than I'd really want.
We were an ISP call center, particularly an ISP for VARs, meaning that people would take our dial-up/ISDN (the pre-DSL days of 1995-2000) product and slap their brands on it. The customers (meaning the VARs) were responsible for the cost of call delivery and, since many of them were smaller telecom companies or larger companies that already had favorable bulk circuit rates, would often provision their own T1 into our facility.
In some cases there would be Ts between their call centers and ours, allowing their CS to transfer to our tech support and then back again to address billing issues.
And if you paid attention to my original post, I said our IVR also directed calls to four other call centers located in Baltimore, Little Rock, Houston and Kansas City. That's not to say our switch was providing 100% of their calls or handled all of their queue management but that when our call center was full the overflow would be directed to Houston, Little Rock, Kansas City or, for certain customers, Baltimore.
At one point we had more than 70 active brands. We had about ten religious organizations (several of them national or large regionals, like the Presbyterian Church), probably a dozen universities (mostly smaller ones) and some larger companies (Penzoil). Eventually we bought/acquired/something Digex's user base but due to a massive management snafu of near biblical proportions exascerbated Digex virtually doing everything in their power to make it as hard as possible on the end users, virtually none of them stayed through the migration.
If someone, anyone, can set up an Asterix PBX from scratch then it is a relatively easy to configure PBX.
I have been part of a Nortel PBX roll out. Nortel Meridian 61C, about 2 dozen T1s incoming, around 150 handsets, redundant IVRs (Symposium as primary that also did pre-queing for 4 other call centers, voicemail-based IVR as backup, old fashioned rotary groups as tertiary), with an early generation (1999) VoIP circuit.
With experienced installers (Greg & Danny were great) it was a by-the-book PBX install, meaning it took about 2 weeks to get all the circuits configured correctly (we were dealing with like 6 C/LECs and we had inter company links to two different organizations), all the users in place, program all the sets, configure our layers of redundancy, and go live.
Easy? No, not very. A Meridian Option 61c came with literally dozens of manuals, almost completely taking up our 10ft wide shelf over the console. Each covered only a particular subset of commands. One volume was the index.
I doubt that any install is easy just because of the number of options a decent phone system supports. And PBXs are, in fact, critical tech. I had what I was told were paranoid amounts of redundancy in our call center configuration and I wound up needing it. Our Windows NT Symposium server came with a defective SCSI cable that took nearly a month to identify (during our 3am-5am maintenance window, JOY) and during one of the windows the IVR system failed (I think a backup tape broke, causing a system fault and a CPU fail over but I could be wrong) meaning that our call center was now relying on 1960s-style rotary hunt groups, where each phone rings two or three times before being forwarded to the next one in the list.
Yeah, eventually, like 2-4 years following the release of the Linux SDK. Other than a handful of cutting edge developers, it will take several years for the bulk of commercial coders to reach a point where a whole hog rebuild of their code is justified.
At that point the OP's hardware will be many years old and likely in need of a replacement due to physical damage. Which will be fine, since pretty much all his apps will migrate to the new device, barring a few that might use hardware-specific hacks.
Actually there were other products that did that. I had a dinky little 386 back in the days of Windows 3 and I had a product, I think from QEMM, that was a memory manager and task switcher. It let me run Lotus123 and WordPerfect (both for DOS) at the same time, which was very handy when writing lab reports.
I hope you meant $200/hundred DIDs b/c otherwise Verizon is ripping you a new one or the market has gone to shinola. I left telecom about six years ago but I remember putting in orders for a hundred DIDs at a time and I know it wasn't $20k.
Our call center wasn't that large onsite but we provided services for for lots of VARs who'd use their own brand and did a lot of call re-routing. I think at one point there were more than a hundred 1-800 numbers pointed at our Meridian. We only had about 30 T1s for all those 800s so most of 'em where divided in the switch into 3-12 channel pools. Wasn't the most effective use of our channels but each 800 was billed separately and several clients preferred it this way. Kept their costs down if they were low volume since our shared pool costs were cost effective at around 4-5 channels. We kept most of the PRIs with DIDs so that we could rearrange the circuit loading by repointing the 800 and managing the channels on the switch, which tended to be necessary when clients would have a big marketing push and request going from 2 channels to a high volume pool. Usually they requested it after their clients complained about getting busy signals and we reminded them about the way they cheaped out on the contract.
