Anybody who has read the Revised Report on Algol 68 knows that the language as defined there consists of the set of production trees of the grammar of the language.
Each production tree needs to be represented in a representation language, which is the form we usually seen when we think about programming languages. However, back then machines had such wildly differing wordsizes and character sets that this distinction was necessary. After publication of the Revised Report, with its included representation meant publication of programs, but not for actual implementation on hardware, they developed a `standard' hardware representation. Even in that representation there were multiple representations of the keywords in the language, e.g.
.WHILE
and
WHILE
were allowed representations for the exact same keyword.
Implementors were strongly suggested to use the standard hardware representation, but if documented, they were fully free to chose whichever representation they like. Hence, an implementation may perfectly well use
I've carefully considered which phone to choose when last upgrading. The balance between the 6310i and the T610 was as follows:
Features they've both got: - triband GSM - bluetooth, gprs, wap - good sound quality - well organized menus - nice shape - no fiddly moving parts - addressbook - calendar
Features the 6310i is better at than the T610: - reception - power connector
Features the T610 is better at than the 6310i - *will* sync with iSync, as opposed to the 6310i which will not - higher resolution color screen - built in toy-camera - very much nicer ring - smaller without getting unusable
Now, given that I live in Europe and we generally get good reception *anyway*, I didn't really care about the reception quality. I *do* care about the iSync compatibility though. As I carry my phone practically everywhere, that means I have my calendar with me at all times. I cannot get over how extremely useful this is. Oh, and dialling from the full addressbook in my iBook has it's advantages as well. Nowadays I just keep the numbers of people that I actually tend to call in my phone, making dialling from the phonebook quite a bit quicker.
If Nokia were to come out with a worthy successor to the 6310i that had roughly the same format but with a higher resolution screen and SyncML capability... that phone would absolutely rock and be my first choice. Unfortunately, the T610 currently beats the 6310i on features... and it attracts more attention from women as well. (Actually, almost all other owners of T610's I've seen are women...).
Given the fact that the people working on formally proving the maths that go into Intel's new processors is correct have written a book on how to prove that the maths is correct, I would say, *yes*, it will be able to do Math correctly.
For more info, you might look at John Harrisson's pages at http://www.cl.cam.ac.uk/users/jrh/ . He's at Intel doing some of the work, and he wrote a PhD on how to do the stuff he's doing for Intel nowadays.
Now, I may be just stupid, but I'd say the people at the High Field Magnet Laboratory in Nijmegen have a much stronger claim to world records... (33T continuous, 60T pulsed).
Interesting that this should trigger a YRO-like response, where cows have had to wear a unique ID printed/burned on a yellow eartag, one for each ear, for years already (at least in the Netherlands, but I've seem them in Germany as well).
Doing identification by retinal scanning seems *much* more humane to me than identification by plastic yellow flaps shot through the ears. It'd take us back to the times when identification was mostly done based on the cow's hide patterns, but without the trickiness in identifying the pattern if the hide has been damaged...
The fire did not start in the server rooms. What happened was that a fire started in one of the lecture rooms (and a smallish one at that, number A108) that just happened to be in the same wing of the TWRC building that also housed the server rooms (yes, multiple). It then proceeded to take out two entire wings and threaten other buildings nearby as well as the library.
Now, I would *love* to see a halon system capable of stopping that...
Owh, and the fire seems to be under control by now, as evidenced on http://webcam.traserv.com/ (which you can contrast to http://www-infstud.sci.kun.nl/~arthurvl/ispy.jpg (taken at about 09:40 CET this morning)).
Because an extra battery on a 14.1" iBook gives you merely 55 Wh extra power, whereas this gives you 130Wh extra, while also enabling you to not switch batteries. So, the difference is between 110 Wh in total with a batteryswap (and hence a hibernate or shutdown) in the middle or 185 Wh without the need to swap batteries. I know what I would choose.
(*drool* 185 Wh... that amounts to about 18 hours of computing off the grid!)
Simplest solution: go read-only for system-data
on
fsck-less Booting?
·
· Score: 3, Insightful
The simplest solution is to go read-only for all system data, such as binaries and static configuration. Even better: use something of a commit-system to commit configuration changes to disk and have the disk be read-write only when committing changes.
Even though you'll still run with fsck this will not be a problem, as stuff can't have changed for reasons other than hardware failure... and you're not going to work around that in software anyway.
Dumbbells, chairs & broomsticks
on
Exercise for Geeks?
