When I asked the author about the power draw (in between jibes about his new found/. fame), he stated that the power draw under heavy load was more like 25 watts. Idling, the AC power draw runs about 10 to 11 watts (figure ~3w disapated in the PS brick, leaving maybe ~8w for the mini).
A couple of things to bear in mind, the nominal AH rating for batteries is often given at the 20 hour discharge (e.g. a 1A draw for a 20 AH battery) and the battery pack may not have been delivering the full 4 AH.
One reason for picking the Mac mini was it had a good performance per watt metric and it was reasonably cheap. It has about twice the power consumption of a Soekris x86 system, with about a 4X higher clock for about 2X the price of the Soekris.
If you are collecting more than 4G of data in one session, well YMMV - but 4G is a LOT of data, perhaps consider your approach.
My recent forays to Crucial show 4 sticks of 2GB reg/ecc PC2700 DDR memory will set one back a bit over $3k. For 8 to 16GB of data, the most economical route would be a dual Opteron box, things start getting expensive above 16GB.
Finally some c compilers got around even that. When you compiled you would tell the compiler what model to use. I think your options where usually "tiny" 64k of data and code max could be made into a com file, "Small" 64k of code and 64k of data, and "large" which broke the 64k limits. I think then came "huge" which I can not remember if it broke the data structure limits or if that ran in protected mode.
From the top of my head - no guarantee that my memory is accurate.
"tiny" was were the code and data segments were the same - and the intent was for converting into.com files. The.com format was essentially a translation of the CP/M executable format. SCP was planning on a multiuser version of 86-DOS and it would have very easy to switch between processes - as long as they stayed to one segment.
"small" was 64K for the data, IIRC it was possible to have multiple segments for the code (the 8086 supported long (segment:offset) jumps.
"large" allowed for more than one data segment, though an object (i.e. array) had to fit in one segment. MS-Fortran, for example, allowed named common blocks to reside in a different segment than the default data.
"huge" allowed for arrays to exceed 64K, the code set up far pointers to address members of the array.
Sure glad I don't have to worry about segments anymore. OTOH, OS/2 1.x did a good job of making use of segments for memory protection - any attempt at accessing memory outside of the assigned block would result in an immediate halt.
But not because of backwards compatibility issues so much as bad code, written by bad coders.
IMHO, the goodness of open source code comes more from people trying to port the code to other platorms than from the "millions of eyeballs" looking over the code.
I would put front porch evangalist and witch doctor Wayne Green before them.
Talk about a bad flashback from the 60's (and 70's...). Never Say Die (from W2NSD/1). Was really surprised to seem him at an ARRL convention in the early 80's...
Besides starting BYTE (which was sort of a spin-off of 73 magazine), Wayne was one of the people pushing the Kansas City standard for cassette storage. That made it much easier for people to exchange data.
A truly colorful character.
What I consider to be the dawn of the personal computer age was the "Mac's Service Shop" column in PopTRonics (column originally ran in Electronics World) in the march or April 1972 issue - the focus of the column was the HP-35. It was clear that world was about to change.
The first "West Coast Computer Faire" was also quite a revelation - nothing since then has had the same impact. Also depressing on how few companies present at the Faire are still in business.
OS/2 v1.x made good use of the segment registers on the 80286. Heard stories that the MS programmers preferred using OS/2 for writing Windoze code - memory access violations (e.g. going beyond the requested memory segment) would result in an immediate halt to execution with a pointer to where the un-allowed memory access occurred.
In making OS/2 2.0 portable to other architectures, IBM and MS went to a flat memory model which ruled out using the segment registers as Intel intended.
Our school's computer classes were given access to the SINGER computer that was setup to run BART.
I don't recall hearing abuot a Singer computer used for running BART, but they did have quite a collection of rare hardware. One example was that they had three of the four of a certain model of Philco computers in existence (ca 1975). They also had some Westinghouse Prodac 2000 boxes.
I've also heard that there was a small bug in the simulation program that led them to think capacity was going to be higher than in the real world.
FWIW, first time I rode BART was when the Richmond line was opened on Jan 29, 1973 - was a freshmen at the big U at the time.
