To have a such a thing would normally go hand in hand with the industry having a professional body that both unifies the professionals within that industry and controls access to the profession through some meritocratic discrimination.
Industry bodies in software and electronics seem to have avoided this so far. The most influential body in the industry is probably the IEEE. Thier influence has been far more prevalent in the techincal standards arena rather than the standardization of the skills of its own members.
It is probably worth exploring the model in the medical, accounting and legal industries. All three are considered elitist, insular and self serving.
To have the benefits of being in such a club you would also need to conceed the limitations of freedom that go along with it. You would not be able to just sit, design, code and sell work. You would first need to be tested, certified and admitted to the club and be subject to review or exclusion from the club based on your behaviour.
Open or free software may get labelled as 'unsafe' since it has contributions from 'uncertified' coders.
Thus I suspect the ramifications of professionalisation are a little to much to swallow for many used to the relative freedoms of the hi tech industry.
802.11b specifies a transport that goes as fast as 11mbps before DCF overheads are accounted for. If someone has a device that goes faster than this then it is not 802.11b. It is something else.
If it can still interwork with 802.11b devices then it will be doing so at a maximum of 11mbps.
If you want something faster, then why not use something that will not be obsoleted by a standardized solution such as 802.11a or the still-in-progress 802.11g?
270,000 licenses is too many to supply the needs of state workers, but may be very sesnible if it is to supply needs of online government services delivered over the internet to Californian residents.
If they cleaned up their processes and made most interaction with state government available to people over the internet, they may save bucketloads of time and money for everyone involved.
Of course this assumes that the government is competent to do so...
> (Ever tried to fix surface mount components with a soldering iron at your kitchen table? Don't!!)
I've always found that working with SMT is easier than through hole. You have gravity on your side. It will hold the component on the pad while you tack it in place.
Just use a decent soldering iron that has a small enough tip and don't make the mistake of using too small a tip. A too small tip doesn't hold enough heat to flow the solder onto larger SMT pins.
Also make good use of brush on flux and desolder braid. They are your friends when reworking SMT boards.
When laying out your own PCB, SMT components let you get away with drilling far fewer holes and zero ohm resistors let you 'jump' over tracks without using vias.
When it comes to probing, all your signals are generally available on one side. Most SMT parts (except BGA and LCC styles) don't shroud their leads like stand-up electrolytics and transistors do.
One of the primary barriers to messing with this sort of stuff in America is the crappiness of component supply for the hobbyist. I have yet to see anything that comes close to the likes of Radio Spares or Farnell in the UK.
As it happens, I am also dropping the cable TV due to a steadily deteriorating service and I have a satellite dish on order.
They have just announced price rises for the cable TV and they have been adding more and more channels, for which they demand more and more cash. Each new channel seems to slow down the already slow and crappy browser further and requires me to trawl through more and more unavailable channels to get to the few I can actually access.
In contrast, the satelite people have a nicer, faster, slicker browser, can sell me a combined Tivo/receiver box and for less money offer me more channels.
Maybe Comcast can stop the rot, but I'm not waiting to find out.
Myself and 100% of my friends who I know have cable with AT&T are in the process of switching to DSL and dropping AT&T due to the steady degradation of service, culminating in the @Home debacle.
Do Comcast know that AT&T may be about to lose a huge proportion of their customers?
I am an AT&T cable modem customer and have had the bad fortune of being transferred from @Home to AT&T's own ISP.
They suck massively.
1) Their support are not answering the phone
2) My IP address that has been static since I signed up over a year ago has suddenly changed and it appears that static addressing in any form has gone up in smoke. This screws anyone relying on a static IP.
3) They have been playing fast and loose with the service agreement (that I signed), but instituting an AUP (that I didn't sign)that directly contradicts it.
4) Their DNS service has been very erratic
This is not the sort of crap I want to be paying for and I am actively shopping for a replacement (I.E. DSL). I expect any other user who wants to do more than play with the latest microsoft browser will be doing the same and dumping AT&T as soon as possible.
This highlights an interesting split in the direction of modern operating systems research.
In the one corner we have MIT with Cesium that implements files and everything else in the form of an object oriented database.
