I agree that you should buy what is optimal for your situation (which may be different from other
people's situations). If you regularly use two
serial ports simultaneously, it will probably be
cheaper for you to buy those serial ports on the
motherboard if you are going to buy today.
By the way, Aten sells USB serial adapters for
$19.
That page also has a link to Centrix which, at $9 shipping per order, is remaindering USB serial adapters for $4 and USB parallel adapters for $2. I would not argue that that repesents an equilibrium price though.
The article says, "This mobo comes with an array of on-board multimedia and I/O features [...] including [...] Serial ATA RAID
(courtesy of a Silicon Image controller chip) [...]".
I don't see any indication from the article
whether the Intel motherboard that uses the
i845ge also provides Serial ATA. Also, I
would be interested in knowing how this Silicon Image chip is attached. For example, if it is
only connected by a 32-bit 33MHz PCI bus, then
it will only be able to transfer data across the
bus at 133 megabytes per second. No single disk
drive goes that fast, but if it has a bunch of
Serial ATA ports, it might be an issue. I saw
some posting on slashdot that said that most recent chipsets do not physically attach their IDE interfaces through
the PCI bus, but rather do something faster even though the devices logically look to the CPU like they are on the PCI bus.
I have provided numerous links to support my
factual claims. If you cannot provide any references to support your claims that "in terms of volume x86 makes up under 1% this year compare this to the ARM and MIPS world", "more bluetooth devices are made each day than 802.11b in a year", and "all new phones have Bluetooth", then there is no way to distinguish you from someone who just makes up false information. I have done enough research for my statements to provide links that directly show the numbers that support my claims. You have not done this with yours.
Anyhow, it's pretty apparent that you are unable to provide links that show the numbers that you claim. If your claims were true, you could easily find them by google and then you should have and would have posted them.
I'm sorry if I am no longer able to prioritize time to respond further to you if you do not post references or retractions for your aforementioned claims.
take a long hard look at the semiconductor world and in terms of volume x86 makes up under
1% this year compare this to the ARM and MIPS world which make up your Set Top Box, NIC,
and mobile phone
This report estimates sales of Arm, MIPS, SH and PowerPC chips in 2001 was 539 million units.
"For the first quarter of 2002, PC makers worldwide shipped 31.4 million PCs".
So, that's 125 million PCs/year. That's a lower
bound on annual x86 unit volume. Even if every PC were a uniprocessor, and even if that source's definition of "PCs" included every x86
application, from embedded to supercomputers, 125 million is still a lot more than 1% of 539 million.
Please
post references or retractions for your claims that "in terms of volume x86 makes up under 1% this year compare this to the ARM and MIPS world",
"more bluetooth devices are made each day than 802.11b in a year", and "all new phones have Bluetooth" (even if you meant all new designs of cellular and cordless phones, a reference would still be in order).
it terms of volume more bluetooth devices are made each day than 802.11b in a year
"In 2Q02, 802.11b hardware shipments grew 15%, totaling 3.9 million units." 3.9 million x 4 quaters/year = 15.6 million/year
if volumes are at least flat for the next
three quarters. x 365 = 5.7 billion bluetooth devices produced per year? I read in a magazine
recently that the total number of computers ever produced is expected to reach one billion in the next few months.
Could you please post a reference for your claim
that "more bluetooth devices are made each day than 802.11b in a year" or a retraction? Thanks in advance.
I don't know what protocols the two companies
discussed in the article use, but there are
GPL'ed implementations of mesh routing
protocols linked from these pages about
MobileMesh, and
Ad Hoc On Demand Distance Vector.
802.11b supports independent physically overlapping networks. Each network has a
name (an "essid"). For example, if you had
a network name "starbucks" and another named
"free", and were using GNU/Linux, you could do:
For more efficient transmission, you can
even program your access points to use different frequencies. There are twelve overlapping frequency bands used 802.11b, which provide for
three or four completely independent networks.
Attempting to associate with a network named
"Any" or "" will usually result in associating the
network with the strongest signal, depending on your driver and card. This is also true in
other operating systems.