·
· Score: 2, Informative
As with all activities: use your mind. What you want to be doing is exercise your muscles. While this can be done without any extra weights, most fit people will only really benefit with extra weights. A good set of dumbbells shouldn't set you back more than some EUR 60 (or $ 60).
Now, given that, you want to try to target particular muscles. Learn the moves that exercise the biceps, triceps, deltoids etc. Step on and off chairs to train your legs. Do pushups. Do situps. Stretch.
The problem with all this is discipline: it really doesn't help if you don't at least do this for about 20 minutes every day or 40 minutes every other day.
Ofcourse, you could also go for ballroom dance, historic dancing, indoor climbing, acrobatics, juggling...
If you are not too concerned about processorperformance, you might want to look
at Soekris Technologies' Net4501 or Net4521.
It is not very expandable, but seems to fulfill your requirements nicely. More information is at Soekris' website. Other options would be the PC104 series of modules... but you'd have to find your own enclosures.
How very interesting to note once again that Americans tend to think that everything important in computers has been American. For instance, I don't see the Electrologica machines, the X1 and X8, mentioned anywhere, even though they were the first to incorporate interrupts. Oh well.:)
Re:Proof that sed can in fact do everything...
on
SedSokoban
·
· Score: 1
I just saw the spazioweb.inwind.it/seders website,
and noticed that Christophe Blaess built a Turing machine emulator in sed as well last year... apparently there's more people as geeky as me.
Anyway, his implementation has the number of characters the sed-implementation allows minus three as a limit on the total number of states in the Turing machine, which mine doesn't.
Proof that sed can in fact do everything...
on
SedSokoban
·
· Score: 2, Interesting
Well, this seems as good a place as any to mention the sed script I wrote some 4 years ago that emulates a Turing machine. The script is actually capable of performing any calculation one can write a Turing machine program for... addition of two numbers is provided as an example.
It is very interesting to see the fairly old technology of Pie menus implemented yet again. Somehow, one would think that there should be a lot more implementations, given that pie menus show up again and again...
> Hackers? In the land of legal weed and prostitution? The Dutch better be preparred.
Actually, weed is quite illegal in the Netherlands. However, it is officially tolerated.
Prostitution is legal, though, yes, so if that gets you going, please come.
Note that Dutch law is *very* strict on cracking however. It has been forbidden to crack accounts on computers for some seven years already, and people *have* been arrested for such acts. Besides, the Acceptable Use Policy is pretty clear on what is and is not accepted behaviour (basically: be nice to the network or the network crew won't be to you). Yes, the Dutch *are* well prepared.:)
First a bit of background: I'm 25 and therefore technically one generation older than you are. However, my little sister is 15, so technically in your generation. My little sis and I get along splendidly, but then again, she always had two older brothers to hang around with. She also gets along fine with people her own age, and is really very much a 15-year old.
As I am the oldest in my family, when I was fifteen I had no such 'luck'. I have always been youngest in class, and always spent time with people at least roughly a year older than I was. Therefore I was always striving to keep up, and usually succeeding... but for the social aspects of life.
Anyway, I was wondering, what are the ages of the people you regularly hang out with?
And what do you think of people that are your age?
One terabyte, for your information, is *not* 1024 gigabyte. If you had taken a look at the definitions of the prefixes in the systeme international, you would have known that tera is the prefix for 10^12, not for
2^40, as you seem to imply here. The systeme international actually *does* have prefixes for binary powers. 2^40 bytes would be a tebibyte, using those prefixes.
As not such an old fart that has been "programming" since he was nine,
I must say that at the age of 15, when I tried to learn OOP, I had much difficulty grasping the concepts. This may have had to do with the very limited set of articles the library had on it, but I really do think I wasn't able to comprehend it at that time.
On a side note, a lot of the concepts in programming are inherently hard. The concept of a variable is definitely not trivial. In the past year I've been teaching a group of 6 12 and 13 year olds to program. We started out teaching them recursive refinement (which is really easy and they really grasped quite well). Ofcourse, next you have to introduce variables. This is HARD. It took us wel over a month to get the kids to internalize the concept of a variable. And even now they occasionaly miswrite assignments, with the destination and source of the assignment swapped.
For those of you that might think teaching OO before teaching what a variable is: DON'T! We've tried that at the university with second year chemistry students. They really had trouble following the course, until we reclarified what a variable is.
