Indeed - a machine like the Sam Coupe is what Amstrad should have done to the Spectrum. The Coupe was a good upgrade on the Spectrum, and mostly compatible. Amstrad, owning the ROM code, could have made a machine like that completely compatible. Unfortunately, the makers of the Coupe, MGT just didn't have the oompfh to get the Sam Coupe out at a time when it would still sell well and didn't own the ROM,-and Amstrad wanted to concentrate on the PC so only made incremental improvements (the tape drive then built in disk drive) to the Spectrum.
Yes, the BBC Micro was definitely the mightiest of the 8 bits - without a doubt. The BASIC interpreter was definitely the best of the bunch (it had a built in assembler, too) and it had great support for hardware add-ons. We had Beebs at school and an Econet network (which IIRC was simply a matter of adding an extra chip to a socket in the machine). A friend and I wrote a MUD for the BBC and Econet, loosely modelled on Shades. It was an ungodly mix of BBC BASIC and 6502 assembly (and it was a surprise it ran at all), but it had client-server aspects as well as peer-to-peer aspects before either of us who wrote it had heard either of those terms. Fun times.
I've never seen a computer that used I/O mapped screen memory on a Z80. Memory mapping was natural for the Z80 too. The Spectrum's framebuffer was memory mapped and started at address 0x4000. Memory mapped I/O was also common on Z80 based machines, but having a 16-bit I/O address space as well meant that you could save valuable memory address space for...well, memory.
DRAM refresh was also not a problem for the Z80 - the Z80 actually had built in DRAM refresh circuitry - this is one reason it was so popular, because you didn't need a big pile of glue logic to do DRAM refresh because the Z80 provided a/RFSH pin which did this for you. No wait states were caused by the refresh.
The Z80 is not an 8080, it merely has binary compatibility with the 8080. The chip is electronically quite different.
Also, static RAM uses considerably less power than dynamic RAM. DRAM was cheaper, that's all.
Two inaccuracies: The fastest 6502 instructions were not one cycle, but two cycles. The fastest Z80 instructions were four cycles.
A very good 6502 programmer could write a program for the 1MHz 6502 in Commodore machines run as quickly as a run-of-the-mill Z80 programmer could on the Spectrum.
While the slowest 6502 cycles instructions were around 7 clock cycles, and the slowest Z80 instructions (the index register instructions) were real dogs, one or two of them taking up to 20 cycles to complete, this was more than made up by register pairing. A Z80 could do a 16 bit add in 11 cycles, where the 6502 would take on the order of 20 (and use more memory).
There was a lot more to the Z80 than a slight improvement over the 8080 - it had not a few but many more instructions (all the DD and ED prefixes), including the addition of extra registers such as the index registers. The Z80 also had more interrupt modes than the 8080, including the very useful IM 2 which meant you could trivially wrest interrupt control from the ROM program (which simply wouldn't be possible with an 8080 based machine). It not only had the block move instructions (LDIR and LDDR) but also block I/O transfer instructions and block search instructions which helped keep the memory footprint of many programs down.
* 48K is not 1/4 of 64K. Besides, the Spectrum had more memory available for the BASIC programmer than the C64. * It was a rubber keyboard, not chiclet. It was actually much better for playing games than a spring keyboard.
The colour attribute design wasn't a flaw - it was part of how they made the machine affordable by keeping it simple. The Spectrum's framebuffer is laid out in such a way that you can get a very good frame rate out of programs on the machine without requiring (expensive) hardware support. Commodore owned a semiconductor fab and so could afford to put lots of flashy hardware in the machine. Sinclair had to buy their chips on the open market, and make do with the ULA - the forerunner to the CPLD/FPGA - to provide I/O. The 1 bit per pixel main frame buffer overlaid by a simple 8x8 attribute system made it cheap to produce the hardware.
I presume you're from the US since you spelled colour the US English way. You also have to remember at the time, the early 1980s, Britain was a poor country. Remember how we had the first world and the third world? Some people theorized the 'second world' so to speak was countries like Russia. No one admitted to being a second world country, but in the early 1980s, Britain would have fit the bill for being "second world" - industrial strife, high unemployment, high inflation and generally poor economic conditions. Sinclair understood the market - the computer had to be fast and useful, and be able to rival the Commodore 64 and its ilk for entertainment, but yet be much, much cheaper - therefore Sinclair had to resort to various tricks to economize on the hardware. They couldn't throw silicon at it like Commodore could. The C64 cost around £400 in the UK in the early 1980s, and the 48K Spectrum was £170 at its launch (£120 for the 16K model).
