Sex and drugs.. (not a troll, honest guv')
on
Trigger Happy
·
· Score: 2
A serious question. I'm not sure about the "size" of your economy.
How much is the porn industry worth?
How much is the cocaine industry worth?
i.e. Is $17billion really that much compared to other traditional "big bucks" industries.
I know they aren't even the biggest industries, but they are sufficiently taboo such that it's hard to estimate quite how large they are, and likewise sufficiently embarassing for people to not want to publicise them).
I cache my hifi remote control on the bed-side table.
My girlfriend flushes the cache, and puts it back to next to the hifi (this leads me to believe she has twice as many X chromosomes as necessary)
In this process, whatever state the remote is in one thing remains constant - there is only _one_ remote control..
In the olden days, RAM caches were not like that.
The cached information was duplicated in all lower cache levels. But, fortunately, if you read on in the article, you get to this:
"
Exclusive cache designs mean that the information contained within one layer is not contained within the layer above it. In the Thunderbird and Duron, this means that the information in the L2 cache is not contained within the L1 cache, and vice versa.
"
Now _that's_ more like the paradigm I'm used to.
I've been told it's a fair sized win for AMD's chips, but I'll reserve judgement until I get my hands on one. However, I'll say in public - I am a believer...
In some ways you are demonstrating the effect of the cache, yes. However you assume 2-way associativity. Some processors are 4-way (increase the number of accesses to be 5 with the same low 12 bits).
_BUT_ you're also forcing the "cache unfriendly version" to have several other speed problems:
- It isn't fair to certain processors as there are restrictions on using addresses with the same low 4 address bits back-to-back (original Pentium certainly).
- You've chosen to access non-aligned data (accessing 16 bit values on an odd, and cache line spanning, address)
References:
Abrash - The Zen of Assembly Language Optimisation
Agner Fog - Pentopt (www.agner.org methinks)
Unhappy with the performance of someone elses memory hungry code (50-100MB working RAM footprimt) I wrote my own version of the utility. It was only marginally faster than the original. However, I increased the performance by a factor of 10 when I realied that I could cache up jobs to do, then in turn perform those jobs on each 4MB chunk of data (Dec Alpha with 4MB L3 cache). I managed to increase performance even further by aiming the code at the L2 cache instead! The total number of bytes read/written was identical, but simply changing the order in whihc they were done increated performance 12-15 times.
However, as soon as you do take into account caching issues, you sometimes start making non-portable decisions. (not always though, as generally most architectures have the problem but the lines are simply drawn in different places).
We've seen the power of the masses before (distributed.net type stuff - DESCrack, RC5, GIMPS etc...). We don't use one supercomputer to attack these problems, why should we want a super-server to solve the usenet archiving problem?
Let it be chopped into manageable sized chunks, run by people who have a genuine interest and relevance to the field under discussion. If noone wants to archive alt.fan.right.said.fred then it fades away, but if the official fan club wants to, let them. So this has the benefit of maybe thinning down the archives.
Also remember about compresion. The amjoity of a usenet post is redundant, or at least highly predictable, and will compress down immensely.
It's write-once read-many, so the compression scheme can be chosen to be decoder-light.
The small glimmer of truth in what you say is that maybe some large non-commercial center should manage the archives. Where? MIT perhaps? Whatever. That would be great. However, every American tax payer would then be paying for it. So Europeans and South Africans and Australians and Japanese and Brasilians and... would be getting the service for free.
There is no such thing as a free lunch.
FatPhil
(A European who thinkgs that we Eurpoeans should pay our part to keep the archives up and running)
Re:Selling USENET Archive?
on
Deja For Sale
·
· Score: 1
You did ask for them to archive your post by omitting the X- header which turns off archiving...
FatPhil
False Memory Syndrome?
on
Deja For Sale
·
· Score: 1
Shoot me if I'm wrong, but I remember www.dejanews.com being _web accessible_ back in 1993 (which was the first time I used it, it may go back much further). The article claims that the company Deja was set up in 1995 in order to do the above. I think that is revisionist history. My version is that in 1995 they decided that they wanted to _make a profit_ from their web site, that's all.