Perhaps it's more of a plug than a disclaimer, but I should mention I'm involved in
LANRoamer,
an open source system that you can use to sell
passers-by access to your wireless network and
other participating networks.
I'd really like to know whether you had useful patches for apache, bsd, or xfree86 that you held back because of the non-GPL license. Heck, I'd like to see anyone come up and claim that. Those making the most noise about licenses are usually those with nothing to contribute.
The person you responded to is an
extremely productive free software developer.
I don't know if he wants to be identified too easily, so I'll keep this list short:
He wrote a replacement for the graphics subsystem for emacs under X windows (don't think it was ever incorporated, alas). I saw a demo of a chess front end in an X window.
He is a principal coauthor of a free-for-noncommercial-use encrypted telephone program.
He was a paid developer for the Free Software Foundation. You probably ran his code the
last time you ran gcc, emacs or a number of
other GNU utilities.
We give a deterministic
O((log n)**12) time algorithm for testing whether a number is prime.
[Sorry, the Slashdot filter does not allow me to
superscript the 12.]
The algorithm takes O(log2(n)**12) time, where n
is number being factored. If we optimistically assume that this algorithm can test the primality
of a 16-bit number in one microsecond, then here is how long it would take to test time primality of some larger numbers.
2**12 times as long for a 32-bit number = 4096 microseconds = 4 milliseconds,
4**12 times as long for a 64-bit number = 16,777,216 microseconds = 16 seconds,
8**12 times as long for a 128-bit number =
68,719,476,736 microseconds = 68,719 seconds = 19 hours,
16**12 times as long for a 256-bit number =
281,474,976,710,656 microseconds = 9 years.
I don't know what a realistic base time for this
algorithm really would be, and I don't know
where the cross over point against existing exponential time deterministic primality testing
algorithms would be, but at least this provide a sense of how log2(n)**12 grows.
I wrote: "It is true that the our days are curerntly slightly shorter than 86,400 seconds [...]". I meant longer. The Earth's rotation as decelerated a bit since 1820, not accellerated. Sorry for any confusion.
That sentence can mean many different things. It is true that the our days are curerntly slightly
shorter than 86,400 seconds, and 365 times that difference adds up to about one second, but that has nothing to do with whether the days are now getting longer or shorter. The duration of one
rotation of the Earth on its axis is not getting shorter by one second every
year. It is getting longer by 1.4 milliseconds
every century, and I would guess that that deceleration will be weaker as the Earth slows down and the moon gets farther away.
The earth's rotation is slowing anyway. This is the reason that they insert those "leap seconds" every few years to compensate
for the lost time.
At first, I did not believe that such a small
change could account for the leap seconds, but
you're right
:
Through the use of ancient observations of eclipses, it is
possible to determine the average deceleration of the Earth to be roughly 1.4 milliseconds per day per century.
[...] Over the course of one year, the difference accumulates to almost one second, which is compensated by the insertion of a leap
second into the scale of UTC with a current regularity of a little less than once per year. Other factors also affect the Earth, some in unpredictable
ways, so that it is necessary to monitor the Earth's rotation continuously.
In order to keep the cumulative difference in UT1-UTC less than 0.9 seconds, a leap second is added to the atomic time to decrease the difference
between the two. This leap second can be either positive or negative depending on the Earth's rotation. Since the first leap second in 1972, all leap
seconds have been positive and there were 22 leap seconds in the 27 years to January, 1999. This pattern reflects the general slowing trend of the
Earth due to tidal braking.
When I ordered Covad DSL ($50/month for 384/128kbps), the salesperson was very clear that sharing one's line to sell wireless access to one's neighbors was perfectly OK with them and something that they regarded as a competitive advantage of their service.
DSL has less media sharing and is easier to upgrade on an individual basis. This may be
why DSL providers in my experience generally seem
to be ambivalently neutral to definitely positive about wireless access sharing, while cable modem
providers have generally been quite concerned
and proactive about
any kind of bandwidth hogging scenarios (not
just wireless sharing).
Maybe it is to NATO's advantage that this information is public.