Re:PocketPC is better than Palm - well, until now
on
Palm to go Linux
·
· Score: 1
I was around back then and the problem was memory storage. Most 2001 era devices had 8-16MB storage. External storage was still limited mostly to 32MB MMC cards which were only slightly less expensive than a 32MB MP3 player. I had a Visor and could use one of the Springboard MP3 player addons if I wanted. But with only the ability to store a handful of MP3s and only a couple minutes of video it was a demand that couldn't be cost justified. The MP3 phenomena really didn't take off until the 128MB and 256MB MP3 players hit the market at an affordable price. Even today I would be surprised if more than 20% of WinMob/Treo users play MP3/video on their device on a weekly basis.
Free-hand writing recognition has been crappy all the way around IME. I tried it repeatedly on Palm (it was 3rd party, but existed) and on WinCE and WinMob. Could be because I'm a lefty but handwriting on everything but a Newton (freaky piece of kit that was) was pretty much pointless until the last 2-3 years when ~300Mhz CPUs became available. Even when the recognition was acceptable the lag was interminable. I can still grafitti faster than anyone I know can thumb-board, even if I pull out my old PalmIIIx rather than my Treo650.
Color started with the IIIC and Prism series which were cheaper than the equivalent Jornadas, IPaqs, and Casio EMxxs of the day. True, the Windows devices had better screens but that was a temporary market condition since Palm was still using the aging PalmOS v3.5 while Windows was the brand new CE3.0.
Many GPS units back in 2000 ran PalmOS to make their $750 cost more appetizing since you at least got a $300-400 PDA mixed in. All the original GPS manufacturers had Palm-enabled GPS units so if you had problems it was more likely due to the specific GPS receiver & software.
Slashdot Mirror
It isn't meaningless, but it is a context-less number. Let's try this a different way.
Bob the scientist goes and gets samples from the air, soil, water, and fish at a site. His sensor can go down to 0.02 micrograms of pesticide per liter of sample. When he checks the results he finds the sensor found no pesticide in the air, 0.05 micrograms/L of pesticide in the soil, 0.02 ugrams/L in the water, and 0.15 ugrams/L in the fish.
The context provides the useful part of the data. The soil shows significant levels of pesticide, indicating it was the area directly sprayed. The air sample shows no pesticide to the limit of the sensor so the spraying was done more than a few hours ago. The water is somewhat contaminated but also at the limit of the sensor so it probably isn't that bad as long as there is rain to further dilute the compound. However the fish sample is several times higher than the water or soil sample, implying the pesticide has been used repeatedly and that the ecosystem may start suffering damage.
If the sensor was limited to detecting 10 micograms/L then you couldn't detect the pesticide prior to visible symptoms in the wildlife. By having a more sensitive sensor you have greater lead time to finding problems. For forensic-type activities, it also means it is easier to track down the point source of the pollutant.
It's a density function so it doesn't matter of what. 0.02 micograms/liter is equivalent to 0.00002 micrograms/cc
Converting to "parts per million" does require attention to "liter of what" as molecule size becomes a factor, just as "percent mass" would also require the "of what" to know the mass of the solvent.
The "of what" would depend on what the source of the sample was: ground water, soil, air, urine, blood. 0.02 micrograms of pesticide per liter of soil may be a non-issue while 0.02 micograms of pesticide per liter of water could indicate a lake is turning toxic. 0.02 micograms of pesticide per liter of blood is probably a sign that the host should call their priest of choice.
I just read the iPhone manual. Most mail requires you to set the "manual check/autcheck every 15/30/60 minutes" flag. Not too sure about Yahoo, which it states "If you have a Yahoo! email account, email is instantly transferred to iPhone as it arrives at the Yahoo! server." If the guy had a yahoo! account, it could be quite difficult to disable the email check feature. Either way, the guy had to set up the email on the machines. I thought the magic Itunes registration process configured the email.
I am a little surprised that you apparently can't disable the GSM/GPRS without also killing the WiFi. Were I on a foreign trip I might find it worthwhile to have my favorite WiFi enabled gizmo handy for websurfing in Starbucks and the like even when I didn't want to use plan minutes.
Actually, you're wrong about your Treo. Even after you put it into sleep mode it will still log itself onto the network and check email if it is configured to do so automatically. Don't believe me? Setup the included Versamail and enable the "autosynch" option.
