Palm+bike, old, Handspring, circus stuff.
on
Palm on a Bicycle
·
· Score: 1
Gee, I feel like I'm missing something here. I *always* use my palm (both of them, actually) when riding a bike.
When I was a teenager, I knew a guy who could do a _handstand_ on his handlebars, but a handspring?!? Sounds like a good way to wipe out, get hurt and trash your bike to boot!
I've kind of given up on educating sysadmins about their Open Relays. I never received a single response out of over 50 incidents where I tried to contact them.
In one case, 210.217.41.84 was an open relay running SCO with an old sendmail. I tried to contact them at various email addies and got no response. I even got a Korean associate to try to contact an IT type at the company that owned the server; blank stares.
I checked that IP address 6 months ago, sendmail said "out of memory." I checked it today, its gone now. All this was over a period of at least a year; I wonder who finally pulled the plug and why...
I don't think the Chinese gonvernment much cares about what's going out their firewalls. In fact, they might consider it humorous that you [westerner] are being hassled from Chinese netblocks.
What they probably don't realise is that those "broken" servers, with pirated, unpatched, out-of-date versions of MickeySoft OS's that have open mail relays also have known, exploitable holes. These pirated OS machines could be hijacked from inside or outside of China and put to some really interesting uses.
I'll double the reply to your post which said you obviously don't work at corporate AR.
A close family member of mine works at HP and for several years was assigned to "The Wal-Mart Account." He often commuted (600 miles or so) to "lovely Benton-Hell" to handle the technical fallout of management decisions and other major crap. When corp Wal-Mart says to corp HP, "jump," HP says "I'm already jumping, do you want higher, to the left, to the right?" WM *always* plays the *big guys* (IBM, HP, Sun, others) off each other, so you gotta stay on their good side.
I have to wonder if Wal-mart's cost-cutting (retail and behind-the-scenes-techo-retail) is a part of HP's early adoptance of Linux when they've clearly had a top leading eunichs product for oh so many years.
Circumstantial evidence I've seen, plus WM's "cost-cutting," means I'll believe it if someone tells me WM has HP installing Linux on Sun boxen, talking to IBM iron.
Advertisers flirt with loss of Trademark protection.
Sorry, my WordStar fingers (using CTL key combos to cursor around) caused me to post before I got started.
The quote, "You don't pick up a facial tissue, you pick up a Kleenex." in the article caught my eye. People also often ask for a Band-Aid(tm) when they mean bandage. I don't know how hard Kimberly-Clark works to protect the Kleenex(tm) brand, but I'm sure they're quite close to losing it (not loosing it, damn it, loosing is not a word) all the time. There's a fine line between making your product name known in every household and making it a household word (or whatever the requirement is to lose your trademark protection).
I thought [Mike Fischer, vice president of entertainment marketing for Sega], a guy who oughta know better, was a bit irresponsible in making that statement.
You came very close to saying what I wanted to post.
I think the issue of patenting DNA is similar to that of software patents, to which I disagree. In fact, DNA *is* our software and the software of all life (as we know it). That little robosome machine that transcribes our software could be viewed as a sort of execution unit.
If someone writes a cool new app in C++, they might want to patent it (or some part of it). These days (at least in the US), they'd stand a good chance of being awarded a patent. Imagine a bio-engineer who "wrote" a cool new app in DNA, say, a cholesterol eating bacterium. Seems to me he's got a similar right to protect his intellectual property, which in my opinion, is none.
My surprise was borne of my experiences in Korean and in China over the past 13 years. I do think it's time I wake up to the fact that the Korean population is becoming more and more literate in English. My colleague (who heads our Korean branch office) is in his 40's and it's often difficult to communicate with him, let alone engage in word-play. His teen-aged kids, however, can chat happily for hours in English, jokes included.
For the record, I can read & write Hangul. I don't always understand what I'm reading, but because of the beauty and simplicity of Hangul, I can read it.
