It depends upon whether the signal in question is DSSS or FHSS.
Direct sequence spread spectrum involves multiplying the signal by a pseudorandom noise signal, and won't have the sync problems (though it will have other problems).
Frequency hopping Spread Spectrum involves changing frequencies really fast - so the time sync is important.
OK, so Gator is saying: Spyware is software installed without the users knowledge.
Our software is only installed with the user's knowledge.
Ergo, we are not spyware.
OK, so, define a new term: Tattleware - software that reports your actions to a third party
Then use that term to describe Gator.
Additionally, it would seem to me that one could argue in court You say your software is not spyware, since it is always installed with the user's permission. Fine - prove your assertion. We want a list of all computers with Gator on them, and proof that the users consented to the installation.
The TCXOs (temperature compensated crystal oscillators ("X" being the industry standard abbreviation for crystal - get over it)) used in moble equipment are usually rated about.5 ppm, so a 500 Hz shift isn't that much.
The more important aspect is the timing skew - GSM and CDMA require the mobile and the base station to have a VERY accurate idea of the time of flight delay between them, so as to keep the transmissions in their allocated time slots (IIRC GSM requires something like a 5 microsecond accuracy, but not being at work yet I can't get the specs right now.)
Moving that fast means the timing skew is going to shift significantly between bursts.
However, most high speed trains are moving to having a cell on the train itself, which then links to the landline system via a dedicated link from train to land.
One of the central tenents of all good engineering is "As complex as is needed, and NO MORE." Any more you add after that point is just "more to go wrong".
Eight wheels, eight suspension systems, eight control systems. True, perhaps the system is designed with failure in mind, but think about owning this as a vehicle for normal use - how often will you be taking in to be fixed, because one or more wheels have broken?
It's just like the fools who buy 4 wheel drive SUVs when what they need is a minivan - now they have what amounts to a whole extra powertrain to go wrong.
Now, if the intent was for this to be used in unusual circumstances (forestry work, extreme rough road work or the like) I could believe this was "as complex as needed but no more".
The "Three click rule" and the 7 +/- 2" rule are good rules for designing simple UIs (of which web pages can be considered a subset), but simple inspection can reveil the problem with this idea.
Suppose a UI were to scrupulously follow both rules. Then you would have a maximum number of choices of 9 ^ 3 = 729 choices. No more.
That may be great IF the number of choices you have is less than 729, and IF the choices can naturally be grouped in bunches of 9.
However, any complicated application may easily exceed this.
Moreover, people CAN deal with more than 7 choices, as long as the choices are somewhat related - Baskin Robins 31 flavors are all exactly that - "flavors". Imagine if a BR menu offered 31 choices of foods, drinks, plate colors, locations in the restaurant, server names, music, etc. ALL AT ONCE.
7 +/- 2 and 3 click are useful GUIDELINES. Just as saying "Using goto in C/C++ is generally a bad idea", or "pointing a loaded gun at any part of your body is a bad idea" are pretty good guidelines, there are times when you need violate them (e.g. error handling in the absence of exceptions, demonstrating a bullet-resistant vest, and designing a complicated piece of test equipment).
You should just use them AS GUIDELINES - "Hey, I really have a lot of items in this menu, perhaps I should take a break and see if I can come up with a different way to group them?"
On a radio tech forum I follow (here) they have a list of Mikey jokes - here's a sample:
Q. How does Michael Jackson pick his nose? A. From a catalogue.
Q: Have you heard about Michael Jackson's New Book? A: It's called "The In's and Out's of Child Rearing". Q: Why was Michael Jackson spotted at K-Mart? A: He heard boys' pants were half-off!
Q. What's the difference between Mr. Potato Head and Michael Jackson? A. Michael Jackson has had more noses.
Q. When is it bed time at Michael Jackson's house? A. When the big hand is on the little hand!
Q. What do Michael Jackson and a jockey have in common? A. They both ride three year olds.
The mindless anti-GM zelots who protest every GM creature without good reason really piss me off. No matter the organism, no matter the modification, they are against it - not because they have a real REASON or anything, but because "It's GM, it must be bad!"
I am not saying "all GM is good, let's go" - quite the contrary:
I am against BT-enhanced corn - not because it may kill butterflies, but because it guarantees that a perfectly good insecticide will be rendered USELESS in a few years. Rather than making a BT based spray, and using it ONLY in cases of severe infestation, and then making sure you KILL ALL THE BUGS, it will now be used everywhere, no matter what, but at a level that will allow the naturally resistant bugs to live, breed, and dominate!