In the Netherlands trade secret protection does not hold if the group of people the secret is distributed to is too large. This holds *even* if each and every person in that group has vowed to keep it secret. The point is that a large enough group of people in and of itself is a `public', so distributing a trade secret to a large enough group of people implies publishing the trade secret and therefore lifting trade secret status. Neat, huh?:)
My response to that congressional aide: It's disingenuous to use terms like "theft" and "piracy," ancient notions of law and property, in the 21st Century. They have little contemporary meaning in cyberspace.
I've got to disagree with that. Theft on the net is still theft. Yes, the original work is still present, you're not depriving anyone of someone they own, but if someone chooses to use the net to sell something then taking it without paying is still theft. It maybe more benign than breaking and entering, but given enough people doing it it can be more damaging.
Note that one can only steal actual property. Stealing ideas is a non-concept. There is no such thing as "owning" an idea. The problem lies in the definition of the word idea. Is software an idea, an implementation of an idea or a description of an idea? If it is an implementation, it should logically be patentable... if it is a description, it should be copyrightable (these two are significantly different).
Copyright nor patents preclude free distribution. However, currently there are problems applying both copyright and patents to new developments. With the extremely small division between an idea and it's implementation in software and companies patenting as broad as they possibly can we see the patent system breaking down not because of inherent flaws in it, but merely because of insufficient care exercised in determining which patents are valid and which are not.
I think the point I'm trying to make is to let people realize that crying "the system is breaking down, therefore it must be fundamentally wrong" is pretty stupid. If my PC were to run Amazon, it would break down pretty quickly... but that does not give me any reason to think Amazon's systems are built on fundamentally different concepts than my PC is...
I do agree that the current system is too expensive to keep on track, and that would be a good reason to redesign it. That does not mean it's design is inherently wrong...
Actually, if you were to compile C code to the 128 bit VLIW of the Crusoe, you *could* be benefiting from the runtime aliasing, as you would also have access to the ldp (load-and-protect) and stam (store-under-alias-mask) instructions. The problem however is in the fact that Transmeta reserves the right to change the instructionset underlying the CodeMorphing software with each and every revision of the processor. And that would mean rewriting the compiler...
As the article states, the UCITA can, given certain circumstances, be applied to the sale of computers. Now, I could easily see Microsoft[1] buying up hardware manufacturers and sell the computers with a license clause effectively stating that one cannot run anything but Microsoft software on the hardware.
Wouldn't this be *terrible* for open source?
[1] Microsoft is just taken as shorthand for `big evil software empire', and you can supply your own variant any time.
You are of course completely right in saying the measurement is part of the science involved, and without measurement any attempts at improvement are based on guesses at best. However, the problem with most management methods is that they do not put the measurement tools in to the work flow of the workers but only graft them on to the side.
In that case updating the measurements *is* auxiliary to the actual job at hand, and will not be done.
As an example: an editorial board can benefit greatly from knowing how long diffent stages of editing an article take, and how well a certain editor performs at each stage. Maintaining this information is a major hassle, however, as it has nothing to do whatsoever with editing articles. But, if one incorporates maintaining this information in the meta-work of finding the next article to work on, it suddenly is incorporated in the normal work to do, and the information will be kept up to date.
Errr... *actually* the court ruled that Karin Spaink did not infringe at all.
The court *did* rule however that ISP's *would* commit copyright infringement if a page on their machines contains a link to a copyrighted page *and* the owner of the copyrights on said page made it clear to the provider that it considered the link to be a copyright infringement *and* the ISP will not remove the link.
Basically, said ruling makes either the ISP or the copyright owners judges insofar that *they* must now decide what does and does not constitute copyright infringement. And that is *really* scary stuff.
In case you guys haven't heard of it yet: in a court ruling in a case of Scientology against a number of providers a judge ruled as follows (straight from a translation which can be found at http://www.xs4all.nl/~kspaink/cos/verd2eng.html): DECLARES it to be the law that by having a link on their computer systems which when activated brings about a reproduction of the works that CST has the copyright to on the screen of the user, without the consent of the plaintiffs, the Service Providers are acting unlawfully if and insofar that they have been notified of this, and moreover the be reasonably doubted, and the Service Providers have then not proceeded to remove this link from their computer system at the earliest opportunity.
Now this can be extrapolated to mean that *any* company or person that complains to a service provider about a link to works that company or person has copyrights for can force the service provider to remove the link under penalty of law... I'd personally love to see (say) Universal take on (say) Altavista like this...
Slashdot Mirror
knows that the language as defined there consists of the
set of production trees of the grammar of the language.
Each production tree needs to be represented in a
representation language, which is the form we usually
seen when we think about programming languages.