I had the same reaction. I think a lot of other folks won't catch it, though (like the ignoramous who 1st replied to your post).
I'm betting the CEO/President/founder/whatever is named Park. It's a little surprizing to me that a non-native english speaker would be savvy enough to think up that play on words/names.
So, are we supposed to pronounce it gamepark or gamepok (the Korean family name anglicized to Park is actually pronounced pok or pahk)?
Err, I read the article. I'm left a little confused. It seems they're talking less about non-uniform sampling than how to best recreate the function of an existing dataset.
For down-to-earth, "non-uniform" sampling methods, I've used variations of two in the past:
1) Relatively low sample rate until a known significant event in the sample stream is detected.
Then shift to the appropriate sample rate (or even variable, based upon a timing profile).
This has been especially useful when decoding (radio) "bursty" datastreams where there's long periods of "silence"
interrupted infrequently by data transmissions. No sense in wasting buffer space or overhead until you know you've got data to look at.
2) A variation I've used more often: operate at full (Nyquist or better) spec sample rate, storing to a fixed-size buffer whose contents are discarded/ignored if, upon inspection, it contains no interesting data. If data of interest shows up, say, halfway into the buffer, that becomes the beginning of the buffer (can you say "ring buffer"?) and you continue to sample til the buffer is used up. Best example of this would be the way a DSO (Digital Sampling O-scope) works.
I've implemented one system which used 4 different sample rates within a single data set. The data is a pressure waveform from a pump/valve dynamic response test. In it are regions of interest which range in sampling priority from to "nothing's happened" to "gotta know what happened up to 4.7 mS after this event".
That's non-uniform sampling in my book. No need to "reconstruct missing data".
As has been pointed out in other post(s), QNX has been around a long time. In fact, they first called it Qunix, but AT&T (Bell Labs) slapped'm down on that long ago.
I'm heard first-hand testimonials attesting to its bullet-proof operation which makes it a great choice for controlling machinery. You can also install, de-install just about any service/driver/app without needing to reboot.
Where I work, we make large, expensive automated testing equipment (lotsa horsepower, moving parts, other dangerous shit). We wanted to eval QNX about 3 years ago, but they told me they only provide free eval copies to their $100K plus customers. We make about 7 to 12 machines per year; they slammed the door in my face.
Now (and their previous free non-comm version) that the've got a pkg I can use to eval, it's too late. Even if we were still in a position to choose QNX, I doubt we'd easily forget our previous snubbing.
The RC5 client has optimized cruncher cores for several CPU's, using instructions that are either unique to that CPU, or that are uniquely faster than in other CPU's.
Very little production software is optimized for a specific instruction set these days (MMX and 3DNow! instructions used in gfx software being partial exceptions).
On the PowerPC point, the relatively lower clock rate is terribly misleading; most instructions execute in fewer clock cycles than in Pentium type computers.
I recall, in the olden days, counting clock cycles & such to optimize Z80 asm code for size or speed as necessary. Ahhh, the good ol' days (not)!
" (Note: The Sabre bomb-proof bunker isn't located in any of the cities listed. I know because I can look out the window and see the top of it. See this [sabre.com].)
"
My brother say's it's underground, at/near the end of a runway at TUL. He's an HP guy and has been there on American Airlines account(s). That ring true? And does Sabre use much HP gear?
Here's a US town being proactive for Broadband...
on
Broadband Obstacles
·
· Score: 2, Interesting
Checkout http://www.stillwater.org/extras/appa.htm
and http://stillwater.brightok.net
The former is a dated document, but the project is still ongoing. The fiber loops are complete and the city of Stillwater and it's tech partner, Chickasaw, are still rolling out fiber to each neighborhood, "one at a time." One of the sweeter things about it is that the Incumbent Local Exchange Carrier (SBC) is not a part of it and powerless to stop it. That's particularly interesting 'cause Stillwater's SBC office has such [phone] connectivity that Creative Lab's N. Am. Tech support is there.