I am against the various Monsanto "self-destructing" breeds of wheat, because that just is one more way in which farming is converted from a relatively self-sufficient and sustainable activity into an activity dependant upon the corporations to survive. Instead of farmers being able to get next year's seed from this year's harvest, they have to go back to Monsanto every year. Improve wheat all you want, guys, but the same arguments RMS makes about software apply to crops - cost to create high, cost to replicate LOW, so GPL them (or at least Aladdin license them!) In twenty years when Monsanto's patents on those strains expire, will you see Monsanto create a strain without the self-destruct gene, or will Monsanto just let the parent stocks die?
Worry about introducing a new lifeform into a biosphere forigen to it - but the cain toad has done more damage than this fish ever will, and the cain toad wasn't gene-tweaked! Being gene-tweaked is not sufficent nor necessary cause to be damaging to an ecosystem!
The mindless anti-GM zelots can prevent things that really help - I would love to see a GM crop that fixed nitrogen like a legume, yielded lots of bio-desiel and plastic precursors, and could be grown year after year in brackish soil, concentrating the salt in the stalks - imagine the boost to the environment and the boost to the third world farmer! But you can bet that, even if an RMS-inspired botanist created such a crop and released it free of charge (think George Washington Carver), the mindless anti-GMers would prevent it from seeing the light of day!
In short, BE worried about things, but have a clearly reasoned, well thought through idea of WHY you are worried - not just because the thing has "scary" words in it like "genetically modified", "nuclear", or "diesel"!
Bruce commented in the linked article that he did not understand why there is the problem with dependancies in RPM when APT seems to have fixed this.
Repeat after me - It's not RPM's fault, it is the fault of the packagers!
The problem with ANY packaging system that allows for dependancies is the JACKASS PACKAGE CREATOR who defines his dependancies as libfoo.1.2.3.so.pl1.thursday.3oclock.mine.mine as opposed to libfoo.1.so
Debian fix this by being very controlled in what they let in - overly anal-retentive packages get bounced, and packages that depend upon packages not in the standard must clearly state WHAT they need and where to get it.
Redhat itself does a fair job of this, but with everybody and their dog creating RPMS higglty-pigglty, the third party RPMs suck.
Once again, repeat after me, class: No packaging system will solve the problem of incompetent people creating packages
IF the Linux standard base were to be adopted,
IF commonly distributed packages were either
Linked against only packages in the approved LSB,or
Provide along with the packages they needed
IF packages didn't needlessly replace things that didn't pertain to them (e.g. Nautilus wanting to replace my copy of Mozilla with its own),
THEN we could do away with the "dependancy hell" problem, no matter WHAT the underlying packaging system was
Also - with regards to the MAN vs. INFO issue - what about having a simple web server, bound to 127.0.0.1:<mumble> that would respond to INFO and MAN requests, returning HTML. Then a user could use ANYTHING from Lynx to Mozilla to view help, and you could get all the benefits of cross-linking. Ideally, this server could run from inetd, stand-alone, or as a CGI under Apache.
What I am trying to do is to get the author of Downloader for X to include BT protocol in his program. As his program already supports HTTP, FTP (with file splitting for multithreaded downloads), scheduling, inter-process communication (X drag-n-drop) and has a decent GUI, if he would add BT it would be just about perfect.
Go to his site, take a look, and perhaps drop him a line on his forums asking for BT - maybe if he gets/.ed he may just do it.
I had submitted a story some time ago (mirrored in my journal) about a rather strange stand-up game in Japan called "BOONG-GA BOONG-GA". The idea is to "spank" an animated character on-screen. This "spanking" is done with controller shaped like a hand with the index finger outstreched.
Election commissioners are usually appointed, not elected (funny when you think about it)....
Not funny at all - necessary. If an election commissioner were to be an elected official, then he could alter the election rules to favor himself upon re-election, and thus be in a conflict of interest.
Well, I was trying to head off the semi-trolls ("Uh-hu-hu U DON'T NO HOW 2 SPELL CRYSTAL!")
And in Japan, nothing is far from anything, especially cell towers - they have those things packed tighter than sardines.
But since the phones are talking to a cell on the train, it matters not.
It depends upon whether the signal in question is DSSS or FHSS.
Direct sequence spread spectrum involves multiplying the signal by a pseudorandom noise signal, and won't have the sync problems (though it will have other problems).
Frequency hopping Spread Spectrum involves changing frequencies really fast - so the time sync is important.