The 6502 had fewer registers and fewer instructions - it took more code to do the same thing. A Z80 could do a 16 bit add in 11 cycles - it took the 6502 around 20 cycles to do the same thing. The fastest 6502 instructions took 3 clock cycles to complete, the fastest Z80 instructions took four.
Machines like the BBC Micro got better performance than the Spectrum not from the 6502, but because they had more hardware support which meant the CPU didn't have to do everything. But a BBC Model B, while undoubtedly a mighty machine and much more powerful than a Spectrum cost three times as much as a 48K Spectrum. Again, the Commodore 64, at the Spectrum's launch, was three times more expensive and had less RAM available to the user for BASIC programs.
The "classic" Z80 (as used in the Spectrum) is still made and can be bought from most electronics supplies firms (only in CMOS versions these days, but the CMOS version is a drop-in replacement for the old NMOS version). Zilog also make several advanced variants designed for microcontrollers, including one with a built in Ethernet MAC (the eZ80). They are cheap and easy to use, and are popular because of this.
No, I'm not joking. ZX BASIC was miles ahead of Commodore BASIC.
The CPU itself is also unarguably faster. While the 6502 and 6510 generally can perform more instructions per cycle (the fastest 6502/6510 instructions complete in 3 clock cycles rather than the Z80's 4 clock cycles), this is more than made up for by the Z80 having more registers and 16-bit register pairing - meaning programs need far fewer instructions to write. Add to that the Spectrum being clocked 3.5 times faster than the C64, it makes a noticable difference.
An example: to do 16 bit addition, the 6502 would need 20 cycles to do what the Z80 can do in a mere 11 cycles.
I've also made a 25th anniversary hardware project for the Sinclair Spectrum - an add-on board to be used for helping diagnose problems with sick Sinclair Spectrums:
It uses LEDs to display the test progress and status, so even if you can't get a picture out of the Spectrum, you can at least find out if the CPU and memory is working, and a good idea whether the ULA is servicable.
Not judging by quite a lot of people who I knew, many of them would get very belligerent with cops (even when they were stopped for reasons that were entirely justified).
Oh for heaven's sake, some jerk taunting other school students or teachers has NOTHING to do with learning. Get a grip. If they want to criticise teachers, they can do it in a polite and reasoned manner.
Have you ever tried to do one? It's almost entirely impractical, and you'll end up coming to the conclusion that "printf("Hello world\n");" infringes on someone's patent. I've been doing a bit of coding today for an embedded system I'm designing, and although all of the code is run of the mill trivial stuff, I'd not be surprised if I've probably infringed at least four or five patents in the process. Fortunately, I don't live in the US so I don't care.
Part of patent reform should include that if someone can prove beyond reasonable doubt that they independently came up with the idea that they are now being sued over, this should be evidence that the patent was not non-obvious in the first place (or else someone wouldn't have independently thought it up) - and this should automatically invalidate the patent.
The USPTO themselves admit that only 5% of patents are worthy, they even have a term for these 5% - "pioneer patents" - i.e. patents that are truly novel and non-obvious. The other part of patent reform should be that only these "pioneer patents" should be accepted.
I had a crappy Dodge Ram truck (with so many dents and dings that I nicknamed it "Others Dodge while I Ram" - although I was not responsible for any of the dents, I got it that way).
I was stopped FIVE TIMES in less than a year. The Police always claimed a valid reason, but I only ever once got a ticket (so I suspect once they had pulled me over and noted I was indeed licensed and insured and the vehicle wasn't stolen, they didn't write a ticket because what they told me was really just an excuse. Also, I make a point of being polite even when pulled over without justification, and I think the police are so shocked when someone's polite to them, they completely forget to write a ticket:-)).
Once I had some more money, and bought a two year old F150 in like new condition - with no change in driving style, I was only stopped once in the following five years, and that was entirely valid (I had forgotten to get the inspection done, and a cop at a junction noticed that the sticker was out).
I'm convinced I got stopped in the Dodge all the time was simply because it looked crappy.
The big deal was that it was done by a CMOS oscillator, i.e. something that can be fabbed in today's semiconductor factories in mass, not something that's limited to the laboratory.