Indeed.
Some people have forgotten that we do have "healing" in computer systems.
In large distributed computing tasks, if one test is taking too long and the node which should be processing it seems to have lost contact, the master node will reassign the job to another node.
Nothing particularly clever about that, as long as it's designed in from the outset.
FatPhil
Beware Penrose. He's a bit of a Mathematical Fly in the Physicists Ointment. The books you mention are mainly quite good, he lulls you into a state of trust by feeding you lots of pretty mathematical models, and then WHAM! - he throws a curveball at you so wild that it has no direct connection with what you've just read, but as you are now numbed into trusting everything he says he gets away with pushing mumbo-jumbo on you for the last few chapters.
It happens in many books. Check out "Evolving The Mind" (go to a library - _don't_ buy it like I did), as a wonderful introduction to how the brain evolved in animals, but suddenly about half way through the book the author flips and becomes a mumbo-jumbo spouter...
Vexed. Oh god. You've put the game back in my mind. I used to move blocks around all day and all night, haunting me. Great game, lousy dream though.
FatPhil
Give me some credibility - I only play quake with only one hand when I'm masturbating furiously with the other. In which case simple matters like getting fragged really become unimportant.
Anyway, did you do an EMC test on the heaters?
If it's throwing off heat, it's either throwing off light or sub-luminal frequencies?
In fact, just an "on" light would be good enough to destroy their claim!
I bought one for home use 2 years ago.
It was more expensive than my 486 was 8 years ago, but still to this day, it is the best machine I own. The build quality is astounding, and it is a joy to program (I'm a mathematician, and 64 bit registers rock my world).
I believe that when I buy a 1G Athlon it won't be much better for me than the Alpha, and of course that's 2-3 years later!
I use Linux, the distribution is irrelevant as long as you get one that is well enough tested before they put it on the shelves.
The AlphaLinux RedHat mailing list have provided answers to every question I've ever asked about every issue, and they don't even ask which distribution I use...
If you wish for your post to be taken seriously, then please talk sense.
As far as I know the Athlon (note the spelling) only has three pipelines which can perform operations like multiplies. Where does your figure of 6 come from?
The area of Palestine reserved for a Jewish state by Balfour et al. was less than 50% of the current area of Palestine at the time. Currently the Israelis are controlling over 100% of that area.
That's the same as The United states of America expanding into Canada and Mexico.
The US couldn't do that though, because they give their best assault weapons away - to the Israelis.
That little gem from a reasonably highly ranked officerin the USAF. (who was complaining about how his men weren't well enough supplied)
FatPhil
(Who's lived in the part of Palestine which was never supposed to be anything but Palestine).
In the past I've programmed the TI C80MVP which was a multi-processor chip. It had 4 Integer DSP chips (32bit, 64bit accumulator) and a Master/FP chip all on one slab of silicon. Each of the DSP cores was pipelined so as to permit 3 simultanious MACCs (multiply and add in one operation - a side effect of the multiplication algorithm used is that you can squeeze in an extra addition _for free_). Therefore the chip could (should!) have 12 simultanious multiplies on the go.
That was back in the early/mid 90s.
A factor of 10 in 7 years - that's less than Moore's law.
TI obsoleted the chip years ago. It was a dog to program. I believe the C6000 series superseded (that's the correct spelling, BTW, fr L. Sedere, to sit) it. The C6000 series was intended to be as powerful but far easier to program.
It's non-linear optics.
Basically if you pump the signal beyond the normal linear behaviour you an get different frequencies to behave completely differently from each other. You can get a "mirror" using total internal reflection for any single frequency in the FDM.
Non-linear here means that when you add the signals together you don't get the sum of the inputs as the output!
Hashing does take O(ln(n)) time for the most general case (n is the number of objects stored).