In the other corner we have Bell-Labs/Lucent Plan 9 that implements databases and everything else in the form of files.
Plan 9 makes programming things easier since the file api lets you do most things. I wonder how much MIT's system makes things easier to do? I always found databases a pig to interface to (except in plan9).
Machine Beauty is a book by David Gelernter that gives an excellent exposition of why beauty in technology is a good thing. I recommend it to anyone who wants a deeper understanding of the issues.
I have in my possession an original copy of the wonderful tome "Apple graphics and arcade game design" by Jeffrey Stanton, circa 1982. Apple as in Apple ][.
This covers such advanced topics as:
Programming Apple Hi-Res graphics from BASIC and machine language.
Raster Graphics and bit mapping
Moving things around the screen
Simulating gravity.
Doing cunning things with 6502 assembler to make things run faster.
As a pre-teen getting to grips with the innards of Apples, this one proved a lot easier to read than anything by Rodney Zaks.
Perhaps if everyone was required to write their first game in 6502 assembler on an Apple ][, they might acquire a better outlook on life. Or maybe just a headache.
Intentional radiators on planes (carried by passengers at least) are not presently permitted and so the whole issue is moot until there is a change in these regulations.
The reason that the Bluetooth and 802.11 crowd keeps working on this is that there is some hope for movement by the FAA/CAA/Other regulatory bodies. The arguments are pretty simple..
1) 2.4GHz is already polluted by wideband radiation in planes, since they use microwave ovens to heat the food. So there can be some level of confidence that you 802.11 tranciever isn't going to bring the plane down.
2) Electrical interference is a function of both the strength of the interferer and the succeptability of the interferree. Plane equipment is supposed to be built to stringent succeptability requirements. When someone says you phone/pda/toothbrush will interfere with the plane, ask why the equiment on the plane is operating outside the succeptability requirements mandated for planes. The responsibility to make interference not be a problem has be foisted on the passengers. This is a bad thing. Passengers forget to turn phones off all the time. It should be the job of the plane manufacturers to make planes safe in the presence of passengers.
3) There is lots of lobbying going on.
However being charged money to get on the net is not really a necessity is it? One person sets his laptop up as a gateway/NAT router and everyone else sets up as an ad-hoc 802.11 network. That would save people a bit of cash.
If all you wanted to do was play quake with your peers, you wouldn't need net access at all. 802.11 can work peer to peer. You don't need an AP just to communicate between a group of machines.
So lobby lobby your MP/Senator/FAA rep/garage mechanic/EU minister to change the rules. It will improve the quality of your traveling life.
I thought the worst tech job out there was a well paid position as a microelectronics designer in Oregon. It must be since I've been unable to find anyone willing or able to take up the job for bloody ages.
I have gone in the opposite direction, from the UK to the US. However, my understanding is that generally you will experience less (but not a whole lot less) bureaucracy moving to the UK (and most other EU countries) that an EU citizen experiences moving to the US.
The ground rules are pretty much the same. If you are educated, have no criminal record and can find a company to sponsor you, then you will be able to get work authorization.
A thing to be aware of are that you will be going against the tide for a reason. EU techy citizens move to the US because of higher pay, lower taxes and a chance to work within a richer group of tech companies. You can expect lower pay, higher taxes and a more limited range of employers over there.
Nevertheless, I think its is pretty much a requirement of having an interesting life to live in more than one country. So go and do it.
This encapsulates two common misconceptions about Bluetooth.
1) Bluetooth is not a LAN.
It is fow wire replacement. The primary profiles (headset, synchronization, serial device) support that assertion. An investigation of the piconet/scatternet architecture reveals just how inefficient Bluetooth would be at trying to route data through a network. In fact the spec reads more like the plan was to obstruct network functionality for the benefit of wire replacement functionality. Point to point punting of Audio and packet data fits easily and efficiently into Bluetooth. Routing does not.
2) You can't just stick a fat PA on the front and operate at 100s of meters.