However, back then machines had such wildly differing
wordsizes and character sets that this distinction was
necessary. After publication of the Revised Report, with its
included representation meant publication of programs,
but not for actual implementation on hardware, they
developed a `standard' hardware representation.
Even in that representation there were multiple representations
of the keywords in the language, e.g. and were allowed representations for the exact same keyword.
Implementors were strongly suggested to use the standard hardware representation, but if documented, they were fully
free to chose whichever representation they like. Hence, an
implementation may perfectly well use instead of as long as it is clearly documented.
I've carefully considered which phone to choose when
last upgrading. The balance between the 6310i and the
T610 was as follows:
Features they've both got:
- triband GSM
- bluetooth, gprs, wap
- good sound quality
- well organized menus
- nice shape
- no fiddly moving parts
- addressbook
- calendar
Features the 6310i is better at than the T610:
- reception
- power connector
Features the T610 is better at than the 6310i
- *will* sync with iSync, as opposed to the 6310i which will not
- higher resolution color screen
- built in toy-camera
- very much nicer ring
- smaller without getting unusable
Now, given that I live in Europe and we generally get good
reception *anyway*, I didn't really care about the reception
quality. I *do* care about the iSync compatibility though. As I carry my phone practically everywhere, that means I have my calendar with me at all times. I cannot get over how extremely useful this is. Oh, and dialling from the full
addressbook in my iBook has it's advantages as well. Nowadays I just keep the numbers of people that I actually
tend to call in my phone, making dialling from the phonebook quite a bit quicker.
If Nokia were to come out with a worthy successor to the 6310i that had roughly the same format but with a higher
resolution screen and SyncML capability... that phone would
absolutely rock and be my first choice. Unfortunately, the
T610 currently beats the 6310i on features... and it attracts more attention from women as well. (Actually, almost all other owners of T610's I've seen are women...).
Given the fact that the people working on formally proving the maths that go into Intel's new processors is correct have written a book on how to prove that the maths is correct, I would say, *yes*, it will be able to do Math correctly.
For more info, you might look at John Harrisson's pages
at http://www.cl.cam.ac.uk/users/jrh/ . He's at Intel
doing some of the work, and he wrote a PhD on how to
do the stuff he's doing for Intel nowadays.
Now, I may be just stupid, but I'd say the people at the
High Field Magnet Laboratory in Nijmegen have a much stronger claim
to world records... (33T continuous, 60T pulsed).
Where is the world record?
Interesting that this should trigger a YRO-like response,
where cows have had to wear a unique ID printed/burned on a yellow eartag, one for each ear, for years already (at least in the Netherlands, but I've seem them in
Germany as well).
Doing identification by retinal scanning seems *much*
more humane to me than identification by plastic yellow
flaps shot through the ears. It'd take us back to the times
when identification was mostly done based on the cow's
hide patterns, but without the trickiness in identifying the
pattern if the hide has been damaged...
The fire did not start in the server rooms. What happened was that a fire started in one of the lecture rooms (and a smallish one at that, number A108) that just happened to be in the same wing of the TWRC building that also housed the server rooms (yes, multiple). It then proceeded to take out two entire wings and threaten other buildings nearby as well as the library.
Now, I would *love* to see a halon system capable of stopping that...
Owh, and the fire seems to be under control by now, as evidenced on http://webcam.traserv.com/
(which you can contrast to http://www-infstud.sci.kun.nl/~arthurvl/ispy.jpg (taken at about 09:40 CET this morning)).
Because an extra battery on a 14.1" iBook gives you merely 55 Wh extra power, whereas this gives you 130Wh extra, while also enabling you to not switch batteries. So, the difference is between 110 Wh in total with a batteryswap (and hence a hibernate or shutdown) in the middle or 185 Wh without the need to swap batteries. I know what I would choose.
(*drool* 185 Wh... that amounts to about 18 hours of computing off the grid!)
The simplest solution is to go read-only for all system data, such as binaries and static configuration. Even better: use something of a commit-system to commit configuration changes to disk and have the disk be read-write only when committing changes.
Even though you'll still run with fsck this will not be a problem, as stuff can't have changed for reasons other than hardware failure... and you're not going to work around that in software anyway.
As with all activities: use your mind. What you want to be doing is exercise your muscles. While this can be done without any extra weights, most fit people will only really benefit with extra weights. A good set of dumbbells shouldn't set you back more than some EUR 60 (or $ 60).
Now, given that, you want to try to target particular muscles. Learn the moves that exercise the biceps, triceps, deltoids etc. Step on and off chairs to train your legs. Do pushups. Do situps. Stretch.