If the guy says "you can't fill it beyond x%" then he's not permitting you to have arbitrary n, and the order expression means nothing - i.e. his O(1) expression means nothing. He asks for your maximum, calculates a lowest upper bound, and creates an algorithm which always works with time less than that upper bound. Voila, constant time. But not O(1)!
I know I'm not beiong coherent, but I've had a long day.
In order to decide between (i.e. hash) N objects, at least lg(N) bits must be calculated for the hash value. No you can't do them all at once, as I'm chosing N=2^1048576. When you fix that, I'll increase N again. I'm allowed to do that, you're not allowed to stop me.
Firstly - in user-interface terms, a drag is distinct from a click.
How to get past Tescos -
If the patent is to exclude Tescos, then it must be formulated specificly to do so. Find something that tescos does that you don't want, and patent something without that thing. (so if Tescos pops up a dialogue box at a particular, patent the process _not_ using a dialog box at that stage).
Now the Improvement -
Reorder the process, and add one extra feature which makes perfect sense in the real world.
Reorder: patent the entering of payment details _before_ the shopping commences, perhaps.
Extra feature: patent the entry of a self-imposed spending-limit, and add the feature of telling you how close you are to your spending limit after each item selection.
Logic error!
Less than 1% of a huge amount is not necessarily a huge amount.
If it were 1.0e-36% it would be a small amount, wouldn't it?
However, some tar heeled chappies (I think) indicates the composition is 0.62%, but this is from very old data. Far more than Carbon, for reference. So there is some hope...
FatPhil
The boot 'ROM' code is probably in Flash, so could be blasted with x86 code. The peripherals which have boot code have x86 code already, as the Alpha supports x86 emulation even at the lowest levels.
FatPhil
Slashdot Mirror
A serious question. I'm not sure about the "size" of your economy.
How much is the porn industry worth?
How much is the cocaine industry worth?
i.e. Is $17billion really that much compared to other traditional "big bucks" industries.
I know they aren't even the biggest industries, but they are sufficiently taboo such that it's hard to estimate quite how large they are, and likewise sufficiently embarassing for people to not want to publicise them).
FatPhil
(Sex budget $0, Drugs budget 0$, FYI)
I cache my hifi remote control on the bed-side table.
My girlfriend flushes the cache, and puts it back to next to the hifi (this leads me to believe she has twice as many X chromosomes as necessary)
In this process, whatever state the remote is in one thing remains constant - there is only _one_ remote control..
In the olden days, RAM caches were not like that.
The cached information was duplicated in all lower cache levels. But, fortunately, if you read on in the article, you get to this:
"
Exclusive cache designs mean that the information contained within one layer is not contained within the layer above it. In the Thunderbird and Duron, this means that the information in the L2 cache is not contained within the L1 cache, and vice versa.
"
Now _that's_ more like the paradigm I'm used to.
I've been told it's a fair sized win for AMD's chips, but I'll reserve judgement until I get my hands on one. However, I'll say in public - I am a believer...
FatPhil
You cheeky man you!
In some ways you are demonstrating the effect of the cache, yes. However you assume 2-way associativity. Some processors are 4-way (increase the number of accesses to be 5 with the same low 12 bits).
_BUT_ you're also forcing the "cache unfriendly version" to have several other speed problems:
- It isn't fair to certain processors as there are restrictions on using addresses with the same low 4 address bits back-to-back (original Pentium certainly).
- You've chosen to access non-aligned data (accessing 16 bit values on an odd, and cache line spanning, address)
References:
Abrash - The Zen of Assembly Language Optimisation
Agner Fog - Pentopt (www.agner.org methinks)
FatPhil
Unhappy with the performance of someone elses memory hungry code (50-100MB working RAM footprimt) I wrote my own version of the utility. It was only marginally faster than the original. However, I increased the performance by a factor of 10 when I realied that I could cache up jobs to do, then in turn perform those jobs on each 4MB chunk of data (Dec Alpha with 4MB L3 cache). I managed to increase performance even further by aiming the code at the L2 cache instead! The total number of bytes read/written was identical, but simply changing the order in whihc they were done increated performance 12-15 times.