Part of the purpose is to allow many devices to interoperate. That is why the protocols are defined so much further up the stach than say 802.11. As a part of that, the 768kbps shared within a piconet may get pretty slow shared amongst a piconet. It would get stupidly slow shared between all the devices you could fit in 100s of meters (E.G.at a trade show). In addition, the high power devices sharing space with low power devices would lead to serious link budget imbalance in the RF. Rather like a pair of quiet guys trying to have a conversation at the front row of a moterhead concert.
As far as the protocols go, they are not too open (the latest specs are available to the SIG signees , but don't worry, they're a boring read). However the IEEE are rewriting them as 802.15 and you'll find them to be better written.
My advice, if you want to do Bluetooth protocol or application development is to concentrate on the upper layers (Above HCI), since the lower layers will get commoditized and componentized in due course and there is hardware involved, it won't be free anyway.
Re:So what does linux on a microcontroller offer?
on
Microcontroller Linux
·
· Score: 2
It looks to me like the situations you describe are more a case of the hardware moving to meet the OS, rather than the other way round. This is certainly not a bad thing for the right product. The Tivo and the the in card MP3 players are two prime examples. They both benefit greatly from using the Linux codebase.
What I'm questioning is the use of linux in things that are an order of magnitude smaller. Like cell phones, handheld PDAs and personal location systems. The interfaces are generally dictated by the microcontroller and/or whatever ASIC might be in the system. Hard disks are out of the question.
Running traditional linux apps seems to be an unlikely goal - I'm not paying to run a webserver from my cell phone and X applications are not ideal for a 128x72 LCD screen.
Quite possibly, the need for a real filesystem might be a good dividing line. Where you need such a thing, linux gives you lots of goodies for providing and using the filesystem. In a system which has no such need, linux looks too much like an overwight dinosaur.
So what does linux on a microcontroller offer?
on
Microcontroller Linux
·
· Score: 4
I have a rather direct interest in microcontroller based operating systems (I design microcontroller based products) so Linux on a micro might seem to be something I should be interested in. Yet I can't really indentify why I might want to use it.
The things I look for in a microcontroller OS are stuff like..
A small footprint
Lightweight realtime multitasking (if multitasking is required)
Rapid boot (millseconds)
Persistent state support
Will fit in a single flash chip with space for the application
Hosts whatever protocol stack is going in the product (stuff like GSM, IS95, TCP/IP, IS707 seem to be the usual suspects).
Most importantly.. A really really good remote debugger, cross target development environment and totally uncrashable kernel
So could someone clue me in as to why I should use something like this rather than say Nucleus?
Slashdot Mirror
DECT is not a user of the 2.4Ghz Band. It uses 1.9Ghz.
To have a such a thing would normally go hand in hand with the industry having a professional body that both unifies the professionals within that industry and controls access to the profession through some meritocratic discrimination.
Industry bodies in software and electronics seem to have avoided this so far. The most influential body in the industry is probably the IEEE. Thier influence has been far more prevalent in the techincal standards arena rather than the standardization of the skills of its own members.
It is probably worth exploring the model in the medical, accounting and legal industries. All three are considered elitist, insular and self serving.
To have the benefits of being in such a club you would also need to conceed the limitations of freedom that go along with it. You would not be able to just sit, design, code and sell work. You would first need to be tested, certified and admitted to the club and be subject to review or exclusion from the club based on your behaviour.
Open or free software may get labelled as 'unsafe' since it has contributions from 'uncertified' coders.
Thus I suspect the ramifications of professionalisation are a little to much to swallow for many used to the relative freedoms of the hi tech industry.
802.11b specifies a transport that goes as fast as 11mbps before DCF overheads are accounted for. If someone has a device that goes faster than this then it is not 802.11b. It is something else.
If it can still interwork with 802.11b devices then it will be doing so at a maximum of 11mbps.
If you want something faster, then why not use something that will not be obsoleted by a standardized solution such as 802.11a or the still-in-progress 802.11g?
270,000 licenses is too many to supply the needs of state workers, but may be very sesnible if it is to supply needs of online government services delivered over the internet to Californian residents.
If they cleaned up their processes and made most interaction with state government available to people over the internet, they may save bucketloads of time and money for everyone involved.
Of course this assumes that the government is competent to do so...