By the way, I use my PowerBook as the world's most expensive Spectrum tape recorder - I use 'playtzx' to play TZX files to the machine (basically, a TZX file allows a Spectrum program to be encoded in a file in exactly the same way it was on a cassette, including flash loaders and other fancy loaders)
Probably best to go for even simpler than that - an old Sinclair Spectrum does all of the above, and it's small enough to be completely understandable in a short period of time. Plus Z80 assembly language is a bit nicer than 8088 (even though the ISA is related).
Tell me about it - I have one rubber key Spectrum, two Spectrum+ and a toast-rack 128K Spectrum. I had to repair the rubber keyed one and one of the Spectrum+ machines - both had bad 4116 (lower RAM) chips and bad keyboard membranes. By the way, you can buy brand new Spectrum keyboard membranes and rubber mats for a very reasonable price from http://www.rwapsoftware.co.uk/ . He's just had another run of them made.
I agree. "Blogosphere" is one of those awful dot-bomb era buzzwords, bandied about by marketdroids and the press to make them look cool and informed about the Internet. There seems to be an inverse relationship between the amount of times someone mentions "the blogosphere" and their understanding of how the internet actually works.
While I agree that the Iraq war was a complete waste of time and money, and has made things worse not better, we have to put some reality into the solar equation here.
Your calculations are predicated on 24 hours a day of direct sunshine. That doesn't happen, so you already must reduce your estimate by 50%. Most solar installations (particularly the kind of domestic on the roof kind) are fixed, so only really produce peak power within half an hour each way of local mid day, assuming they are perfectly positioned for the mid day sun. Three hours after mid day, the typical monocrystalline panel will be down to about 30% of peak output.
You actually need something like ten times the peak wattage to generate a given average wattage (so if your house on average uses 400 watts, you really need 4kW peak of solar to provide those paltry average 400w). Surface area is not a problem - most houses have plenty of surface area on their rooves even for panels with half the efficiency of today's monocrystalline panels. However, cost is a problem. For solar to ever become attractive, it needs to be about 1/20th of the price it is now per peak watt - currently, it's only cost-effective for buildings where getting the grid to the building will cost a good percentage of a solar installation (this incidentally is why you see them powering some electronic road signs now - it's cheaper to run them off a 50 watt panel than it is to get the electricity company to run a wire to the sign). Solar has lots of potential for domestic use if we can make it *cheap* - it's unobtrusive (unlike wind, which is cheap but you may annoy your neighbours) and needs little maintenance (current panels are guaranteed for 25 years) - it's just about twenty times too expensive for anyone except the biggest efficiency freak to consider. Personally, I can go for the low hanging fruit - riding my bicycle into work instead of driving saves more energy than powering my house from nothing but renewable energy (not to mention 125 miles a week of cycling over hilly terrain is pretty good for fitness).
PV cells are also very inefficient on a cloudy day.
I have an 80 watt peak panel. In direct sunshine, perpendicular to the panel, it can produce 80w. If there is a very thin layer of cirrus cloud, or 7 miles visibility in haze, it produces no more than 50w. If there is a thin layer of overcast cloud where you can still see faint shadows on the ground, it produces no more than about 15 watts. On a proper overcast day, the power produced is negligable (no more than 3 or 4 watts, well under 10% of rated power).
It depends on what measures are used for "efficency". If they are using sunlight converted to electricity, you are right. However, the article is a typical fluff piece (that I wish Slashdot wouldn't link to - Slashdot is for the more technically minded, so should be linking to technical articles, not stuff in the popular press or stuff that's so light in details it doesn't really tell you anything).
However, if by 'efficiency', they mean cost-per-watt of capacity, you certainly can (in theory!) make solar panels 60 times more efficient.
There's really no such thing as an oil company any more - they all call themselves ENERGY COMPANIES (which incidentally, is one of the signs that there isn't an awful lot of *cheap* oil left).
BP, for instance, make solar panels. You can bet if BP made a breakthrough in solar panel technology, they'd be rushing to market it.
And in any case, solar doesn't really compete with oil - the things that solar is good for are things that oil isn't and vice versa.
Slashdot Mirror
Indeed - a machine like the Sam Coupe is what Amstrad should have done to the Spectrum. The Coupe was a good upgrade on the Spectrum, and mostly compatible. Amstrad, owning the ROM code, could have made a machine like that completely compatible. Unfortunately, the makers of the Coupe, MGT just didn't have the oompfh to get the Sam Coupe out at a time when it would still sell well and didn't own the ROM ,-and Amstrad wanted to concentrate on the PC so only made incremental improvements (the tape drive then built in disk drive) to the Spectrum.