However, as soon as you do take into account caching issues, you sometimes start making non-portable decisions. (not always though, as generally most architectures have the problem but the lines are simply drawn in different places).
FatPhil
We've seen the power of the masses before (distributed.net type stuff - DESCrack, RC5, GIMPS etc...). We don't use one supercomputer to attack these problems, why should we want a super-server to solve the usenet archiving problem?
Let it be chopped into manageable sized chunks, run by people who have a genuine interest and relevance to the field under discussion. If noone wants to archive alt.fan.right.said.fred then it fades away, but if the official fan club wants to, let them. So this has the benefit of maybe thinning down the archives.
Also remember about compresion. The amjoity of a usenet post is redundant, or at least highly predictable, and will compress down immensely.
It's write-once read-many, so the compression scheme can be chosen to be decoder-light.
FatPhil
The small glimmer of truth in what you say is that maybe some large non-commercial center should manage the archives. Where? MIT perhaps? Whatever. That would be great. However, every American tax payer would then be paying for it. So Europeans and South Africans and Australians and Japanese and Brasilians and ... would be getting the service for free.
There is no such thing as a free lunch.
FatPhil
(A European who thinkgs that we Eurpoeans should pay our part to keep the archives up and running)
You did ask for them to archive your post by omitting the X- header which turns off archiving...
FatPhil
Shoot me if I'm wrong, but I remember www.dejanews.com being _web accessible_ back in 1993 (which was the first time I used it, it may go back much further). The article claims that the company Deja was set up in 1995 in order to do the above. I think that is revisionist history. My version is that in 1995 they decided that they wanted to _make a profit_ from their web site, that's all.
FatPhil
Indeed. Some people have forgotten that we do have "healing" in computer systems. In large distributed computing tasks, if one test is taking too long and the node which should be processing it seems to have lost contact, the master node will reassign the job to another node. Nothing particularly clever about that, as long as it's designed in from the outset. FatPhil
Beware Penrose. He's a bit of a Mathematical Fly in the Physicists Ointment. The books you mention are mainly quite good, he lulls you into a state of trust by feeding you lots of pretty mathematical models, and then WHAM! - he throws a curveball at you so wild that it has no direct connection with what you've just read, but as you are now numbed into trusting everything he says he gets away with pushing mumbo-jumbo on you for the last few chapters.
It happens in many books. Check out "Evolving The Mind" (go to a library - _don't_ buy it like I did), as a wonderful introduction to how the brain evolved in animals, but suddenly about half way through the book the author flips and becomes a mumbo-jumbo spouter...
FatPhil
Vexed. Oh god. You've put the game back in my mind. I used to move blocks around all day and all night, haunting me. Great game, lousy dream though. FatPhil
Give me some credibility - I only play quake with only one hand when I'm masturbating furiously with the other. In which case simple matters like getting fragged really become unimportant.
FatPhil
Is it some kind of corny pun on it being a "true theatre" experience?
Back to the drawing boards guys - how about "large domed screen" (TM) technology?
FatPhil
Say I'm playing quake using only the keyboard?
I get shot and my mouse commits ritual suicide off my desk?
I've already wasted money on a rumble pack, never again, a gimmick of the naffest kind.
FatPhil
(love the .sig)
Anyway, did you do an EMC test on the heaters?
If it's throwing off heat, it's either throwing off light or sub-luminal frequencies?
In fact, just an "on" light would be good enough to destroy their claim!
Phil
I bought one for home use 2 years ago.
It was more expensive than my 486 was 8 years ago, but still to this day, it is the best machine I own. The build quality is astounding, and it is a joy to program (I'm a mathematician, and 64 bit registers rock my world).
I believe that when I buy a 1G Athlon it won't be much better for me than the Alpha, and of course that's 2-3 years later!