NATO countries often
disagree on Balkan politicies and try to
put different political spins on the reliability of information
driving these decisions. Plus, others may believe
that the information that they're getting about the Balkans is
being filtered in some politically-oriented way. By being
public about this relatively innocuous data, it makes NATO
come off as presenting "just the facts", avoid creating some
other little opportunities for member countries to argue about
what the facts are, and eliminates some
politics about which photos to release.
Of course, if it started to look like fighting was imminent,
and the short term knowledge of what NATO saw or missed during
yesterday's flight becomes dangerous information,
then I would expect then I would expect NATO to secure the
video feed, or perhaps even at any point when a few more
fighting words than usual were emerging from Serbia.
USB host mode available on Casio E200 and NEC P300
on
Toshiba e740 Pocket PC
·
· Score: 2
The Casio Cassiopeia E200 and NEC MobilePro P300 can be USB hosts.
The Cassiopeia E200's cradle has both a USB client connector
(for syncing) and a USB host connector. You
can also buy a bottom lip for the Casio
and that has the USB host connector, so you could drive a tree of
USB devices while mobile. I saw a picture of it in a magazine review. I
believe USB current out from the lip device
is limited to 100mA (the USB specification
allows current limitiations of 100mA to 500mA).
I do not know if the NEC MobilePro P300's USB host mode requires it to be
plugged into its cradle.
It's not possible to approve LGPL without opening up a hole that allows J. Random Megacorp to
make an LGPL licensed librhpatents.so, which lets them use the patents with closed source proprietary apps.
One could easily license the patents for
use where the program consists entirely of free software. As soon as you link in a proprietary byte of code, the resulting work is not covered by the permission grant. Just about any reasonable definition of free software would do, since accidentally licensing demo-ware or something should not be a problem if the patent is for defensive purposes only. It would not even be necessary to create a list of approved copyright notices.
Theoretically, this should be easier to write under patents than under copyrights, since patents restrict end use, while copyrights are only supposed to restrict a few actions, like copying and live performance.
I am not a lawyer. So, please do not use this message as legal advice.
Chapter two of The Nature of Copyright: A Law of Users Rights by L. Ray Patterson and Stanley W. Lindberg discusses the origin of the word copyright ("Copyright In The Beginning: A Publishers Right"). According to
the book, copyright does not come from a "right to copy", but rather from copie, which was a noun meaning "manuscript"
(not a verb meaning "to duplicate"). The title of the copie
was recorded in a registry maintained by the Stationers Guild in the late 1500's to mid 1600's to record who they gave a monopoly to to publish a given book. This right lasted in perpetuity, but was only legally binding on members of the guild,
which is apparently where the big push came from to turn this arragement into a law in order to
prevent non-members from breaking the monopoly.
Authorship of the work apparently did not matter.
Instead, a guild member could claim a monopoly in this registry for six pence (no idea what that translates to in today's dollars).
This is the correct spelling of the
adjective bizarre
(meaning weird, as opposed to the noun bazaar, which is
a type of market).
Your use of the adverb euphemistically
is also wrong. Webster.com defines euphemism as
"the substitution of an agreeable or inoffensive expression for one that may offend or suggest something unpleasant; also: the expression to be substituted" (my emphasis).
If you don't need the latest version of Pocket PC, there are some clearance prices available. I saw the often overlooked Fujitsu AudioVox Maestro at CompuUSA for $287 a few weeks ago. It has a CF slot, 320x240 color display, ARM processor, and 16MB of RAM.
More recently, Fry's has had the Casiopeia E-125 on clearance for $200-225. The E-125 has a CompactFlash slot, a 150MHz MIPS R4000, 320x240 16-bit color display, USB and serial base unit. There is some magic file name my Casiopeia E-105 runs if it is booted with a CompactFlash card in its slot, which is handy for booting Linux, although I have only booted a precompiled binary that I downloaded.
On the other hand, I should warn you that it looks like Linux-MIPS development has been pretty quiet for the past two years, and MIPS-based PDA's seem to be disappearing. Also, I think that you have to go through the base unit to plug it in a USB device, which impedes my interest in possibly using an E-125 as a USB analyzer.
