Extending the dataflow model to structured data with a nice 2d shell that let you tie your pipes and filters together in ways that the UNIX shell couldn't dream of seemed like the obvious next step.
CMS Pipelines did some of this. Not with structured data, nor literally 2D, but you could split and combine flows, and code feedback loops.
People in California should get a chuckle over this. State universities (and a lot of other state agencies) currently have a furlough program in place. Workers are paid some fraction of a salary and work the commensurate time. I think it's 90% for universities. So what used to be a 30-lecture class is now 27 lectures as the professor must take three "furlough" days. That's all the state can afford.
Somehow the number of credits remains the same. I'm not sure how that works. Maybe the state thinks the whole "furlough" business will be over before the accreditation agencies notice.
Original article also compares a Peta of floating ops per *second* to an Exa of byte "processing and storing" per *day.* Journalism profs should save that article for class discussion.
You build a computer that works in base 10 instead of base 2 and then you can call it arbitrary.
The base 10 units for disk drive size were established by IBM, which made the first disk drive. (That drive held 5,000,000 characters, replacing 62,500 punched cards. Each punched card had held 2^6.3219 characters). They also produced base 10 computers. The IBM 1401 which was available with 4K (4000 characters, or digits if you were doing arithmetic) to 16K of memory.
The question of when, if ever, drive manufactures should have made the switch from base 10. Perhaps the introduction of fixed-block disks, which have binary block lengths (512) and cannot be user-formatted to say, 1000 byte blocks. would have been a good time. But that would have confused things too.
An image with more contrast (greater average difference between adjacent pixels) probably has more detail. But compressability, as has already been noted, is probably just as good a measure.
Yes, everything sounds like AS/400. And System/38 before that. Which was sometimes known as "baby FS," where FS was IBM's Future System project. In order to contains every object in a single address space, FS was to have an 80-bit address space. FS was canceled in 1975 (per Wikipedia). System/38 had a 48 address space (or rather, object space), and AS/400 increased it to 64, or maybe 65. System/38 had two layers of microcode: Horizontal (like other IBM machines) and Vertical (which was the VM interpreter for the instruction set exposed to end users). But few customers used the VM directly. They just programmed in RPG or COBOL, or bought program suites written in RPG or COBOL.
Not having heard of Fixstars before, I'm not sure what to make of this. Their website reports that it is "the pioneering company of the Cell Broadband Engine," presumably leading the way for later entrants such as IBM, Toshiba, and Sony.
I remember TerraSoft mostly for overpriced hardware obviously intended for developers with a corporate checkbook behind them. I would look at the prices and decide that Apple was the cheaper source for PowerPC systems.
How about Fixstars? Looking at the web store, I can get Sony PS3 ($450), GigaAccel PCIe card (no price or delivery date listed), a fully populated IBM BladeCenter Chassis ($170,800), or a YDL PowerStation with "Quad-core 2.5GHz IBM 970MP CPUs" ($1895).
The YDL Powerstation sounds interesting and affordable. "Quad-core CPUs" (plural!) for $1895. How many quad-core CPUs, exactly? There is a "learn more" link, but really nothing more there. Less, in fact. That's about the level of detail I used to see from TerraSoft.
they would have had multiple operators dedicated to watching over particular areas in case of such a fault. It looks like someone wasn't paying attention.
You were misled by a terrible headline. The 0.557 petaflop computer is the fastest *for open science.* Roadrunner, at Los Alamos, tops the list. It does 1 petaflop.
In short, beware of tools which produce graphs. Or worse, pie charts.
My story: management spent 25,000 on a malloc checker which produced pie charts. Our code didn't use malloc - at least not directly - so the code tested clean. Waste of money, but management liked the pie charts.
On the other hand, lint, gcc warnings (use more than -Wall, which is not really "all"), and Coverity are all good things to use.