DSL (and phone) service thru Chickasaw won't be in my 'hood for a while, but their wireless net is.
All this is really groovy to talk about, but the bottomline? The price is still too high.
Assuming this technology catches on (and really works), it would be a great boon to the consumer electronics industry.
At least initially (the first coupla decades?), however, it could cause real harm to the hobby and small run electronic apps. Imagine if all major electronic goods are "printed" at a fab, much like today's chip foundries. The economy of mass production should drive the prices of such goods way down, and that's good. Now imagine that you've got this cool idea for a little 'tronic widget that might be of interest only to yourself and a few geek fellows. You just want to make 25 or 50, even 100 of them.
Your first obstacle is that the components needed just to build your prototype now cost about 2 to 10 times more than they used to (because discrete transistors, LED's capacitors and such are no longer mass produced). If you do get you prototype built and debugged, how can you afford to build (and who can afford to buy) your small lot of 50? It might be kinda like me trying to get UMC in Taiwan to fab me an ASIC to implement my hobby color organ or some such; they'll do it, if I can pay the 10K or 20K setup and tooling fee's and am willing to pay 2500 USD per chip.
Maybe some day, the ubertechie might be able to buy their own PC printer, but I see the possibility that hobby electronics and small projects would suffer in the meantime.
Almost forgot, this technology would also bring "no user servicable parts inside" to a whole new level!
If small is your goal, My vote is a qualified "PC/104Plus" (http://www.pc104.org/). At 3.8 x 3.6 Inches per board, it makes the name "compactPCI" a lie.
PC/104Plus is a PCI compatible (electrically, not physically), pass-through bus, where you stack on your peripheral cards. Most of the popular chipset (CPU's, Ethernet, Graphics, etc) functions are available.
You can build a full-featured PC in a really small package (such as this one at VersaLogic, no association to me, at http://versalogic.com/Products/DS.asp?ProductID=11 4 ).
The downside, most PC/104Plus boards will cost you anywhere from 3 to 20 times the price of a "normal" board.
"That's why I own"... and Radio Shack Cat #: 44-233 Bulk Tape eraser.
I gaurantee it to fsck up your floppies, video tapes, & harddrives (I don't gaurantee that someone may not be able to recover your data later).
I mostly use it to degauss CRT's that've been subjected to a mag field strong enough that its internal degaussing coil can't clear up those funky color splotches.
The link, http://isbn.nu/0312862075 listed in the main post attributes a novella titled "Neuromancer" to Mr. Vinge. I think they must be confused. William Gibson's novel by that name has been compared to Vernor Vinge's (earlier) work, "True Names" but is in no way related to Mr. Vinge.
Unless there's some other connection of which I'm unaware...
I'm gonna check the "well-known numbers" RFC, but
I did a little scan of one of the infectoids:
Ports open at:
21
25 (open mail relay too!)
80
135
139
443
445
1025
1027
2057
2162
2174
2200
2210
2214
2219
2227
2228
2257
2282
I recogize some of those ports, but surely
windows doesn't need all those ports open?
Whew! That review was scathing! Any time someone
bashes that hard on a product/company I get
suspicious of their data.
But in this case, I believe it. I've been madly
in love with AMD's Athlon/Duron line since I first
tried it. I upgraded all my employers PC's to
Duron processors (the day-to-day performance diff
tween Duron and Athlon not even noticable).
The dnet client on an AMD Duron/Athlon will whip
any Pentium hands down, clock-for-clock.
I do believe Intel really shit their nest this
time.
Slashdot Mirror
Gee, I feel like I'm missing something here. I *always* use my palm (both of them, actually) when riding a bike.
When I was a teenager, I knew a guy who could do a _handstand_ on his handlebars, but a handspring?!? Sounds like a good way to wipe out, get hurt and trash your bike to boot!