But it is GATOR that is suing, claiming not to be spyware - thus they ARE the plaintif (civil case, not criminal, therefor plaintif, not prosecution).
OK, so Gator is saying:
Spyware is software installed without the users knowledge.
Our software is only installed with the user's knowledge.
Ergo, we are not spyware.
OK, so, define a new term:
Tattleware - software that reports your actions to a third party
Then use that term to describe Gator.
Additionally, it would seem to me that one could argue in court
You say your software is not spyware, since it is always installed with the user's permission. Fine - prove your assertion. We want a list of all computers with Gator on them, and proof that the users consented to the installation.
500 Hz at 900 MHz is less than 1 ppm.
.5 ppm, so a 500 Hz shift isn't that much.
The TCXOs (temperature compensated crystal oscillators ("X" being the industry standard abbreviation for crystal - get over it)) used in moble equipment are usually rated about
The more important aspect is the timing skew - GSM and CDMA require the mobile and the base station to have a VERY accurate idea of the time of flight delay between them, so as to keep the transmissions in their allocated time slots (IIRC GSM requires something like a 5 microsecond accuracy, but not being at work yet I can't get the specs right now.)
Moving that fast means the timing skew is going to shift significantly between bursts.
However, most high speed trains are moving to having a cell on the train itself, which then links to the landline system via a dedicated link from train to land.
The same way your local radio stations manage to have a "90% music hour" with 20 minutes of commercials - by overlapping the talks.
One of the central tenents of all good engineering is "As complex as is needed, and NO MORE." Any more you add after that point is just "more to go wrong".
Eight wheels, eight suspension systems, eight control systems. True, perhaps the system is designed with failure in mind, but think about owning this as a vehicle for normal use - how often will you be taking in to be fixed, because one or more wheels have broken?
It's just like the fools who buy 4 wheel drive SUVs when what they need is a minivan - now they have what amounts to a whole extra powertrain to go wrong.
Now, if the intent was for this to be used in unusual circumstances (forestry work, extreme rough road work or the like) I could believe this was "as complex as needed but no more".
I am SO depressed that all the /. crowd can come up with are lame Contact references.
YOU CALL YOURSELVES GEEKS! DROP AND GIVE ME TWENTY!
Vega, as ALL REAL GEEKS know, was the home of Mother Thing of Robert Heinlein's "Have Space Suit, Will Travel".
And if they are watching Earth circa 1978, we'd better be damn thankful they don't rotate us 90 degrees just on general principles!
I can think of a couple of ways to recreate that functionality under Linux:
/home on a tmpfs volume.
lndir + tmpfs - Mount system read only, on login lndir the user's home dir to a directory in tmpfs.
LVM snapshots - create an LVM snapshot of the system in a running state - restore on restart.
Mount system RO, mount
Yeah, they tried this crap on us - "If you use Linux, you will have to GPL all your code!"
We shot them down in flames on that one.
There can be only one!
Now, hold still while I chop off your head....
Although the aftermath of a battle sure is hard on the servers, what with all the lightning and such....
The "Three click rule" and the 7 +/- 2" rule are good rules for designing simple UIs (of which web pages can be considered a subset), but simple inspection can reveil the problem with this idea.
Suppose a UI were to scrupulously follow both rules. Then you would have a maximum number of choices of 9 ^ 3 = 729 choices. No more.
That may be great IF the number of choices you have is less than 729, and IF the choices can naturally be grouped in bunches of 9.
However, any complicated application may easily exceed this.
Moreover, people CAN deal with more than 7 choices, as long as the choices are somewhat related - Baskin Robins 31 flavors are all exactly that - "flavors". Imagine if a BR menu offered 31 choices of foods, drinks, plate colors, locations in the restaurant, server names, music, etc. ALL AT ONCE.
7 +/- 2 and 3 click are useful GUIDELINES. Just as saying "Using goto in C/C++ is generally a bad idea", or "pointing a loaded gun at any part of your body is a bad idea" are pretty good guidelines, there are times when you need violate them (e.g. error handling in the absence of exceptions, demonstrating a bullet-resistant vest, and designing a complicated piece of test equipment).
You should just use them AS GUIDELINES - "Hey, I really have a lot of items in this menu, perhaps I should take a break and see if I can come up with a different way to group them?"
Be careful about judging location from accent - you may be fooled.
True story: my employer was in talks with a major player in the industry for a contract (sorry, can't get much more precise than that).
