A forth interpreter + compiler + editor can live in less than 20k of memory. This provides a lot of extra flexibility in how one does things like: handle BIOS extentions (juust write em in forth and they can run on any CPU architecture - not just x86), write service tools, disk drive low-level formatters,....
Forth is more compact than machine code. This might not seem relevant in multi-megabyte machines, but it does make it a lot easy/cheaper to add small code snippets to hardware devices.
You can modify the behaviour of the Forth compiler itself, on the fly. This makes it very handy for expressing various structures that are often difficult to express in "static" languages like C and asm.
It gives older programmers an excuse to break out the flared pants again!
Lighting up optical elements is the simple bit. Deciding what to show is the hard bit. To do this you need to know where the observer is standing and what they are looking at. If you have two or more observers then how do you decide which image to show?
Of course I have not read the patent, but the basic idea behind RFID (thump a circuit with RF and it tells you who it is) has been around way longer than that.
The basic idea is used by IFF (indentify friend or foe) and trnasponder devices used in aircraft. Whack 'em with radar and they output a pattern of radar signals that can be detected and decoded (or even viewed as extra returns on the radr screen)
It is close to impossible to write embedded code without using at least some assembler. All C programs - err well just about -- need some assembly code to set things up before calling main().
Another reason why assmbler will always be useful is that many CPUs have valuable instructions that have no C equivalent. Therefore to use these instructions one typically needs to write assembler.
Frequently it is simpler to write a function that accesses hardware directly in assembler than it is to write it in C. With C the compiler will often jerk you around and optimise away specific behaviour that you want.
If they fiddle with the base protocols then there will be interoperability issues. They will want to maintain compatability with previous BT devices. This might mean that when you have a piconet with an old BT device, some of the high speed features might be unsupported (analagous to how my USB2 hub downgrades everything to USB1.1 as soon as you plug in a USB1.1 device).
Of course they're still using 2.4 GHz which is the wild west of the radio spectrum. Still going to have issues headbutting Wifi etc.
And no, of course I didn't RTFM, that would be cheating.
If Linux is central to their strategy then there is no need for x86-ness. x86-ness just makes the CPU design that much harder and commoditises their product. The x86 CPU market is very difficult to compete with unless you have something special to offer. You need fresh high speed/high margin devices every few months to keep alive. It would seem these folks are dumpster diving and trying to produce a lower price CPU to fill the bottom end. A difficult thing to do when old Pentiums etc could be pressed into service instead.
A collegue of mine has VPN over DSL to a corporate network. They do all their phones via VOIP. If you send him a ~1MB email while he's on the phone, the call goes down the toilet. Not exactly a "new millenium experience".
If I'm ignorant, I can do something with bad sonsequences, but it needs intent to be evil. Some one could also exploit my ignorance, but the ignorance does not cause the evil per se.
Similarly, evilness might continue because people are lazy (ie. I prefer to bitch about things on/. or say nothing - rather than do something useful to fight the evil). Again, evil people might exploit my laziness.
In none of these cases does laziness or ignorance cause the evil. Or, as an analogy, people who have kids don't cause kiddy-rape by having kids (though they provide the opportunity for it to happen).
How can ignorance be a cause of evil? Evil requires a motive.
Going the other way...
on
60GB iPod Coming?
·
· Score: 2, Interesting
There are two ways of exploiting Moore's Law.: iether getting more Mbytes or MHz in the same package or achieiving previous size/speed with smaller parts or less power at a lower price. I doubt I'll ever buy an ipod, but I expect most people would rather have a smaller/cheaper/longer playing ipod than one that can store 60GB.
Hoping that Microsoft is just using this defensively is pure crap. There is plenty of prior art out there so MS would have no problems in defeating any attacks on the "double click" front.
IMHO this is just another weapon for their FUD machine. With a patent like this they can yell that they have the power to shut down competitors. It makes people considering moving to, say, Novell desktops think twice.
The problem with patents is that they stand until challenged and are overturned by the USPTO. Until they are overturned they are legally binding - no matter how stupid they are.
Give them a break.... Linux was written by only one person (and a few helpers) in one year (ie one year per programmer). Microsoft has a few hundred programmers so to be fair you should also give them the same time (ie. one year per programmer) -- a few hundred years -- before you start bad mouthing them.
If you think Microsoft is doing this to "be nice" you're being taken for a sucker.
I expect this is coming from two quarters:
1.Finally listening to customers. Customers don't want to keep on hearing "Remember that shit you bought from us last year? Well is is crap and broken and you have to buy the new one! hahahahhahahha". Eventually their assholes start to get sore and they go somewhere else. Now that Novell is back in town with a cool offering, people will be thinking of a switch...
In many countries and I expect US states you have to provide minimum support periods of some years.
Geode's claim to fame is its x86-ness. It can leverege x86 software. With good software now available for ARM and MIPS, the motivation to use x86 for embedded goes away.
Many/most of the power saving tweaks involve taking away stuff that has been used to make Pentiums fast. This means that a Geode is unlikely to perform as fast as a similarly clocked Pentium.
Now this is a whole new architecture, but on the previous (current) generation of Geodes, I found that a 200MHz ARM was faster than a 300MHz Geode. The ARM also only used about 10-15% the power that the Geode was using and was also smaller, lighter etc... Geodes were quite popular in embedded systems but this has largely stagnated. I really struggle to see how Geode will make a comeback.