Yes, the BBC Micro was definitely the mightiest of the 8 bits - without a doubt. The BASIC interpreter was definitely the best of the bunch (it had a built in assembler, too) and it had great support for hardware add-ons. We had Beebs at school and an Econet network (which IIRC was simply a matter of adding an extra chip to a socket in the machine). A friend and I wrote a MUD for the BBC and Econet, loosely modelled on Shades. It was an ungodly mix of BBC BASIC and 6502 assembly (and it was a surprise it ran at all), but it had client-server aspects as well as peer-to-peer aspects before either of us who wrote it had heard either of those terms. Fun times.
I've never seen a computer that used I/O mapped screen memory on a Z80. Memory mapping was natural for the Z80 too. The Spectrum's framebuffer was memory mapped and started at address 0x4000. Memory mapped I/O was also common on Z80 based machines, but having a 16-bit I/O address space as well meant that you could save valuable memory address space for...well, memory.
/RFSH pin which did this for you. No wait states were caused by the refresh.
DRAM refresh was also not a problem for the Z80 - the Z80 actually had built in DRAM refresh circuitry - this is one reason it was so popular, because you didn't need a big pile of glue logic to do DRAM refresh because the Z80 provided a
The Z80 is not an 8080, it merely has binary compatibility with the 8080. The chip is electronically quite different.
Also, static RAM uses considerably less power than dynamic RAM. DRAM was cheaper, that's all.
Two inaccuracies:
The fastest 6502 instructions were not one cycle, but two cycles.
The fastest Z80 instructions were four cycles.
A very good 6502 programmer could write a program for the 1MHz 6502 in Commodore machines run as quickly as a run-of-the-mill Z80 programmer could on the Spectrum.
While the slowest 6502 cycles instructions were around 7 clock cycles, and the slowest Z80 instructions (the index register instructions) were real dogs, one or two of them taking up to 20 cycles to complete, this was more than made up by register pairing. A Z80 could do a 16 bit add in 11 cycles, where the 6502 would take on the order of 20 (and use more memory).
There was a lot more to the Z80 than a slight improvement over the 8080 - it had not a few but many more instructions (all the DD and ED prefixes), including the addition of extra registers such as the index registers. The Z80 also had more interrupt modes than the 8080, including the very useful IM 2 which meant you could trivially wrest interrupt control from the ROM program (which simply wouldn't be possible with an 8080 based machine). It not only had the block move instructions (LDIR and LDDR) but also block I/O transfer instructions and block search instructions which helped keep the memory footprint of many programs down.
* 48K is not 1/4 of 64K. Besides, the Spectrum had more memory available for the BASIC programmer than the C64.
* It was a rubber keyboard, not chiclet. It was actually much better for playing games than a spring keyboard.
The colour attribute design wasn't a flaw - it was part of how they made the machine affordable by keeping it simple. The Spectrum's framebuffer is laid out in such a way that you can get a very good frame rate out of programs on the machine without requiring (expensive) hardware support. Commodore owned a semiconductor fab and so could afford to put lots of flashy hardware in the machine. Sinclair had to buy their chips on the open market, and make do with the ULA - the forerunner to the CPLD/FPGA - to provide I/O. The 1 bit per pixel main frame buffer overlaid by a simple 8x8 attribute system made it cheap to produce the hardware.
I presume you're from the US since you spelled colour the US English way. You also have to remember at the time, the early 1980s, Britain was a poor country. Remember how we had the first world and the third world? Some people theorized the 'second world' so to speak was countries like Russia. No one admitted to being a second world country, but in the early 1980s, Britain would have fit the bill for being "second world" - industrial strife, high unemployment, high inflation and generally poor economic conditions. Sinclair understood the market - the computer had to be fast and useful, and be able to rival the Commodore 64 and its ilk for entertainment, but yet be much, much cheaper - therefore Sinclair had to resort to various tricks to economize on the hardware. They couldn't throw silicon at it like Commodore could. The C64 cost around £400 in the UK in the early 1980s, and the 48K Spectrum was £170 at its launch (£120 for the 16K model).
The Russians have done it.
p hp?t=15756
http://www.worldofspectrum.org/forums/showthread.
The 6502 had fewer registers and fewer instructions - it took more code to do the same thing. A Z80 could do a 16 bit add in 11 cycles - it took the 6502 around 20 cycles to do the same thing. The fastest 6502 instructions took 3 clock cycles to complete, the fastest Z80 instructions took four.