Just what I was wondering: "Maybe, because they hand-code everything, they will pass the validation that all the fancy tools fail at so badly." Anyway, they are not alone. Here are the error couns for the Fortume top 20 companies (top of the Fortune 1000 list) manage on the w3c validator:
Don't forget that the loan shark gets a controlling interest in the company. What are the chances the shark (acting as SCO) is going to ask itself (SNCP) for millions of dollars to pay a judgment to Novell? If/when that happens, the shark needs to ask itself whether its worth the risk, even at 21%. It only makes sense if SNCP really believes that SCO can get the Novell judgment reversed on appeal and win some money from IBM. That could be a big, risky bet.
The deal appears to be that SNCP is committed to loaning SCO $95 million, but only once. If the SNCP loans SCO $95 million on Monday and SCO repays the loan on Tuesday, it appears that the commitment has been fulfilled.
The parent is the vote on the final bill. Here is the vote on the Dodd-Feingold amendment. A "Yea" is a vote for immunity. In my state, Diane Feinstein (D-CA) continued her near-perfect voting record against the Bill of Rights. Barbara Boxer (D-CA) opposed immunity.
The article did not seem to give any data on false positives. A story here has Coverity claiming a 10% false positive rate. But there is no independent confirmation. It would also be interesting to know how hard it is to prove a false positive vs. how hard to fix a true positive. In other words, it it worth Coverity's time to further reduce the false positive rate.
Microsoft didn't invent anything in their patent portfolio. They either bought patents from other companies, or acquired companies which had patents, or simply paid people to invent things for them. Some of those people they paid to invent things used to be paid by IBM until IBM decided it could no longer afford them.
Slashdot Mirror
For me, the kicker is the last paragraph. The likely use is to replace bikes and pedestrians in the Third World, not cars in America.
CMS Pipelines did some of this. Not with structured data, nor literally 2D, but you could split and combine flows, and code feedback loops.
Not so funny to those of us who have endured a DVD-resolution feature-length film on the big screen.
Isn't anyone bothered by government asking commentators to "sign a non-disclosure agreement" about a proposed law disturbing?
How does the government prosecute someone who broke the law? Make the jury sign NDAs? Or maybe use a military court?
People in California should get a chuckle over this. State universities (and a lot of other state agencies) currently have a furlough program in place. Workers are paid some fraction of a salary and work the commensurate time. I think it's 90% for universities. So what used to be a 30-lecture class is now 27 lectures as the professor must take three "furlough" days. That's all the state can afford.
Somehow the number of credits remains the same. I'm not sure how that works. Maybe the state thinks the whole "furlough" business will be over before the accreditation agencies notice.
Original article also compares a Peta of floating ops per *second* to an Exa of byte "processing and storing" per *day.* Journalism profs should save that article for class discussion.
The base 10 units for disk drive size were established by IBM, which made the first disk drive. (That drive held 5,000,000 characters, replacing 62,500 punched cards. Each punched card had held 2^6.3219 characters). They also produced base 10 computers. The IBM 1401 which was available with 4K (4000 characters, or digits if you were doing arithmetic) to 16K of memory.
The question of when, if ever, drive manufactures should have made the switch from base 10. Perhaps the introduction of fixed-block disks, which have binary block lengths (512) and cannot be user-formatted to say, 1000 byte blocks. would have been a good time. But that would have confused things too.
An image with more contrast (greater average difference between adjacent pixels) probably has more detail. But compressability, as has already been noted, is probably just as good a measure.
Yes, everything sounds like AS/400. And System/38 before that. Which was sometimes known as "baby FS," where FS was IBM's Future System project. In order to contains every object in a single address space, FS was to have an 80-bit address space. FS was canceled in 1975 (per Wikipedia). System/38 had a 48 address space (or rather, object space), and AS/400 increased it to 64, or maybe 65. System/38 had two layers of microcode: Horizontal (like other IBM machines) and Vertical (which was the VM interpreter for the instruction set exposed to end users). But few customers used the VM directly. They just programmed in RPG or COBOL, or bought program suites written in RPG or COBOL.
Not having heard of Fixstars before, I'm not sure what to make of this. Their website reports that it is "the pioneering company of the Cell Broadband Engine," presumably leading the way for later entrants such as IBM, Toshiba, and Sony.
I remember TerraSoft mostly for overpriced hardware obviously intended for developers with a corporate checkbook behind them. I would look at the prices and decide that Apple was the cheaper source for PowerPC systems.