There certainly is. And the so-called "left" doesn't tend any more or less toward this than the so-called "right". Certain groups have at different times over the last sixty years been more likely to go with the latter thing you mention than the former. This hasn't had to do with where they lie on the supposed political spectrum. It has to do with who, at that moment in time, has held enough power to be able to abuse it.
Last 60 years?? As if government sponsored censorship didn't exist in the US before 1945.
FDR thought it was perfectly OK to use the IRS to intimidate people who spoke out against him. To say nothing about the tens of thousands of US citizens who were sent off to internment camps just for being children of Japanese immigrants.
Go back to the period of 1917-1920... It was illegal under the Wilson administration to say anything negative about the war effort - even though the US involvement in WW1 was a much bigger mistake in the long run than the current mess in Iraq - the US and the world would have been a much better place if we stayed out.
David Koresh reference...hoarding unregistered weapons
Most jurisdictions in the US do not normally require weapons to be registered. If the FBI really wanted to arrest Koresh, they could have done so on one of his morning jogs.
You don't have to look into the future to see how worries of copyright infringement suits stifle technology. Ampex had the capability of making a home use VTR (that's Video Tape Recorder) in the 1970's, but they were concerned about the major studios suing them for promoting copyright infringement. Ampex figured that the Japanese companies had less to lose, since judgements would have been collected from the import arm, not the main corporation.
If you're using HFSS, then the price of a SB2500 can't be that bad by comparison - you can get 2GB DIMM's for it, so 16 GB is possible in a two processor system. The latest 2500's are shipping with the 1.6GHz US-IIIi.
Lessee, call doesn't get through? Sounds like Stuff*Mart will be going after your VoIP provider, not your ISP. If you have a residential internet connection, your ISP may trot out the T&C which says the service is for recretional use and if you want business class service, then you need a business class connection.
Then again, I wouldn't be surprised if Stuff*Mart is too cheap to spring for a decent ISP.
Nonsense. There were plenty of real OSs around at the time, running on similar processors. The prime example is CP/M. Which, if you had bothered to follow the discussion, you would already know about, since the lawsuit is over whether QDOS was a "slapdash clone" of CP/M. Which, in point of fact, it was.
86-DOS was as much of a clone of CP/M as Linux was a clone of SVR4 Unix.
Patterson cloned the assembly langauge API of CP/M and provided a tool to translate CP/M Z-80 asm to SCP's asm for 8086. There were some significant differences, a function call on CP/M was a jump to location 00, 86-DOS used the INT-21H (please note that Patterson was smart enough to pay attention to the INT's below 20H that were reserved by Intel) - CP/M compatiblity was maintained by putting a call to INT-21H at location 00. Software vectored interrupts did not exist on the 8080.
Another major change was the file system - using the FAT from Microsloth Disk BASIC - and changing the bytes in the directory from a bit map to file size, date and time stamp. Text files created by edlin would have only 1 control-z at the end - where CP/M editors (e.g. Vedit) would pad the files with control-z's to the end of a 128 byte block.
COMMAND.COM was a very different beast from CP/M's command interperter - much of the PIP funcionality built in. Batch files were more sane under 86-DOS than CP/M. TSR's were another feature available under 86-DOS and not CP/M.
The best OS of tthe time??? One former co-worker acerbic commentwas that the world would have been a much better place had IBM chose OS-9 to run on the 68K for their PC. Similarly, the world may have been a different place had Zilog started sampling Z-800's (that's 800 not 8000) in late 1979 and went into production in 1980.
The standard "home" inkjets don't have PostScript, especially the cheap, rock-bottom-price Lexmarks. I wasn't talking about Lexmark's laser printers, which may be fine (I have never used one).
I was referring to the Lexmark laser printers - bought a C510 a few days ago to replace a Xerox NC60 (which Xerox no longer supports - sigh). The documentation for the C510 specifically states that Linux is supported. The competing HP's are basically PCL only, the PostScript support for the competing Xerox is in the "driver".
The beauty of a PostScript printer is that getting the C510 up and running was simply a matter of connecting the parallel cable to the printer. It just simply worked.
OTOH, my wife has had pretty good luck with the HP 3in1.