.
I've kind of given up on educating sysadmins about their Open Relays. I never received a single response out of over 50 incidents where I tried to contact them.
In one case, 210.217.41.84 was an open relay running SCO with an old sendmail. I tried to contact them at various email addies and got no response. I even got a Korean associate to try to contact an IT type at the company that owned the server; blank stares.
I checked that IP address 6 months ago, sendmail said "out of memory." I checked it today, its gone now. All this was over a period of at least a year; I wonder who finally pulled the plug and why...
-
I don't think the Chinese gonvernment much cares about what's going out their firewalls. In fact, they might consider it humorous that you [westerner] are being hassled from Chinese netblocks.
What they probably don't realise is that those "broken" servers, with pirated, unpatched, out-of-date versions of MickeySoft OS's that have open mail relays also have known, exploitable holes. These pirated OS machines could be hijacked from inside or outside of China and put to some really interesting uses.
-
I'll double the reply to your post which said you obviously don't work at corporate AR.
A close family member of mine works at HP and for several years was assigned to "The Wal-Mart Account." He often commuted (600 miles or so) to "lovely Benton-Hell" to handle the technical fallout of management decisions and other major crap. When corp Wal-Mart says to corp HP, "jump," HP says "I'm already jumping, do you want higher, to the left, to the right?" WM *always* plays the *big guys* (IBM, HP, Sun, others) off each other, so you gotta stay on their good side.
I have to wonder if Wal-mart's cost-cutting (retail and behind-the-scenes-techo-retail) is a part of HP's early adoptance of Linux when they've clearly had a top leading eunichs product for oh so many years.
Circumstantial evidence I've seen, plus WM's "cost-cutting," means I'll believe it if someone tells me WM has HP installing Linux on Sun boxen, talking to IBM iron.
900 Lb gorilla is right.
Advertisers flirt with loss of Trademark protection.
Sorry, my WordStar fingers (using CTL key combos to cursor around) caused me to post before I got started.
The quote, "You don't pick up a facial tissue, you pick up a Kleenex." in the article caught my eye. People also often ask for a Band-Aid(tm) when they mean bandage. I don't know how hard Kimberly-Clark works to protect the Kleenex(tm) brand, but I'm sure they're quite close to losing it (not loosing it, damn it, loosing is not a word) all the time. There's a fine line between making your product name known in every household and making it a household word (or whatever the requirement is to lose your trademark protection).
I thought [Mike Fischer, vice president of entertainment marketing for Sega], a guy who oughta know better, was a bit irresponsible in making that statement.
"You don't pick up a facial tissue, you pick up a Kleenex."
You came very close to saying what I wanted to post.
I think the issue of patenting DNA is similar to that of software patents, to which I disagree. In fact, DNA *is* our software and the software of all life (as we know it). That little robosome machine that transcribes our software could be viewed as a sort of execution unit.
If someone writes a cool new app in C++, they might want to patent it (or some part of it). These days (at least in the US), they'd stand a good chance of being awarded a patent. Imagine a bio-engineer who "wrote" a cool new app in DNA, say, a cholesterol eating bacterium. Seems to me he's got a similar right to protect his intellectual property, which in my opinion, is none.
Sorry if I offended. Are you Korean?
My surprise was borne of my experiences in Korean and in China over the past 13 years. I do think it's time I wake up to the fact that the Korean population is becoming more and more literate in English. My colleague (who heads our Korean branch office) is in his 40's and it's often difficult to communicate with him, let alone engage in word-play. His teen-aged kids, however, can chat happily for hours in English, jokes included.
For the record, I can read & write Hangul. I don't always understand what I'm reading, but because of the beauty and simplicity of Hangul, I can read it.
I had the same reaction. I think a lot of other folks won't catch it, though (like the ignoramous who 1st replied to your post).
I'm betting the CEO/President/founder/whatever is named Park. It's a little surprizing to me that a non-native english speaker would be savvy enough to think up that play on words/names.