The big difference is that Palm doesn't configure the Treo to do that automatically. With an iPhone the setup process hooks you up to the email account attached to the phone number with no user action necessary. It's "magic."
As anyone who flies on a plane with a Treo should know, to do a full power down all you need to do is hold the power-off button for ~5 seconds and it will shut down the radio completely. You can tell it is off-net by the fact there's no signal indicator anymore. Matter of fact, my Treo manual highlighted the need to shut down the radio completely when getting on a plane.
I'm not sure what the iPhone process is to go into full radio-off mode but it is hopefully as easy. Whether the manual says so or not I'm not sure. But given "Lord Jobs" rah-rah messages of "it's so easy" then I'm sure many of the faithful declined to read the manual.
About six years ago the engineering firm I work for had a public outreach program where high school students were given PDAs (Palm IIIx) to collect data on the storm and sanitary sewers along creeks. The kids would note which manholes were in the streams, where there were roof downspouts directly attached to the storm sewers, etc. Each manhole has their ID number stamped into the rim and the kids had paper maps as backup.
The data quality was spotty and the teenagers were pretty hard on the equipment but hey, they were teenagers. However the process as a whole was successful. There were virtually no complaints about the difficulty of data entry. Which is pretty encouraging since a non-programmer used a low-cost off-the-shelf PDA database to put it together. It would sync up with Access, which was good enough at the time. The kicker for most PDA data collectors is the sync process. It's worth it to pay someone for a decent data sync plug-in.
Today you should be able to do as well just as easily, if not better.
The biggest problem I've had with non-technical managers of a technical group is an inadequate understanding of the phrase "a good manager can manage anything." They incorrectly assume that because they have been a good manager in the past that they can easily be a good manager in something else. A good manager knows that they need to understand their goals, objectives and, most importantly, resources. A good manager who is dropped in an unfamiliar fire will put in the time and effort to learn what their resources (aka "staff") are capable of, to the limits of their need to understand.
And it's not that I disagree with you. My best IT boss ever was an accountant. He was willing to admit when he didn't understand our jargon and forced us to translate into english, which went a long way towards ferreting out the wastes of flesh that could spout buzzwords but didn't produce drek. He was smart enough to differentiate between "don't want to" and "can not" as part of filtering out the wastes. His lack of IT skills was actually a benefit because if we said "we can't do it" he really couldn't come up with some out-of-date process that was utterly unrelated to current conditions. Since his mental approach was based on hard numbers, he could grok quality assurance requirements in the contracts as well as the staffing performance metrics, which meant he understood the math that said when we needed more staff and didn't just say "do better" but went and fought with the executives. Plus, when we could demonstrate that a capital cost (more detailed user manuals) would reduce operating cost (tech support calls) he would take that to the suits and make them eat it.
Then you miss the point of the PRAM concept. It is designed to evaluate concurrency and to identify parallel-friendly routines. Part of this project is funded by the NSF and includes developing compiler modifications and an API.
But the PRAM model only requires that the parallel memory accesses occur at constant time - that constant could still be huge and it would run PRAM programs.
Right, which is why I said that it would be significant even if it wasn't fast. It's a prototype, first of its kind. It could run like a dog and still be "news for geeks." Kind of like the first quantum computer is/will be even if it doesn't really go faster.
If they have got those times down to something useful that is clearly a step forward, but why then are they claiming superlinear speedup (64 CPUs performing as 100) and "desktop applications"?
I'm honestly not sure what metric they are using. I'd imagine they could run custom software and I'd be surprised if they didn't have a PRAM algorithm interpreter, so they probably ran a suite of functions that followed some hypothetical "desktop software" equivalence to get some performance indexing.
I know they had the hardware available for public access (public meaning the attendees at the ACM International Conference on Supercomputing) so it probably isn't a load of hooey, but it could still be market speak for "does some stuff horrifically fast and could be on the market in 5 years." I'd really expect this to show up as a coprocessor or series of subunits on a traditional CPU at first. I figure it would be like the GPU-based protein folding software or the physics processor, taking tasks flagged as "PRAM-friendly" that throws in some overdrive. Ironically, gaming would get a huge boost from A good parallel processing CPU and API. Many aspects of games could be parallelized (each bot/unit gets their own process, for instance) although I really have no idea if a PRAM processor cluster would be better than using a multicore x86 cpu.
Here's the deal.