> (Ever tried to fix surface mount components with a soldering iron at your kitchen table? Don't!!)
I've always found that working with SMT is easier than through hole. You have gravity on your side. It will hold the component on the pad while you tack it in place.
Just use a decent soldering iron that has a small enough tip and don't make the mistake of using too small a tip. A too small tip doesn't hold enough heat to flow the solder onto larger SMT pins.
Also make good use of brush on flux and desolder braid. They are your friends when reworking SMT boards.
When laying out your own PCB, SMT components let you get away with drilling far fewer holes and zero ohm resistors let you 'jump' over tracks without using vias.
When it comes to probing, all your signals are generally available on one side. Most SMT parts (except BGA and LCC styles) don't shroud their leads like stand-up electrolytics and transistors do.
One of the primary barriers to messing with this sort of stuff in America is the crappiness of component supply for the hobbyist. I have yet to see anything that comes close to the likes of Radio Spares or Farnell in the UK.
As it happens, I am also dropping the cable TV due to a steadily deteriorating service and I have a satellite dish on order.
They have just announced price rises for the cable TV and they have been adding more and more channels, for which they demand more and more cash. Each new channel seems to slow down the already slow and crappy browser further and requires me to trawl through more and more unavailable channels to get to the few I can actually access.
In contrast, the satelite people have a nicer, faster, slicker browser, can sell me a combined Tivo/receiver box and for less money offer me more channels.
Maybe Comcast can stop the rot, but I'm not waiting to find out.
Myself and 100% of my friends who I know have cable with AT&T are in the process of switching to DSL and dropping AT&T due to the steady degradation of service, culminating in the @Home debacle.
Do Comcast know that AT&T may be about to lose a huge proportion of their customers?
I am an AT&T cable modem customer and have had the bad fortune of being transferred from @Home to AT&T's own ISP.
They suck massively.
1) Their support are not answering the phone
2) My IP address that has been static since I signed up over a year ago has suddenly changed and it appears that static addressing in any form has gone up in smoke. This screws anyone relying on a static IP.
3) They have been playing fast and loose with the service agreement (that I signed), but instituting an AUP (that I didn't sign)that directly contradicts it.
4) Their DNS service has been very erratic
This is not the sort of crap I want to be paying for and I am actively shopping for a replacement (I.E. DSL). I expect any other user who wants to do more than play with the latest microsoft browser will be doing the same and dumping AT&T as soon as possible.
This highlights an interesting split in the direction of modern operating systems research.
In the one corner we have MIT with Cesium that implements files and everything else in the form of an object oriented database.
In the other corner we have Bell-Labs/Lucent Plan 9 that implements databases and everything else in the form of files.
Plan 9 makes programming things easier since the file api lets you do most things. I wonder how much MIT's system makes things easier to do? I always found databases a pig to interface to (except in plan9).
Simpler IS better..
Machine Beauty is a book by David Gelernter that gives an excellent exposition of why beauty in technology is a good thing. I recommend it to anyone who wants a deeper understanding of the issues.
I have in my possession an original copy of the wonderful tome "Apple graphics and arcade game design" by Jeffrey Stanton, circa 1982. Apple as in Apple ][.
This covers such advanced topics as:
Programming Apple Hi-Res graphics from BASIC and machine language.
Raster Graphics and bit mapping
Moving things around the screen
Simulating gravity.
Doing cunning things with 6502 assembler to make things run faster.
As a pre-teen getting to grips with the innards of Apples, this one proved a lot easier to read than anything by Rodney Zaks.
Perhaps if everyone was required to write their first game in 6502 assembler on an Apple ][, they might acquire a better outlook on life. Or maybe just a headache.
Intentional radiators on planes (carried by passengers at least) are not presently permitted and so the whole issue is moot until there is a change in these regulations.
The reason that the Bluetooth and 802.11 crowd keeps working on this is that there is some hope for movement by the FAA/CAA/Other regulatory bodies. The arguments are pretty simple..
1) 2.4GHz is already polluted by wideband radiation in planes, since they use microwave ovens to heat the food. So there can be some level of confidence that you 802.11 tranciever isn't going to bring the plane down.