Slashdot Mirror
Generate the next wave of dotbomb style IPOs.
A forth interpreter + compiler + editor can live in less than 20k of memory. This provides a lot of extra flexibility in how one does things like: handle BIOS extentions (juust write em in forth and they can run on any CPU architecture - not just x86), write service tools, disk drive low-level formatters,....
Forth is more compact than machine code. This might not seem relevant in multi-megabyte machines, but it does make it a lot easy/cheaper to add small code snippets to hardware devices.
You can modify the behaviour of the Forth compiler itself, on the fly. This makes it very handy for expressing various structures that are often difficult to express in "static" languages like C and asm.
It gives older programmers an excuse to break out the flared pants again!
When compared to a Model-T even a Skoda looks pretty good.
When compared to being hit on the head, sitting on a rock feels pretty good.
IE is hardly a benchmark to compare anything to. How does FireFox compare with Opera?
Eye candy and a whopping 3% improvement make it onto the big feature list.
Good reason to turn off your phone on the airplane :-).
Lighting up optical elements is the simple bit. Deciding what to show is the hard bit. To do this you need to know where the observer is standing and what they are looking at. If you have two or more observers then how do you decide which image to show?
The basic idea is used by IFF (indentify friend or foe) and trnasponder devices used in aircraft. Whack 'em with radar and they output a pattern of radar signals that can be detected and decoded (or even viewed as extra returns on the radr screen)
Another reason why assmbler will always be useful is that many CPUs have valuable instructions that have no C equivalent. Therefore to use these instructions one typically needs to write assembler.
Frequently it is simpler to write a function that accesses hardware directly in assembler than it is to write it in C. With C the compiler will often jerk you around and optimise away specific behaviour that you want.
Of course they're still using 2.4 GHz which is the wild west of the radio spectrum. Still going to have issues headbutting Wifi etc.
And no, of course I didn't RTFM, that would be cheating.
If Linux is central to their strategy then there is no need for x86-ness. x86-ness just makes the CPU design that much harder and commoditises their product. The x86 CPU market is very difficult to compete with unless you have something special to offer. You need fresh high speed/high margin devices every few months to keep alive. It would seem these folks are dumpster diving and trying to produce a lower price CPU to fill the bottom end. A difficult thing to do when old Pentiums etc could be pressed into service instead.
A collegue of mine has VPN over DSL to a corporate network. They do all their phones via VOIP. If you send him a ~1MB email while he's on the phone, the call goes down the toilet. Not exactly a "new millenium experience".
Geek hint: Do this in that mythical place called "outide", unless you have a very understanding landlord/mother.
What part of this excites geeks? Physical exertion - no. Getting wet - no. French babe - not on this planet anyway.
If I'm ignorant, I can do something with bad sonsequences, but it needs intent to be evil. Some one could also exploit my ignorance, but the ignorance does not cause the evil per se.
Similarly, evilness might continue because people are lazy (ie. I prefer to bitch about things on /. or say nothing - rather than do something useful to fight the evil). Again, evil people might exploit my laziness.
In none of these cases does laziness or ignorance cause the evil. Or, as an analogy, people who have kids don't cause kiddy-rape by having kids (though they provide the opportunity for it to happen).
How can ignorance be a cause of evil? Evil requires a motive.
There are two ways of exploiting Moore's Law.: iether getting more Mbytes or MHz in the same package or achieiving previous size/speed with smaller parts or less power at a lower price. I doubt I'll ever buy an ipod, but I expect most people would rather have a smaller/cheaper/longer playing ipod than one that can store 60GB.
Well there ya go then.... They didn't even invenmt the most important CTL-ALT_DEL either.
I've seen a patent for a piece of tape stuck to a pole, so nothing suprises me any more wrt the USPTO.
IMHO this is just another weapon for their FUD machine. With a patent like this they can yell that they have the power to shut down competitors. It makes people considering moving to, say, Novell desktops think twice.
The problem with patents is that they stand until challenged and are overturned by the USPTO. Until they are overturned they are legally binding - no matter how stupid they are.
Give them a break.... Linux was written by only one person (and a few helpers) in one year (ie one year per programmer). Microsoft has a few hundred programmers so to be fair you should also give them the same time (ie. one year per programmer) -- a few hundred years -- before you start bad mouthing them.
I expect this is coming from two quarters:
1.Finally listening to customers. Customers don't want to keep on hearing "Remember that shit you bought from us last year? Well is is crap and broken and you have to buy the new one! hahahahhahahha". Eventually their assholes start to get sore and they go somewhere else. Now that Novell is back in town with a cool offering, people will be thinking of a switch...
In many countries and I expect US states you have to provide minimum support periods of some years.
As our fellow Geek ESR proves, you don't have to be a muggger's christmas if you're a geek.
why simple application software needs 2G of RAM and multi-GHz CPUs to get the responsiveness I got on a 100MHz 486 with Win3.11.
Many/most of the power saving tweaks involve taking away stuff that has been used to make Pentiums fast. This means that a Geode is unlikely to perform as fast as a similarly clocked Pentium.
Now this is a whole new architecture, but on the previous (current) generation of Geodes, I found that a 200MHz ARM was faster than a 300MHz Geode. The ARM also only used about 10-15% the power that the Geode was using and was also smaller, lighter etc... Geodes were quite popular in embedded systems but this has largely stagnated. I really struggle to see how Geode will make a comeback.