How about Fixstars? Looking at the web store, I can get Sony PS3 ($450), GigaAccel PCIe card (no price or delivery date listed), a fully populated IBM BladeCenter Chassis ($170,800), or a YDL PowerStation with "Quad-core 2.5GHz IBM 970MP CPUs" ($1895).
The YDL Powerstation sounds interesting and affordable. "Quad-core CPUs" (plural!) for $1895. How many quad-core CPUs, exactly? There is a "learn more" link, but really nothing more there. Less, in fact. That's about the level of detail I used to see from TerraSoft.
Read the link carefully. It is not SPC-1, nor modified SPC-1. It is 4K IO in a 70/30 read/write mix.
Maybe that someone was laughing too hard to care.
You were misled by a terrible headline. The 0.557 petaflop computer is the fastest *for open science.* Roadrunner, at Los Alamos, tops the list. It does 1 petaflop.
In short, beware of tools which produce graphs. Or worse, pie charts.
My story: management spent 25,000 on a malloc checker which produced pie charts. Our code didn't use malloc - at least not directly - so the code tested clean. Waste of money, but management liked the pie charts.
On the other hand, lint, gcc warnings (use more than -Wall, which is not really "all"), and Coverity are all good things to use.
Just what I was wondering: "Maybe, because they hand-code everything, they will pass the validation that all the fancy tools fail at so badly." Anyway, they are not alone. Here are the error couns for the Fortume top 20 companies (top of the Fortune 1000 list) manage on the w3c validator:
53 walmart.com
36 exxon.com
26 chevron.com
33 gm.com
76 conocophillips.com
0 ge.com
29 ford.com
52 citigroup.com
105 bankofamerica.com
26 att.com
28 www.berkshirehathaway.com
8 jpmorganchase.com
148 aig.com
55 hp.com
0 ibm.com
144 valero.com
2 verizon.com
180 mckesson.com
5 cardinalhealth.com
1082 www.goldmansachs.com
The RAM case has exceptions, too. The IBM 1401 has 1000 to 16000 bytes, er, characters of RAM, depending on the model.
Don't forget that the loan shark gets a controlling interest in the company. What are the chances the shark (acting as SCO) is going to ask itself (SNCP) for millions of dollars to pay a judgment to Novell? If/when that happens, the shark needs to ask itself whether its worth the risk, even at 21%. It only makes sense if SNCP really believes that SCO can get the Novell judgment reversed on appeal and win some money from IBM. That could be a big, risky bet.
The deal appears to be that SNCP is committed to loaning SCO $95 million, but only once. If the SNCP loans SCO $95 million on Monday and SCO repays the loan on Tuesday, it appears that the commitment has been fulfilled.
The parent is the vote on the final bill. Here is the vote on the Dodd-Feingold amendment. A "Yea" is a vote for immunity. In my state, Diane Feinstein (D-CA) continued her near-perfect voting record against the Bill of Rights. Barbara Boxer (D-CA) opposed immunity.
[BART] vehicular units weigh far too much for the task
...
There was absolutely nothing wrong with the Key Line and the Red Car Line
Don't know about the Key System, but BART cars (about 30 tons) weight about 1/2 as much as Red Cars (about 60 tons).
The article did not seem to give any data on false positives. A story here has Coverity claiming a 10% false positive rate. But there is no independent confirmation. It would also be interesting to know how hard it is to prove a false positive vs. how hard to fix a true positive. In other words, it it worth Coverity's time to further reduce the false positive rate.
"Deprecate" is also technology jargon that means "to mark as obsolete."
I'm not the first one to observe this, but ... how can you obsolete an element of a standard when it has never been part of the standard?
Microsoft didn't invent anything in their patent portfolio. They either bought patents from other companies, or acquired companies which had patents, or simply paid people to invent things for them. Some of those people they paid to invent things used to be paid by IBM until IBM decided it could no longer afford them.
New York is about as far east as you can go before being devoured by sea serpents.
For that matter, ask Google how may radians in a cycle. Do No Evil Inc. hasn't bothered to fix that in the year or so since it was reported to them.