Up 'til now, Parallel Random Access Model (PRAM) computing has been a theory of parallel processing that was a thought model. It hadn't been built. Some people had written programs to emulate a PRAM computer but they were not complete versions.
It could work at a snail's pace and still be a technological accomplishment as it is the very first, complete, working, hardware PRAM computer. It's on par with the Z3, Colossus and Eniac, the first programmable computers (German, English, American, in historical order).
Fortunately, they made the algorithms work well, or at least, if the press release it to be believed, work so that 64 75Mhz computers could produce 100x the performance of a current desktop on at least one particular function. Which is pretty impressive in first-time hardware even if it turns out to be an obscurely used math function known only to about a dozen coders.
Do you own your TiVo, or do you pay a fee for it?
Both. I've got an old Series1 from my inlaws that I don't pay for service and a Series 2 that I do. The Series1 does not get any program guide information due to our lack of a landline.
What we need is a simple tapeless video recorder that lets you watch a show that has already started while it finishes recording. That's all I want, and I want it in a 19" appliance and without any subscription to anything. I haven't found it yet.
Get a cheap TiVo, possibly a used one, and don't pay for service. We program the Series 1 like a regular VCR, by setting the time date and recurrence options. You can still pause live TV and start to watch a show being recorded from the beginning.
You can do that with TiVo as well. The series 2 has a PC server program that will let you copy your files off your TiVo onto a PC. Yeah, they have DRM (to keep the MPAA off their backs) but it's hella weak and there are programs out there that strip it off easily*. Windows Media Player asks for a password and then you can play the video.
Multiple TiVos can actually copy videos back and forth. I've only got an 802.11b wifi adapter on my TiVo and it provides slightly better than real-time data transfer rates. I've been archiving off "Good Eats" to my PC since my wife doesn't appreciate having ~40 episodes stored.
The server program also lets you play MP3s and pictures on your TiVo.
*I've managed to strip the DRM off and copy the videos to DVD but there are frame synch issues that irritate me. The audio is just barely off and the video stream flickers; I think it's discarding some frames. I'm not sure if it is a result of the DRM removal or a setting on the DVD recorder.
Sure, here ya go.
& hl=en&client=firefox-a&rls=org.mozilla:en-US:offic ial&hs=Vm0&um=1&sa=X&oi=froogle&ct=titleb tnG=Search&hl=en&show=dds how=dd&q=Toshiba+SDH400
There is a service called "Tivo Basic" that is a lifetime service attached to some products. It does not have the wishlist or the online TiVo component but it does provide the guide service and program search which appears to be on par with Zap2It's listing system.
It comes on the Toshiba SDH-400 TiVo/DVD player (which I own) and the Pioneer DVR-810H and DVR-57H DVD recorders. You can pick up an 810 for about $250. There are probably more but I stopped at the first Google hit.
http://www.google.com/products?q=Pioneer+DVR-810H
http://www.google.com/products?q=Pioneer+DVR-57H&
http://www.google.com/products?btnG=Search&hl=en&
Well, back in the 18th century when the original system was set up using physical objects was about the best they could do. The meter was based on a percentage of the earth's meridian and the kilogram is derived from volume of water at maximum density. However that turned out to be a variable number, even using the same container, because of issues with pressure.
...you need the lump of iridium-platinum. And that measurement will be limited by the tiny change in mass that appears to be happening with no explanation.
In the late 19th century the iridium-platinum prototype was created that were accepted to be equal to the mass of a "kilogram" of water under the conditions expected at the time of definition. Actually, three of them were created and stored in different locations to provide a check against any mass drift in the prototypes. Multiple copies have been made over the years delivered to nations across the planet to provide their own base references. There is apparently some change in mass but no one can explain why, and this is important.
The problem is that in the interim the other units have been changed to something that can be derived. Time is based on a certain number of cycles in a cesium atom at absolute zero. Length is defined as the speed of light in a vacuum in one second. The kicker here is that as improved measuring devices appear, the accuracy of the meter and second only improve as the new measurements add extra digits off the decimal place. Anybody can do them, anywhere. Build an experimental chamber that gets closer to absolute zero and has more sensitive cesium detector = more accurate measurement of a second. Build an experimental chamber that gets closer to absolute vacuum and a more accurate measurement of a second = more accurate measurement of a meter.
To develop a more accurate measurement of a kilogram and
Once a kilogram is defined as X atoms of silicon, anyone with a silicon atom counter can make their own base unit for equipment calibration.