The Socket
WL6000-320 802.11b CF card is supposed to use less power than other CF cards when idle, has the smallest antenna stub that I've seen and is CompactFlash type I, while most other cards are the of the thicker type II variety, which cannot plug into as many devices. To me, these features would justify the
$140 mail order price ($180 at Fry's), even though I worry that the small antenna stub might translate into a weaker antenna.
What has stopped me from buying, however, is the claim that this
card does not work under current Linux drivers
because
it uses a currently undocumented MAC chip, although the rest of the hardware is supposed to be standard Intersil Prism.
Does anyone know if this problem is still the case? If the hardware were documented I would probably feel comfortable enough buying the chip and adapting one of the Prism drivers to it. (One the other hand, I would use a different card or do without rather than run a proprietary driver.)
By the way, the SMC card the MicroCenter had for $140 looks identical to the Socket card from the outside. I would be intersted in knowing if they are the same, and if anyone has used the SMC card under Linux.
The 2900 series was a bunch of four bit wide components for building a processor, such as arithmetic logic unit and register file. You could arrange these four bit "slices" in parallel to build a processor of whatever bit width you wanted. A few components were later released in larger widths, to reduce component counts.
I believe that Three Rivers Computer in Pittsburgh, PA use Am2900 bit slice chips to
build a successor the the Xerox Alto workstation.
Carnegie-Mellon University had a bunch of these when I was in high school. The "PiRQ" (if I remember the spelling and capitalization) had a fourteen inch fixed hard disk, a portrait format bitmapped monochrome screen, a digitizer tablet, and a pretty slick looking enclosure.
The Am29000, on the other hand, was an urelated product that came much later. The Am29000 was the first mass produced single chip RISC processor.
It was a 32 bit processor with separate busses for instructions and data, so that you could use dual ported memory or some other scheme to arrange for data accesses not to break the streaming of instructions.
Maybe that can make He-gel or H2-gel and get the *solid* material to be lighter
than air...
If the aerogel has a density of 3 milligrams per cubic centimemeter, and air has a 1.2 milligrams per cubic centimer, then replacing all of the air in this aerogel with an imaginary massless gas would result in a density of 1.8 milligrams per cubic centimeter, still heavier than air.
However, with further engineering, perhaps someone will invent a solid lighter than air. At that point, I the floating cities that you imagine would probably become as common as floating cities made of blimps are today.
Slashdot Mirror
By the way, Aten sells USB serial adapters for $19.
That page also has a link to Centrix which, at $9 shipping per order, is remaindering USB serial adapters for $4 and USB parallel adapters for $2. I would not argue that that repesents an equilibrium price though.
I don't see any indication from the article whether the Intel motherboard that uses the i845ge also provides Serial ATA. Also, I would be interested in knowing how this Silicon Image chip is attached. For example, if it is only connected by a 32-bit 33MHz PCI bus, then it will only be able to transfer data across the bus at 133 megabytes per second. No single disk drive goes that fast, but if it has a bunch of Serial ATA ports, it might be an issue. I saw some posting on slashdot that said that most recent chipsets do not physically attach their IDE interfaces through the PCI bus, but rather do something faster even though the devices logically look to the CPU like they are on the PCI bus.
USB parallel port adapters cost $11 on pricewatch.
Anyhow, it's pretty apparent that you are unable to provide links that show the numbers that you claim. If your claims were true, you could easily find them by google and then you should have and would have posted them.
I'm sorry if I am no longer able to prioritize time to respond further to you if you do not post references or retractions for your aforementioned claims.
This report estimates sales of Arm, MIPS, SH and PowerPC chips in 2001 was 539 million units. "For the first quarter of 2002, PC makers worldwide shipped 31.4 million PCs". So, that's 125 million PCs/year. That's a lower bound on annual x86 unit volume. Even if every PC were a uniprocessor, and even if that source's definition of "PCs" included every x86 application, from embedded to supercomputers, 125 million is still a lot more than 1% of 539 million.