Oh, I also picked HP because they have very good Linux support and Lexmark's Linux support sucks or is non-existent.
Interesting, I just bought a Lexmark over an HP because of better support for UNIX (i.e. Linux, Solaris) on the part of the Lexmark. It has PostScript (albeit using non-Adobe code), whereas comparable HP's don't.
I seem to recall the floppy ROM's showed up in early 1977. Could be wrong as the first copy I picked up was in early 1977. It was a pretty decent magazine at the time.
This could set an ugly precedence where an individual sells a product, a corporation sees a small business selling it, they decide to patent the product or trademark the term, and then turn around and sue so that the individual can no longer start another business again because they owe the Corporation for the rest of their life.
Ummm, have you heard the story about a small Chicago area ISP that trademarked "Internet Explorer"? While not an exact parallel with the Ultracade/Mame mess, it does bring up a history of corporations squashing the little guy.
I'd be willing to bet that they are related to Sprites - guessing the potential between the top of the thunderclouds and E-layer of the ionosphere could be tens of millions of volts. The mean free path of the particles may be long enough that they can pick up some significant energy.
On a related note, in the early 1980's, QST was reporting a very strong correlation between thunderstorm activity and sporadic E-layer propagation. Those reports came to mind when first reading about sprites over ten years later.
or simply decides to stop supporting their own format.
An all too frequent occurence - happened to a few co-workers when Lotus stopped supporting Manuscript. One of them commented that some of the stuff he wrote in the early 1990's is basically unreadable, but the stuff he did in TeX in the 1980's is maintainable.
One of the smartest things Sun did in regards to StarOffice/OOo was to GPL the core engine (meaning that the files should be readables decades from now) and documenting the file format.
The last type is connected to a network that was built by private companies with private sector dollars. That network is "slightly" regulated in that the cable company is given a monopoly on the township for a limited time span.
I don't think I've ever seen a cable company that really paid for the whole cost of their network. If you're cable is on a pole line, the they are most likely using pole lines built and paid for by the electric and phone companies. This doesn't include the easement for the pole lines. Similarly, if the cable is underground, they make use of easements for their right of way. They may have paid to put the cable there, but they most likely DO NOT own the right of way.
Now along comes VoIP. It flaunts its non-common carrier status. Vonage got an FCC ruling (now being appealed by state regulators) that it is not a telephone company subject to common carrier rules (and taxes).
IMHO, Vonage can't have it both ways. If they don't want the regulations or taxes of a normal telco, then they don't get the protecions that a normal telco should get.
It wouldn't be all that difficult for an ISP to make the QoS for Vonage's packet's really lousy. The flip side of this is if a customer wants good QoS for VoIP, then that customer should expect to pay for it.
Depending on your definition of neglegence, this either clearly wasn't (wasn't any worse than typical businesses) or could have been (a known risk which best practices clearly say not to do).
From what I've heard, the break-in took place in a building that did have a fair amount of security associated with it. My guess is that SAIC will be considerably more paranoid with the data after this incident - especially in regards to physical security. The corporate culture is pretty much like a start-up, with a lot of emphasis on keeping overhead low.
ISTR hearing about a computer being stolen from UCSD with similar types of data - with the central UCSD campus about 0.5 mile from where the SAIC break-in occured.
I knew a few people who were involved with the company when it was SAI - and rember hearing stories (ca 1976) about how it was metastizing across La Jolla from the old corporate headquarters in La Jolla - which was originally Scripps Hospital then Scripps Clinic.
One the key points of SAI stock was that you didn't have to sell it when you left the company.
Slashdot Mirror
A couple of things to bear in mind, the nominal AH rating for batteries is often given at the 20 hour discharge (e.g. a 1A draw for a 20 AH battery) and the battery pack may not have been delivering the full 4 AH.
One reason for picking the Mac mini was it had a good performance per watt metric and it was reasonably cheap. It has about twice the power consumption of a Soekris x86 system, with about a 4X higher clock for about 2X the price of the Soekris.
The 640K limit was from the memory map of the IBM PC, 86-DOScould support almost the full meg of the 8086 address space.