So, we sent the lead engineer we would put on the project, and the head of the division that was in.
After a few hours of negotiations, the reps from the company we were talking to said, "Excuse us, but we thought your company was located in America?"
The engineer (A French citizen) and the exec (a British citizen) looked at each other, and said "Yes, we are located in Wichita, KS. Why do you ask?"
On a radio tech forum I follow (here) they have a list of Mikey jokes - here's a sample:
Q. How does Michael Jackson pick his nose?
A. From a catalogue.
Q: Have you heard about Michael Jackson's New Book?
A: It's called "The In's and Out's of Child Rearing".
Q: Why was Michael Jackson spotted at K-Mart?
A: He heard boys' pants were half-off!
Q. What's the difference between Mr. Potato Head and Michael Jackson?
A. Michael Jackson has had more noses.
Q. When is it bed time at Michael Jackson's house?
A. When the big hand is on the little hand!
Q. What do Michael Jackson and a jockey have in common?
A. They both ride three year olds.
I am not saying "all GM is good, let's go" - quite the contrary:
The mindless anti-GM zelots can prevent things that really help - I would love to see a GM crop that fixed nitrogen like a legume, yielded lots of bio-desiel and plastic precursors, and could be grown year after year in brackish soil, concentrating the salt in the stalks - imagine the boost to the environment and the boost to the third world farmer! But you can bet that, even if an RMS-inspired botanist created such a crop and released it free of charge (think George Washington Carver), the mindless anti-GMers would prevent it from seeing the light of day!
In short, BE worried about things, but have a clearly reasoned, well thought through idea of WHY you are worried - not just because the thing has "scary" words in it like "genetically modified", "nuclear", or "diesel"!
What - you think I want to look at a spammer all the time? Sheesh!
I want him in his hole, where he belongs!
I want a spammer for Christmas.
And some soundproofing foam.
I already have the duct tape, the soldering iron, the needlenose pliers, and the wallplug ending in bare wires.
" You, sir, are a BOFH. Bravo!"
No, a TRUE BOFH would not use a wimpy 9 volt battery on shrimp.
A true BOFH would use a 9000 volt cattleprod on the luser interrupting him.
How are you detecting a drop? Are you just keeping both interfaces up and letting the metric decide, or do you have a script that detects the fault?
Repeat after me - It's not RPM's fault, it is the fault of the packagers!
The problem with ANY packaging system that allows for dependancies is the JACKASS PACKAGE CREATOR who defines his dependancies as
libfoo.1.2.3.so.pl1.thursday.3oclock.mine.mine
as opposed to
libfoo.1.so
Debian fix this by being very controlled in what they let in - overly anal-retentive packages get bounced, and packages that depend upon packages not in the standard must clearly state WHAT they need and where to get it.
Redhat itself does a fair job of this, but with everybody and their dog creating RPMS higglty-pigglty, the third party RPMs suck.
Once again, repeat after me, class:
No packaging system will solve the problem of incompetent people creating packages
Also - with regards to the MAN vs. INFO issue - what about having a simple web server, bound to 127.0.0.1:<mumble> that would respond to INFO and MAN requests, returning HTML. Then a user could use ANYTHING from Lynx to Mozilla to view help, and you could get all the benefits of cross-linking. Ideally, this server could run from inetd, stand-alone, or as a CGI under Apache.
What I am trying to do is to get the author of Downloader for X to include BT protocol in his program. As his program already supports HTTP, FTP (with file splitting for multithreaded downloads), scheduling, inter-process communication (X drag-n-drop) and has a decent GUI, if he would add BT it would be just about perfect.
/.ed he may just do it.
Go to his site, take a look, and perhaps drop him a line on his forums asking for BT - maybe if he gets
I had submitted a story some time ago (mirrored in my journal) about a rather strange stand-up game in Japan called "BOONG-GA BOONG-GA". The idea is to "spank" an animated character on-screen. This "spanking" is done with controller shaped like a hand with the index finger outstreched.
Not funny at all - necessary. If an election commissioner were to be an elected official, then he could alter the election rules to favor himself upon re-election, and thus be in a conflict of interest.
OK, I have to ask - why not just use a PC?
In other words, what will a more traditional single-board computer (SBC) get you over a PC for your development needs?
Are you wanting something with on board Flash to boot from, are you wanting something that isn't an x86, are you wanting a particular form factor?
In short, what are the needs that preclude using a PC.
"Can you hear me now?"
"Hello? I said, CAN YOU HEAR ME NOW?"
"Hello?"