Please post references or retractions for your claims that "in terms of volume x86 makes up under 1% this year compare this to the ARM and MIPS world", "more bluetooth devices are made each day than 802.11b in a year", and "all new phones have Bluetooth" (even if you meant all new designs of cellular and cordless phones, a reference would still be in order).
"In 2Q02, 802.11b hardware shipments grew 15%, totaling 3.9 million units." 3.9 million x 4 quaters/year = 15.6 million/year if volumes are at least flat for the next three quarters. x 365 = 5.7 billion bluetooth devices produced per year? I read in a magazine recently that the total number of computers ever produced is expected to reach one billion in the next few months.
Poking around the internet, all I see are projections that perhaps bluetooth unit volumes will catch up to 802.11 this year. "It is not fair to run comparisons between Bluetooth and WLAN regarding the number of chips being shipped or expected to be shipped for either technology. Because of its price and the type of products where it is going into, Bluetooth will soon surpass 802.11 chip shipments but dollar volume might still be smaller for some time."
"Our annual Bluetooth study predicted Bluetooth chipset shipments would rise from close to zero in 2000 to just over 10 million units in 2001, and the year will end not far from this figure."
Could you please post a reference for your claim that "more bluetooth devices are made each day than 802.11b in a year" or a retraction? Thanks in advance.
Thank you, afidel, for trying to conduct some reasonably objective tests and posting the results.
Also, is a collection of links on mobile routing protocols, and mesh routing protocols that were originally designed for wired backbone links.
The awful 1995 movie Congo featured a gorilla wearing gloves that, in the movie's world, translated sign language into speech.
iwconfig wlan0 mode Managed essid starbucks
udhcpc --interface wlan0
or
iwconfig wlan0 mode Managed essid free
udhcpc --interface wlan0
For more efficient transmission, you can even program your access points to use different frequencies. There are twelve overlapping frequency bands used 802.11b, which provide for three or four completely independent networks.
Attempting to associate with a network named "Any" or "" will usually result in associating the network with the strongest signal, depending on your driver and card. This is also true in other operating systems.
Perhaps it's more of a plug than a disclaimer, but I should mention I'm involved in LANRoamer, an open source system that you can use to sell passers-by access to your wireless network and other participating networks.
The person you responded to is an extremely productive free software developer. I don't know if he wants to be identified too easily, so I'll keep this list short:
We give a deterministic O((log n)**12) time algorithm for testing whether a number is prime.
[Sorry, the Slashdot filter does not allow me to superscript the 12.]
The algorithm takes O(log2(n)**12) time, where n is number being factored. If we optimistically assume that this algorithm can test the primality of a 16-bit number in one microsecond, then here is how long it would take to test time primality of some larger numbers.
I don't know what a realistic base time for this algorithm really would be, and I don't know where the cross over point against existing exponential time deterministic primality testing algorithms would be, but at least this provide a sense of how log2(n)**12 grows.
I wrote: "It is true that the our days are curerntly slightly shorter than 86,400 seconds [...]". I meant longer. The Earth's rotation as decelerated a bit since 1820, not accellerated. Sorry for any confusion.
That sentence can mean many different things. It is true that the our days are curerntly slightly shorter than 86,400 seconds, and 365 times that difference adds up to about one second, but that has nothing to do with whether the days are now getting longer or shorter. The duration of one rotation of the Earth on its axis is not getting shorter by one second every year. It is getting longer by 1.4 milliseconds every century, and I would guess that that deceleration will be weaker as the Earth slows down and the moon gets farther away.
At first, I did not believe that such a small change could account for the leap seconds, but you're right :
When I ordered Covad DSL ($50/month for 384/128kbps), the salesperson was very clear that sharing one's line to sell wireless access to one's neighbors was perfectly OK with them and something that they regarded as a competitive advantage of their service.
DSL has less media sharing and is easier to upgrade on an individual basis. This may be why DSL providers in my experience generally seem to be ambivalently neutral to definitely positive about wireless access sharing, while cable modem providers have generally been quite concerned and proactive about any kind of bandwidth hogging scenarios (not just wireless sharing).