So, are we supposed to pronounce it gamepark or gamepok (the Korean family name anglicized to Park is actually pronounced pok or pahk)?
Err, I read the article. I'm left a little confused. It seems they're talking less about non-uniform sampling than how to best recreate the function of an existing dataset.
For down-to-earth, "non-uniform" sampling methods, I've used variations of two in the past:
1) Relatively low sample rate until a known significant event in the sample stream is detected.
Then shift to the appropriate sample rate (or even variable, based upon a timing profile).
This has been especially useful when decoding (radio) "bursty" datastreams where there's long periods of "silence"
interrupted infrequently by data transmissions. No sense in wasting buffer space or overhead until you know you've got data to look at.
2) A variation I've used more often: operate at full (Nyquist or better) spec sample rate, storing to a fixed-size buffer whose contents are discarded/ignored if, upon inspection, it contains no interesting data. If data of interest shows up, say, halfway into the buffer, that becomes the beginning of the buffer (can you say "ring buffer"?) and you continue to sample til the buffer is used up. Best example of this would be the way a DSO (Digital Sampling O-scope) works.
I've implemented one system which used 4 different sample rates within a single data set. The data is a pressure waveform from a pump/valve dynamic response test. In it are regions of interest which range in sampling priority from to "nothing's happened" to "gotta know what happened up to 4.7 mS after this event".
That's non-uniform sampling in my book. No need to "reconstruct missing data".
As has been pointed out in other post(s), QNX has been around a long time. In fact, they first called it Qunix, but AT&T (Bell Labs) slapped'm down on that long ago.
I'm heard first-hand testimonials attesting to its bullet-proof operation which makes it a great choice for controlling machinery. You can also install, de-install just about any service/driver/app without needing to reboot.
Where I work, we make large, expensive automated testing equipment (lotsa horsepower, moving parts, other dangerous shit). We wanted to eval QNX about 3 years ago, but they told me they only provide free eval copies to their $100K plus customers. We make about 7 to 12 machines per year; they slammed the door in my face.
Now (and their previous free non-comm version) that the've got a pkg I can use to eval, it's too late. Even if we were still in a position to choose QNX, I doubt we'd easily forget our previous snubbing.
The RC5 client has optimized cruncher cores for several CPU's, using instructions that are either unique to that CPU, or that are uniquely faster than in other CPU's.
Very little production software is optimized for a specific instruction set these days (MMX and 3DNow! instructions used in gfx software being partial exceptions).
On the PowerPC point, the relatively lower clock rate is terribly misleading; most instructions execute in fewer clock cycles than in Pentium type computers.
I recall, in the olden days, counting clock cycles & such to optimize Z80 asm code for size or speed as necessary. Ahhh, the good ol' days (not)!
" (Note: The Sabre bomb-proof bunker isn't located in any of the cities listed. I know because I can look out the window and see the top of it. See this [sabre.com].)
"
My brother say's it's underground, at/near the end of a runway at TUL. He's an HP guy and has been there on American Airlines account(s). That ring true? And does Sabre use much HP gear?
Checkout http://www.stillwater.org/extras/appa.htm
and http://stillwater.brightok.net
The former is a dated document, but the project is still ongoing. The fiber loops are complete and the city of Stillwater and it's tech partner, Chickasaw, are still rolling out fiber to each neighborhood, "one at a time." One of the sweeter things about it is that the Incumbent Local Exchange Carrier (SBC) is not a part of it and powerless to stop it. That's particularly interesting 'cause Stillwater's SBC office has such [phone] connectivity that Creative Lab's N. Am. Tech support is there.
DSL (and phone) service thru Chickasaw won't be in my 'hood for a while, but their wireless net is.
All this is really groovy to talk about, but the bottomline? The price is still too high.
Assuming this technology catches on (and really works), it would be a great boon to the consumer electronics industry.