Sony is teh suxxorz where drivers are concerned. I've had to deal with multiple Sony laptops and Clies. They are beautiful pieces of hardware while they work but once they freak out it becomes a total nightmare. Freaky custom installers, lack of downloadable files ($25 for a freaking driver CD? Bite me!), and sometimes custom cabling that they have no supplies for (I'm looking at you, external CDRom drive that needed a weird two-port USB cable for power!)
Apple can be picky about drivers, no doubt. But with the exception of a blue G3 (G4?) tower with a defective mobo, I've never had an Apple degrade the same way every Sony computer device has. (I only speak of their computers. I've got a Playstation 1 that refuses to die, along with an old receiver.)
We've been here a while. I think my friend who turned me onto slashdot has a 3-digit ID.
I guess that depends on the expected usage of the unit over its operating life. If you assume it will be used twice a year plus an annual maintenance startup then you have a nominal life of 33 years, which is on par or greater than traditional internal combustion generators. If, on the other hand, you were in someplace with relatively frequent power outages, say monthly, then it would only last a decade.
The flip side is that if you have need for continuous generation power, particularly in remote areas where the cost of transporting the fuel is significant, the increased efficiency over the life of the unit could be a major cost savings.
I wonder what the startup time is on the cells. The lack of moving parts and high efficiency sounds like it would be ideal for a backup generator since you could get twice the duration for the same fuel tank. The big question is how long it will take to reach nominal load. If you need an excessive amount of batteries to make the transition it could still be unfeasible.
One would think that you could get racks of the things to get generation capacity in excess of 5KW since the units already consist of multiple tubes. It would simply mean removing the individual DC/AC converters and using one big one.
Anyone have any idea what the maintenance cycles are on fuel cells and how long you can let one sit idle?
I really, really long for a big palm pilot, somewhere around the size of a tablet PC. PalmOS supports XGA resolutions and it would be great for web browsing on WiFi/BT DUN, doing light data entry (Docs to go is pretty good, IMO), VNC/RDP to my PCs, run my handy PalmOS apps, support BT stereo audio for my MP3s, and have at least 2 flash memory slots to ensure plenty of storage.
This....is like they took that idea and threw away the good parts. They are NOT running a PalmOS emulator on Linux or else they wouldn't need to port Opera or Docs to Go. They are NOT making it tablet-like since it has no touch screen and the display won't fold all the way over.
Unless Palm hacked up the OS or used special libraries, there's nothing that the Foleo does that couldn't be done with a Nokia N800.
And if I could get a PalmOS emulator for the N800 so I could keep my palm apps, I might be satisfied with that, even if it is a little smaller than I'd really want.
We were an ISP call center, particularly an ISP for VARs, meaning that people would take our dial-up/ISDN (the pre-DSL days of 1995-2000) product and slap their brands on it. The customers (meaning the VARs) were responsible for the cost of call delivery and, since many of them were smaller telecom companies or larger companies that already had favorable bulk circuit rates, would often provision their own T1 into our facility.
In some cases there would be Ts between their call centers and ours, allowing their CS to transfer to our tech support and then back again to address billing issues.
And if you paid attention to my original post, I said our IVR also directed calls to four other call centers located in Baltimore, Little Rock, Houston and Kansas City. That's not to say our switch was providing 100% of their calls or handled all of their queue management but that when our call center was full the overflow would be directed to Houston, Little Rock, Kansas City or, for certain customers, Baltimore.
At one point we had more than 70 active brands. We had about ten religious organizations (several of them national or large regionals, like the Presbyterian Church), probably a dozen universities (mostly smaller ones) and some larger companies (Penzoil). Eventually we bought/acquired/something Digex's user base but due to a massive management snafu of near biblical proportions exascerbated Digex virtually doing everything in their power to make it as hard as possible on the end users, virtually none of them stayed through the migration.
Huh, I wonder if that's where the idea for the "giant mutant animal" movies of the 50s came from.
If someone, anyone, can set up an Asterix PBX from scratch then it is a relatively easy to configure PBX.
I have been part of a Nortel PBX roll out. Nortel Meridian 61C, about 2 dozen T1s incoming, around 150 handsets, redundant IVRs (Symposium as primary that also did pre-queing for 4 other call centers, voicemail-based IVR as backup, old fashioned rotary groups as tertiary), with an early generation (1999) VoIP circuit.