Of course, if it started to look like fighting was imminent, and the short term knowledge of what NATO saw or missed during yesterday's flight becomes dangerous information, then I would expect then I would expect NATO to secure the video feed, or perhaps even at any point when a few more fighting words than usual were emerging from Serbia.
The Cassiopeia E200's cradle has both a USB client connector (for syncing) and a USB host connector. You can also buy a bottom lip for the Casio and that has the USB host connector, so you could drive a tree of USB devices while mobile. I saw a picture of it in a magazine review. I believe USB current out from the lip device is limited to 100mA (the USB specification allows current limitiations of 100mA to 500mA).
I do not know if the NEC MobilePro P300's USB host mode requires it to be plugged into its cradle.
One could easily license the patents for use where the program consists entirely of free software. As soon as you link in a proprietary byte of code, the resulting work is not covered by the permission grant. Just about any reasonable definition of free software would do, since accidentally licensing demo-ware or something should not be a problem if the patent is for defensive purposes only. It would not even be necessary to create a list of approved copyright notices.
Theoretically, this should be easier to write under patents than under copyrights, since patents restrict end use, while copyrights are only supposed to restrict a few actions, like copying and live performance.
I am not a lawyer. So, please do not use this message as legal advice.
Authorship of the work apparently did not matter. Instead, a guild member could claim a monopoly in this registry for six pence (no idea what that translates to in today's dollars).
Your use of the adverb euphemistically is also wrong. Webster.com defines euphemism as "the substitution of an agreeable or inoffensive expression for one that may offend or suggest something unpleasant; also: the expression to be substituted" (my emphasis).
More recently, Fry's has had the Casiopeia E-125 on clearance for $200-225. The E-125 has a CompactFlash slot, a 150MHz MIPS R4000, 320x240 16-bit color display, USB and serial base unit. There is some magic file name my Casiopeia E-105 runs if it is booted with a CompactFlash card in its slot, which is handy for booting Linux, although I have only booted a precompiled binary that I downloaded.
On the other hand, I should warn you that it looks like Linux-MIPS development has been pretty quiet for the past two years, and MIPS-based PDA's seem to be disappearing. Also, I think that you have to go through the base unit to plug it in a USB device, which impedes my interest in possibly using an E-125 as a USB analyzer.
What has stopped me from buying, however, is the claim that this card does not work under current Linux drivers because it uses a currently undocumented MAC chip, although the rest of the hardware is supposed to be standard Intersil Prism.
Does anyone know if this problem is still the case? If the hardware were documented I would probably feel comfortable enough buying the chip and adapting one of the Prism drivers to it. (One the other hand, I would use a different card or do without rather than run a proprietary driver.)
By the way, the SMC card the MicroCenter had for $140 looks identical to the Socket card from the outside. I would be intersted in knowing if they are the same, and if anyone has used the SMC card under Linux.
The 2900 series was a bunch of four bit wide components for building a processor, such as arithmetic logic unit and register file. You could arrange these four bit "slices" in parallel to build a processor of whatever bit width you wanted. A few components were later released in larger widths, to reduce component counts.
I believe that Three Rivers Computer in Pittsburgh, PA use Am2900 bit slice chips to build a successor the the Xerox Alto workstation. Carnegie-Mellon University had a bunch of these when I was in high school. The "PiRQ" (if I remember the spelling and capitalization) had a fourteen inch fixed hard disk, a portrait format bitmapped monochrome screen, a digitizer tablet, and a pretty slick looking enclosure.
The Am29000, on the other hand, was an urelated product that came much later. The Am29000 was the first mass produced single chip RISC processor. It was a 32 bit processor with separate busses for instructions and data, so that you could use dual ported memory or some other scheme to arrange for data accesses not to break the streaming of instructions.
If the aerogel has a density of 3 milligrams per cubic centimemeter, and air has a 1.2 milligrams per cubic centimer, then replacing all of the air in this aerogel with an imaginary massless gas would result in a density of 1.8 milligrams per cubic centimeter, still heavier than air.
However, with further engineering, perhaps someone will invent a solid lighter than air. At that point, I the floating cities that you imagine would probably become as common as floating cities made of blimps are today.