2) Electrical interference is a function of both the strength of the interferer and the succeptability of the interferree. Plane equipment is supposed to be built to stringent succeptability requirements. When someone says you phone/pda/toothbrush will interfere with the plane, ask why the equiment on the plane is operating outside the succeptability requirements mandated for planes. The responsibility to make interference not be a problem has be foisted on the passengers. This is a bad thing. Passengers forget to turn phones off all the time. It should be the job of the plane manufacturers to make planes safe in the presence of passengers.
3) There is lots of lobbying going on.
However being charged money to get on the net is not really a necessity is it? One person sets his laptop up as a gateway/NAT router and everyone else sets up as an ad-hoc 802.11 network. That would save people a bit of cash.
If all you wanted to do was play quake with your peers, you wouldn't need net access at all. 802.11 can work peer to peer. You don't need an AP just to communicate between a group of machines.
So lobby lobby your MP/Senator/FAA rep/garage mechanic/EU minister to change the rules. It will improve the quality of your traveling life.
I thought the worst tech job out there was a well paid position as a microelectronics designer in Oregon. It must be since I've been unable to find anyone willing or able to take up the job for bloody ages.
The ground rules are pretty much the same. If you are educated, have no criminal record and can find a company to sponsor you, then you will be able to get work authorization.
A thing to be aware of are that you will be going against the tide for a reason. EU techy citizens move to the US because of higher pay, lower taxes and a chance to work within a richer group of tech companies. You can expect lower pay, higher taxes and a more limited range of employers over there.
Nevertheless, I think its is pretty much a requirement of having an interesting life to live in more than one country. So go and do it.
This encapsulates two common misconceptions about Bluetooth. 1) Bluetooth is not a LAN. It is fow wire replacement. The primary profiles (headset, synchronization, serial device) support that assertion. An investigation of the piconet/scatternet architecture reveals just how inefficient Bluetooth would be at trying to route data through a network. In fact the spec reads more like the plan was to obstruct network functionality for the benefit of wire replacement functionality. Point to point punting of Audio and packet data fits easily and efficiently into Bluetooth. Routing does not. 2) You can't just stick a fat PA on the front and operate at 100s of meters. Part of the purpose is to allow many devices to interoperate. That is why the protocols are defined so much further up the stach than say 802.11. As a part of that, the 768kbps shared within a piconet may get pretty slow shared amongst a piconet. It would get stupidly slow shared between all the devices you could fit in 100s of meters (E.G.at a trade show). In addition, the high power devices sharing space with low power devices would lead to serious link budget imbalance in the RF. Rather like a pair of quiet guys trying to have a conversation at the front row of a moterhead concert. As far as the protocols go, they are not too open (the latest specs are available to the SIG signees , but don't worry, they're a boring read). However the IEEE are rewriting them as 802.15 and you'll find them to be better written. My advice, if you want to do Bluetooth protocol or application development is to concentrate on the upper layers (Above HCI), since the lower layers will get commoditized and componentized in due course and there is hardware involved, it won't be free anyway.
What I'm questioning is the use of linux in things that are an order of magnitude smaller. Like cell phones, handheld PDAs and personal location systems. The interfaces are generally dictated by the microcontroller and/or whatever ASIC might be in the system. Hard disks are out of the question.
Running traditional linux apps seems to be an unlikely goal - I'm not paying to run a webserver from my cell phone and X applications are not ideal for a 128x72 LCD screen.
Quite possibly, the need for a real filesystem might be a good dividing line. Where you need such a thing, linux gives you lots of goodies for providing and using the filesystem. In a system which has no such need, linux looks too much like an overwight dinosaur.
The things I look for in a microcontroller OS are stuff like..
A small footprint
Lightweight realtime multitasking (if multitasking is required)
Rapid boot (millseconds)
Persistent state support
Will fit in a single flash chip with space for the application
Hosts whatever protocol stack is going in the product (stuff like GSM, IS95, TCP/IP, IS707 seem to be the usual suspects).
Most importantly.. A really really good remote debugger, cross target development environment and totally uncrashable kernel
So could someone clue me in as to why I should use something like this rather than say Nucleus?