With experienced installers (Greg & Danny were great) it was a by-the-book PBX install, meaning it took about 2 weeks to get all the circuits configured correctly (we were dealing with like 6 C/LECs and we had inter company links to two different organizations), all the users in place, program all the sets, configure our layers of redundancy, and go live.
Easy? No, not very. A Meridian Option 61c came with literally dozens of manuals, almost completely taking up our 10ft wide shelf over the console. Each covered only a particular subset of commands. One volume was the index.
I doubt that any install is easy just because of the number of options a decent phone system supports. And PBXs are, in fact, critical tech. I had what I was told were paranoid amounts of redundancy in our call center configuration and I wound up needing it. Our Windows NT Symposium server came with a defective SCSI cable that took nearly a month to identify (during our 3am-5am maintenance window, JOY) and during one of the windows the IVR system failed (I think a backup tape broke, causing a system fault and a CPU fail over but I could be wrong) meaning that our call center was now relying on 1960s-style rotary hunt groups, where each phone rings two or three times before being forwarded to the next one in the list.
Yeah, eventually, like 2-4 years following the release of the Linux SDK. Other than a handful of cutting edge developers, it will take several years for the bulk of commercial coders to reach a point where a whole hog rebuild of their code is justified.
At that point the OP's hardware will be many years old and likely in need of a replacement due to physical damage. Which will be fine, since pretty much all his apps will migrate to the new device, barring a few that might use hardware-specific hacks.
Actually there were other products that did that. I had a dinky little 386 back in the days of Windows 3 and I had a product, I think from QEMM, that was a memory manager and task switcher. It let me run Lotus123 and WordPerfect (both for DOS) at the same time, which was very handy when writing lab reports.
I hope you meant $200/hundred DIDs b/c otherwise Verizon is ripping you a new one or the market has gone to shinola. I left telecom about six years ago but I remember putting in orders for a hundred DIDs at a time and I know it wasn't $20k.
Our call center wasn't that large onsite but we provided services for for lots of VARs who'd use their own brand and did a lot of call re-routing. I think at one point there were more than a hundred 1-800 numbers pointed at our Meridian. We only had about 30 T1s for all those 800s so most of 'em where divided in the switch into 3-12 channel pools. Wasn't the most effective use of our channels but each 800 was billed separately and several clients preferred it this way. Kept their costs down if they were low volume since our shared pool costs were cost effective at around 4-5 channels. We kept most of the PRIs with DIDs so that we could rearrange the circuit loading by repointing the 800 and managing the channels on the switch, which tended to be necessary when clients would have a big marketing push and request going from 2 channels to a high volume pool. Usually they requested it after their clients complained about getting busy signals and we reminded them about the way they cheaped out on the contract.
I was around back then and the problem was memory storage. Most 2001 era devices had 8-16MB storage. External storage was still limited mostly to 32MB MMC cards which were only slightly less expensive than a 32MB MP3 player. I had a Visor and could use one of the Springboard MP3 player addons if I wanted. But with only the ability to store a handful of MP3s and only a couple minutes of video it was a demand that couldn't be cost justified. The MP3 phenomena really didn't take off until the 128MB and 256MB MP3 players hit the market at an affordable price. Even today I would be surprised if more than 20% of WinMob/Treo users play MP3/video on their device on a weekly basis.
Free-hand writing recognition has been crappy all the way around IME. I tried it repeatedly on Palm (it was 3rd party, but existed) and on WinCE and WinMob. Could be because I'm a lefty but handwriting on everything but a Newton (freaky piece of kit that was) was pretty much pointless until the last 2-3 years when ~300Mhz CPUs became available. Even when the recognition was acceptable the lag was interminable. I can still grafitti faster than anyone I know can thumb-board, even if I pull out my old PalmIIIx rather than my Treo650.
Color started with the IIIC and Prism series which were cheaper than the equivalent Jornadas, IPaqs, and Casio EMxxs of the day. True, the Windows devices had better screens but that was a temporary market condition since Palm was still using the aging PalmOS v3.5 while Windows was the brand new CE3.0.
Many GPS units back in 2000 ran PalmOS to make their $750 cost more appetizing since you at least got a $300-400 PDA mixed in. All the original GPS manufacturers had Palm-enabled GPS units so if you had problems it was more likely due to the specific GPS receiver & software.