At least initially (the first coupla decades?), however, it could cause real harm to the hobby and small run electronic apps. Imagine if all major electronic goods are "printed" at a fab, much like today's chip foundries. The economy of mass production should drive the prices of such goods way down, and that's good. Now imagine that you've got this cool idea for a little 'tronic widget that might be of interest only to yourself and a few geek fellows. You just want to make 25 or 50, even 100 of them.
Your first obstacle is that the components needed just to build your prototype now cost about 2 to 10 times more than they used to (because discrete transistors, LED's capacitors and such are no longer mass produced). If you do get you prototype built and debugged, how can you afford to build (and who can afford to buy) your small lot of 50? It might be kinda like me trying to get UMC in Taiwan to fab me an ASIC to implement my hobby color organ or some such; they'll do it, if I can pay the 10K or 20K setup and tooling fee's and am willing to pay 2500 USD per chip.
Maybe some day, the ubertechie might be able to buy their own PC printer, but I see the possibility that hobby electronics and small projects would suffer in the meantime.
Almost forgot, this technology would also bring "no user servicable parts inside" to a whole new level!
PC/104 and/or PC/104Plus
http://www.pc104.org
If small is your goal, My vote is a qualified "PC/104Plus" (http://www.pc104.org/). At 3.8 x 3.6 Inches per board, it makes the name "compactPCI" a lie.
1 4 ).
PC/104Plus is a PCI compatible (electrically, not physically), pass-through bus, where you stack on your peripheral cards. Most of the popular chipset (CPU's, Ethernet, Graphics, etc) functions are available.
You can build a full-featured PC in a really small package (such as this one at VersaLogic, no association to me, at http://versalogic.com/Products/DS.asp?ProductID=1
The downside, most PC/104Plus boards will cost you anywhere from 3 to 20 times the price of a "normal" board.
The acronym commonly called GPS is really GPSS or, correctly, GPS System [Global Positioning Satellite System].
"That's why I own"... and Radio Shack Cat #: 44-233 Bulk Tape eraser.
I gaurantee it to fsck up your floppies, video tapes, & harddrives (I don't gaurantee that someone may not be able to recover your data later).
I mostly use it to degauss CRT's that've been subjected to a mag field strong enough that its internal degaussing coil can't clear up those funky color splotches.
The link, http://isbn.nu/0312862075 listed in the main post attributes a novella titled "Neuromancer" to Mr. Vinge. I think they must be confused. William Gibson's novel by that name has been compared to Vernor Vinge's (earlier) work, "True Names" but is in no way related to Mr. Vinge.
Unless there's some other connection of which I'm unaware...
Several machines listed in logs as attacking
show up on eeye.com's CodeRedScanner as not
being vulnerable to the index.ida exploit.
Is it possible that the CRII worm seals the hole?
Or maybe those servers have been patched alread.
What IP address was attacking you?
rootcon.com resolves to 208.131.0.34 and is
running Apache.
I'm gonna check the "well-known numbers" RFC, but
I did a little scan of one of the infectoids:
Ports open at:
21
25 (open mail relay too!)
80
135
139
443
445
1025
1027
2057
2162
2174
2200
2210
2214
2219
2227
2228
2257
2282
I recogize some of those ports, but surely
windows doesn't need all those ports open?
Sorry, I was responding to the "Why the Pentium
4 sucks" msg, re: the url
http://www.emulators.com/pentium4.htm
Whew! That review was scathing! Any time someone
bashes that hard on a product/company I get
suspicious of their data.
But in this case, I believe it. I've been madly
in love with AMD's Athlon/Duron line since I first
tried it. I upgraded all my employers PC's to
Duron processors (the day-to-day performance diff
tween Duron and Athlon not even noticable).
The dnet client on an AMD Duron/Athlon will whip
any Pentium hands down, clock-for-clock.
I do believe Intel really shit their nest this
time.