My recent forays to Crucial show 4 sticks of 2GB reg/ecc PC2700 DDR memory will set one back a bit over $3k. For 8 to 16GB of data, the most economical route would be a dual Opteron box, things start getting expensive above 16GB.
From the top of my head - no guarantee that my memory is accurate.
"tiny" was were the code and data segments were the same - and the intent was for converting into .com files. The .com format was essentially a translation of the CP/M executable format. SCP was planning on a multiuser version of 86-DOS and it would have very easy to switch between processes - as long as they stayed to one segment.
"small" was 64K for the data, IIRC it was possible to have multiple segments for the code (the 8086 supported long (segment:offset) jumps.
"large" allowed for more than one data segment, though an object (i.e. array) had to fit in one segment. MS-Fortran, for example, allowed named common blocks to reside in a different segment than the default data.
"huge" allowed for arrays to exceed 64K, the code set up far pointers to address members of the array.
Sure glad I don't have to worry about segments anymore. OTOH, OS/2 1.x did a good job of making use of segments for memory protection - any attempt at accessing memory outside of the assigned block would result in an immediate halt.
IMHO, the goodness of open source code comes more from people trying to port the code to other platorms than from the "millions of eyeballs" looking over the code.
Talk about a bad flashback from the 60's (and 70's...). Never Say Die (from W2NSD/1). Was really surprised to seem him at an ARRL convention in the early 80's...
Besides starting BYTE (which was sort of a spin-off of 73 magazine), Wayne was one of the people pushing the Kansas City standard for cassette storage. That made it much easier for people to exchange data.
A truly colorful character.
What I consider to be the dawn of the personal computer age was the "Mac's Service Shop" column in PopTRonics (column originally ran in Electronics World) in the march or April 1972 issue - the focus of the column was the HP-35. It was clear that world was about to change.
The first "West Coast Computer Faire" was also quite a revelation - nothing since then has had the same impact. Also depressing on how few companies present at the Faire are still in business.
Floobydust??? Must be a Bob Pease fan.
OS/2 v1.x made good use of the segment registers on the 80286. Heard stories that the MS programmers preferred using OS/2 for writing Windoze code - memory access violations (e.g. going beyond the requested memory segment) would result in an immediate halt to execution with a pointer to where the un-allowed memory access occurred.
In making OS/2 2.0 portable to other architectures, IBM and MS went to a flat memory model which ruled out using the segment registers as Intel intended.
I don't recall hearing abuot a Singer computer used for running BART, but they did have quite a collection of rare hardware. One example was that they had three of the four of a certain model of Philco computers in existence (ca 1975). They also had some Westinghouse Prodac 2000 boxes.
I've also heard that there was a small bug in the simulation program that led them to think capacity was going to be higher than in the real world.
FWIW, first time I rode BART was when the Richmond line was opened on Jan 29, 1973 - was a freshmen at the big U at the time.
Last 60 years?? As if government sponsored censorship didn't exist in the US before 1945.
FDR thought it was perfectly OK to use the IRS to intimidate people who spoke out against him. To say nothing about the tens of thousands of US citizens who were sent off to internment camps just for being children of Japanese immigrants.
Go back to the period of 1917-1920... It was illegal under the Wilson administration to say anything negative about the war effort - even though the US involvement in WW1 was a much bigger mistake in the long run than the current mess in Iraq - the US and the world would have been a much better place if we stayed out.
David Koresh reference...hoarding unregistered weapons
Most jurisdictions in the US do not normally require weapons to be registered. If the FBI really wanted to arrest Koresh, they could have done so on one of his morning jogs.
You don't have to look into the future to see how worries of copyright infringement suits stifle technology. Ampex had the capability of making a home use VTR (that's Video Tape Recorder) in the 1970's, but they were concerned about the major studios suing them for promoting copyright infringement. Ampex figured that the Japanese companies had less to lose, since judgements would have been collected from the import arm, not the main corporation.
If you're using HFSS, then the price of a SB2500 can't be that bad by comparison - you can get 2GB DIMM's for it, so 16 GB is possible in a two processor system. The latest 2500's are shipping with the 1.6GHz US-IIIi.
Then again, I wouldn't be surprised if Stuff*Mart is too cheap to spring for a decent ISP.