Machines like the BBC Micro got better performance than the Spectrum not from the 6502, but because they had more hardware support which meant the CPU didn't have to do everything. But a BBC Model B, while undoubtedly a mighty machine and much more powerful than a Spectrum cost three times as much as a 48K Spectrum. Again, the Commodore 64, at the Spectrum's launch, was three times more expensive and had less RAM available to the user for BASIC programs.
The "classic" Z80 (as used in the Spectrum) is still made and can be bought from most electronics supplies firms (only in CMOS versions these days, but the CMOS version is a drop-in replacement for the old NMOS version). Zilog also make several advanced variants designed for microcontrollers, including one with a built in Ethernet MAC (the eZ80). They are cheap and easy to use, and are popular because of this.
No, I'm not joking. ZX BASIC was miles ahead of Commodore BASIC.
The CPU itself is also unarguably faster. While the 6502 and 6510 generally can perform more instructions per cycle (the fastest 6502/6510 instructions complete in 3 clock cycles rather than the Z80's 4 clock cycles), this is more than made up for by the Z80 having more registers and 16-bit register pairing - meaning programs need far fewer instructions to write. Add to that the Spectrum being clocked 3.5 times faster than the C64, it makes a noticable difference.
An example: to do 16 bit addition, the 6502 would need 20 cycles to do what the Z80 can do in a mere 11 cycles.
I've also made a 25th anniversary hardware project for the Sinclair Spectrum - an add-on board to be used for helping diagnose problems with sick Sinclair Spectrums:
http://www.alioth.net/Projects/Spectrum-Diag
It uses LEDs to display the test progress and status, so even if you can't get a picture out of the Spectrum, you can at least find out if the CPU and memory is working, and a good idea whether the ULA is servicable.
Not judging by quite a lot of people who I knew, many of them would get very belligerent with cops (even when they were stopped for reasons that were entirely justified).
My cats keep my hours, but I didn't go out my way to train them to be like that.
Oh for heaven's sake, some jerk taunting other school students or teachers has NOTHING to do with learning. Get a grip. If they want to criticise teachers, they can do it in a polite and reasoned manner.
Have you ever tried to do one? It's almost entirely impractical, and you'll end up coming to the conclusion that "printf("Hello world\n");" infringes on someone's patent. I've been doing a bit of coding today for an embedded system I'm designing, and although all of the code is run of the mill trivial stuff, I'd not be surprised if I've probably infringed at least four or five patents in the process. Fortunately, I don't live in the US so I don't care.
Part of patent reform should include that if someone can prove beyond reasonable doubt that they independently came up with the idea that they are now being sued over, this should be evidence that the patent was not non-obvious in the first place (or else someone wouldn't have independently thought it up) - and this should automatically invalidate the patent.
The USPTO themselves admit that only 5% of patents are worthy, they even have a term for these 5% - "pioneer patents" - i.e. patents that are truly novel and non-obvious. The other part of patent reform should be that only these "pioneer patents" should be accepted.
Only twice?
:-)).
I had a crappy Dodge Ram truck (with so many dents and dings that I nicknamed it "Others Dodge while I Ram" - although I was not responsible for any of the dents, I got it that way).
I was stopped FIVE TIMES in less than a year. The Police always claimed a valid reason, but I only ever once got a ticket (so I suspect once they had pulled me over and noted I was indeed licensed and insured and the vehicle wasn't stolen, they didn't write a ticket because what they told me was really just an excuse. Also, I make a point of being polite even when pulled over without justification, and I think the police are so shocked when someone's polite to them, they completely forget to write a ticket
Once I had some more money, and bought a two year old F150 in like new condition - with no change in driving style, I was only stopped once in the following five years, and that was entirely valid (I had forgotten to get the inspection done, and a cop at a junction noticed that the sticker was out).
I'm convinced I got stopped in the Dodge all the time was simply because it looked crappy.
The big deal was that it was done by a CMOS oscillator, i.e. something that can be fabbed in today's semiconductor factories in mass, not something that's limited to the laboratory.
By the way, I use my PowerBook as the world's most expensive Spectrum tape recorder - I use 'playtzx' to play TZX files to the machine (basically, a TZX file allows a Spectrum program to be encoded in a file in exactly the same way it was on a cassette, including flash loaders and other fancy loaders)
I have:
four Sinclair Spectrums (rubber key 48K, two Spectrum+, and a toast rack Spectrum 128)
a MicroVAX
a Sun Ultra 5 (used as a server)
Out of all of them, the Sinclair machines are the most fun.