I use Linux, the distribution is irrelevant as long as you get one that is well enough tested before they put it on the shelves.
The AlphaLinux RedHat mailing list have provided answers to every question I've ever asked about every issue, and they don't even ask which distribution I use...
FatPhil
If you wish for your post to be taken seriously, then please talk sense.
As far as I know the Athlon (note the spelling) only has three pipelines which can perform operations like multiplies. Where does your figure of 6 come from?
The area of Palestine reserved for a Jewish state by Balfour et al. was less than 50% of the current area of Palestine at the time. Currently the Israelis are controlling over 100% of that area.
That's the same as The United states of America expanding into Canada and Mexico.
The US couldn't do that though, because they give their best assault weapons away - to the Israelis.
That little gem from a reasonably highly ranked officerin the USAF. (who was complaining about how his men weren't well enough supplied)
FatPhil
(Who's lived in the part of Palestine which was never supposed to be anything but Palestine).
It's horses for courses.
In the past I've programmed the TI C80MVP which was a multi-processor chip. It had 4 Integer DSP chips (32bit, 64bit accumulator) and a Master/FP chip all on one slab of silicon. Each of the DSP cores was pipelined so as to permit 3 simultanious MACCs (multiply and add in one operation - a side effect of the multiplication algorithm used is that you can squeeze in an extra addition _for free_). Therefore the chip could (should!) have 12 simultanious multiplies on the go.
That was back in the early/mid 90s.
A factor of 10 in 7 years - that's less than Moore's law.
TI obsoleted the chip years ago. It was a dog to program. I believe the C6000 series superseded (that's the correct spelling, BTW, fr L. Sedere, to sit) it. The C6000 series was intended to be as powerful but far easier to program.
FatPhil
It's non-linear optics.
Basically if you pump the signal beyond the normal linear behaviour you an get different frequencies to behave completely differently from each other. You can get a "mirror" using total internal reflection for any single frequency in the FDM.
Non-linear here means that when you add the signals together you don't get the sum of the inputs as the output!
FatPhil
You don't need to convert them into Electriccal signals.
The reason I know this might be the same reason that I know that Marconi is a totally sucky company to work for.
FatPhil
Hashing does take O(ln(n)) time for the most general case (n is the number of objects stored).
If the guy says "you can't fill it beyond x%" then he's not permitting you to have arbitrary n, and the order expression means nothing - i.e. his O(1) expression means nothing. He asks for your maximum, calculates a lowest upper bound, and creates an algorithm which always works with time less than that upper bound. Voila, constant time. But not O(1)!
I know I'm not beiong coherent, but I've had a long day.
In order to decide between (i.e. hash) N objects, at least lg(N) bits must be calculated for the hash value. No you can't do them all at once, as I'm chosing N=2^1048576. When you fix that, I'll increase N again. I'm allowed to do that, you're not allowed to stop me.
FatPhil
Firstly - in user-interface terms, a drag is distinct from a click.
How to get past Tescos -
If the patent is to exclude Tescos, then it must be formulated specificly to do so. Find something that tescos does that you don't want, and patent something without that thing. (so if Tescos pops up a dialogue box at a particular, patent the process _not_ using a dialog box at that stage).
Now the Improvement -
Reorder the process, and add one extra feature which makes perfect sense in the real world.
Reorder: patent the entering of payment details _before_ the shopping commences, perhaps.
Extra feature: patent the entry of a self-imposed spending-limit, and add the feature of telling you how close you are to your spending limit after each item selection.
FatPhil
Logic error!
Less than 1% of a huge amount is not necessarily a huge amount.
If it were 1.0e-36% it would be a small amount, wouldn't it?
However, some tar heeled chappies (I think) indicates the composition is 0.62%, but this is from very old data. Far more than Carbon, for reference. So there is some hope...
FatPhil
The boot 'ROM' code is probably in Flash, so could be blasted with x86 code. The peripherals which have boot code have x86 code already, as the Alpha supports x86 emulation even at the lowest levels. FatPhil