86-DOS was as much of a clone of CP/M as Linux was a clone of SVR4 Unix.
Patterson cloned the assembly langauge API of CP/M and provided a tool to translate CP/M Z-80 asm to SCP's asm for 8086. There were some significant differences, a function call on CP/M was a jump to location 00, 86-DOS used the INT-21H (please note that Patterson was smart enough to pay attention to the INT's below 20H that were reserved by Intel) - CP/M compatiblity was maintained by putting a call to INT-21H at location 00. Software vectored interrupts did not exist on the 8080.
Another major change was the file system - using the FAT from Microsloth Disk BASIC - and changing the bytes in the directory from a bit map to file size, date and time stamp. Text files created by edlin would have only 1 control-z at the end - where CP/M editors (e.g. Vedit) would pad the files with control-z's to the end of a 128 byte block.
COMMAND.COM was a very different beast from CP/M's command interperter - much of the PIP funcionality built in. Batch files were more sane under 86-DOS than CP/M. TSR's were another feature available under 86-DOS and not CP/M.
The best OS of tthe time??? One former co-worker acerbic commentwas that the world would have been a much better place had IBM chose OS-9 to run on the 68K for their PC. Similarly, the world may have been a different place had Zilog started sampling Z-800's (that's 800 not 8000) in late 1979 and went into production in 1980.
I was referring to the Lexmark laser printers - bought a C510 a few days ago to replace a Xerox NC60 (which Xerox no longer supports - sigh). The documentation for the C510 specifically states that Linux is supported. The competing HP's are basically PCL only, the PostScript support for the competing Xerox is in the "driver".
The beauty of a PostScript printer is that getting the C510 up and running was simply a matter of connecting the parallel cable to the printer. It just simply worked.
OTOH, my wife has had pretty good luck with the HP 3in1.
Interesting, I just bought a Lexmark over an HP because of better support for UNIX (i.e. Linux, Solaris) on the part of the Lexmark. It has PostScript (albeit using non-Adobe code), whereas comparable HP's don't.
I seem to recall the floppy ROM's showed up in early 1977. Could be wrong as the first copy I picked up was in early 1977. It was a pretty decent magazine at the time.
Ummm, have you heard the story about a small Chicago area ISP that trademarked "Internet Explorer"? While not an exact parallel with the Ultracade/Mame mess, it does bring up a history of corporations squashing the little guy.
On a related note, in the early 1980's, QST was reporting a very strong correlation between thunderstorm activity and sporadic E-layer propagation. Those reports came to mind when first reading about sprites over ten years later.
An all too frequent occurence - happened to a few co-workers when Lotus stopped supporting Manuscript. One of them commented that some of the stuff he wrote in the early 1990's is basically unreadable, but the stuff he did in TeX in the 1980's is maintainable.
One of the smartest things Sun did in regards to StarOffice/OOo was to GPL the core engine (meaning that the files should be readables decades from now) and documenting the file format.
I don't think I've ever seen a cable company that really paid for the whole cost of their network. If you're cable is on a pole line, the they are most likely using pole lines built and paid for by the electric and phone companies. This doesn't include the easement for the pole lines. Similarly, if the cable is underground, they make use of easements for their right of way. They may have paid to put the cable there, but they most likely DO NOT own the right of way.
IMHO, Vonage can't have it both ways. If they don't want the regulations or taxes of a normal telco, then they don't get the protecions that a normal telco should get.
It wouldn't be all that difficult for an ISP to make the QoS for Vonage's packet's really lousy. The flip side of this is if a customer wants good QoS for VoIP, then that customer should expect to pay for it.
From what I've heard, the break-in took place in a building that did have a fair amount of security associated with it. My guess is that SAIC will be considerably more paranoid with the data after this incident - especially in regards to physical security. The corporate culture is pretty much like a start-up, with a lot of emphasis on keeping overhead low.
ISTR hearing about a computer being stolen from UCSD with similar types of data - with the central UCSD campus about 0.5 mile from where the SAIC break-in occured.
One the key points of SAI stock was that you didn't have to sell it when you left the company.
Starship????
Perhaps you meant Jefferson Airplane?
Kids nowadays can't get anything straight...