A little song that sums up why the Speccy was (and still is!) so much fun:
http://www2.b3ta.com/heyhey16k/heyhey16k.swf (warning, flash)
Probably best to go for even simpler than that - an old Sinclair Spectrum does all of the above, and it's small enough to be completely understandable in a short period of time. Plus Z80 assembly language is a bit nicer than 8088 (even though the ISA is related).
Tell me about it - I have one rubber key Spectrum, two Spectrum+ and a toast-rack 128K Spectrum. I had to repair the rubber keyed one and one of the Spectrum+ machines - both had bad 4116 (lower RAM) chips and bad keyboard membranes. By the way, you can buy brand new Spectrum keyboard membranes and rubber mats for a very reasonable price from http://www.rwapsoftware.co.uk/ . He's just had another run of them made.
:-)
I still enjoy many of the Spectrum games. This month, by the way, is the 25th anniversary, and I bought a T-shirt for the occasion
http://www.alioth.net/tmp/25YrsOfSpectrum.jpg
I'm such a geek...
I agree. "Blogosphere" is one of those awful dot-bomb era buzzwords, bandied about by marketdroids and the press to make them look cool and informed about the Internet. There seems to be an inverse relationship between the amount of times someone mentions "the blogosphere" and their understanding of how the internet actually works.
While I agree that the Iraq war was a complete waste of time and money, and has made things worse not better, we have to put some reality into the solar equation here.
Your calculations are predicated on 24 hours a day of direct sunshine. That doesn't happen, so you already must reduce your estimate by 50%. Most solar installations (particularly the kind of domestic on the roof kind) are fixed, so only really produce peak power within half an hour each way of local mid day, assuming they are perfectly positioned for the mid day sun. Three hours after mid day, the typical monocrystalline panel will be down to about 30% of peak output.
You actually need something like ten times the peak wattage to generate a given average wattage (so if your house on average uses 400 watts, you really need 4kW peak of solar to provide those paltry average 400w). Surface area is not a problem - most houses have plenty of surface area on their rooves even for panels with half the efficiency of today's monocrystalline panels. However, cost is a problem. For solar to ever become attractive, it needs to be about 1/20th of the price it is now per peak watt - currently, it's only cost-effective for buildings where getting the grid to the building will cost a good percentage of a solar installation (this incidentally is why you see them powering some electronic road signs now - it's cheaper to run them off a 50 watt panel than it is to get the electricity company to run a wire to the sign). Solar has lots of potential for domestic use if we can make it *cheap* - it's unobtrusive (unlike wind, which is cheap but you may annoy your neighbours) and needs little maintenance (current panels are guaranteed for 25 years) - it's just about twenty times too expensive for anyone except the biggest efficiency freak to consider. Personally, I can go for the low hanging fruit - riding my bicycle into work instead of driving saves more energy than powering my house from nothing but renewable energy (not to mention 125 miles a week of cycling over hilly terrain is pretty good for fitness).
PV cells are also very inefficient on a cloudy day.
I have an 80 watt peak panel.
In direct sunshine, perpendicular to the panel, it can produce 80w.
If there is a very thin layer of cirrus cloud, or 7 miles visibility in haze, it produces no more than 50w.
If there is a thin layer of overcast cloud where you can still see faint shadows on the ground, it produces no more than about 15 watts.
On a proper overcast day, the power produced is negligable (no more than 3 or 4 watts, well under 10% of rated power).
This is a monocrystalline solar panel.
It depends on what measures are used for "efficency". If they are using sunlight converted to electricity, you are right. However, the article is a typical fluff piece (that I wish Slashdot wouldn't link to - Slashdot is for the more technically minded, so should be linking to technical articles, not stuff in the popular press or stuff that's so light in details it doesn't really tell you anything).
However, if by 'efficiency', they mean cost-per-watt of capacity, you certainly can (in theory!) make solar panels 60 times more efficient.
There's really no such thing as an oil company any more - they all call themselves ENERGY COMPANIES (which incidentally, is one of the signs that there isn't an awful lot of *cheap* oil left).
BP, for instance, make solar panels. You can bet if BP made a breakthrough in solar panel technology, they'd be rushing to market it.
And in any case, solar doesn't really compete with oil - the things that solar is good for are things that oil isn't and vice versa.