The problem with all this is discipline: it really doesn't help if you don't at least do this for about 20 minutes every day or 40 minutes every other day.
Ofcourse, you could also go for ballroom dance, historic dancing, indoor climbing, acrobatics, juggling...
If you are not too concerned about processorperformance, you might want to look at Soekris Technologies' Net4501 or Net4521. It is not very expandable, but seems to fulfill your requirements nicely. More information is at Soekris' website. Other options would be the PC104 series of modules... but you'd have to find your own enclosures.
How very interesting to note once again that Americans tend to think that everything important in computers has been American. For instance, I don't see the Electrologica machines, the X1 and X8, mentioned anywhere, even though they were the first to incorporate interrupts. Oh well. :)
Oh, some more information is at the University of Amsterdam's Computer Museum's Electrologica X1 and X8 site.
Anyway, his implementation has the number of characters the sed-implementation allows minus three as a limit on the total number of states in the Turing machine, which mine doesn't.
Oh, ofcourse, the sed Turing Machine is on the web as well. :)
Owh, by the way, it's pretty readable sed code... it's had to be for me to finish it. :)
Arthur
It is very interesting to see the fairly old technology of Pie menus implemented yet again. Somehow, one would think that there should be a lot more implementations,
given that pie menus show up again and again...
> Hackers? In the land of legal weed and prostitution? The Dutch better be preparred.
:)
Actually, weed is quite illegal in the Netherlands. However, it is officially tolerated.
Prostitution is legal, though, yes, so if that gets you going, please come.
Note that Dutch law is *very* strict on cracking however. It has been forbidden to crack accounts on computers for some seven years already, and people *have* been arrested for such acts. Besides, the Acceptable Use Policy is pretty clear on what is and is not accepted behaviour (basically: be nice to the network or the network crew won't be to you). Yes, the Dutch *are* well prepared.
First a bit of background: I'm 25 and therefore technically one generation older than you are. However, my little sister is 15, so technically in your generation. My little sis and I get along splendidly, but then again, she always had two older brothers to hang around with. She also gets along fine with people her own age, and is really very much a 15-year old.
As I am the oldest in my family, when I was fifteen I had no such 'luck'. I have always been youngest in class, and always spent time with people at least roughly a year older than I was. Therefore I was always striving to keep up, and usually succeeding... but for the social aspects of life.
Anyway, I was wondering, what are the ages of the people you regularly hang out with?
And what do you think of people that are your age?
One terabyte, for your information, is *not* 1024 gigabyte. If you had taken a look at the definitions of the prefixes in the systeme international, you would have known that tera is the prefix for 10^12, not for
2^40, as you seem to imply here. The systeme international actually *does* have prefixes for binary powers. 2^40 bytes would be a tebibyte, using those prefixes.
As not such an old fart that has been "programming" since he was nine,
I must say that at the age of 15, when I tried to learn OOP, I had much difficulty grasping the concepts. This may have had to do with the very limited set of articles the library had on it, but I really do think I wasn't able to comprehend it at that time.
On a side note, a lot of the concepts in programming are inherently hard. The concept of a variable is definitely not trivial. In the past year I've been teaching a group of 6 12 and 13 year olds to program. We started out teaching them recursive refinement (which is really easy and they really grasped quite well). Ofcourse, next you have to introduce variables. This is HARD. It took us wel over a month to get the kids to internalize the concept of a variable. And even now they occasionaly miswrite assignments, with the destination and source of the assignment swapped.
For those of you that might think teaching OO before teaching what a variable is: DON'T! We've tried that at the university with second year chemistry students. They really had trouble following the course, until we reclarified what a variable is.
In the Netherlands trade secret protection does not hold if the group of people the secret is distributed to is too large. This holds *even* if each and every person in that group has vowed to keep it secret. The point is that a large enough group of people in and of itself is a `public', so distributing a trade secret to a large enough group of people implies publishing the trade secret and therefore lifting trade secret status. :)
Neat, huh?
Copyright nor patents preclude free distribution. However, currently there are problems applying both copyright and patents to new developments. With the extremely small division between an idea and it's implementation in software and companies patenting as broad as they possibly can we see the patent system breaking down not because of inherent flaws in it, but merely because of insufficient care exercised in determining which patents are valid and which are not.
I think the point I'm trying to make is to let people realize that crying "the system is breaking down, therefore it must be fundamentally wrong" is pretty stupid. If my PC were to run Amazon, it would break down pretty quickly... but that does not give me any reason to think Amazon's systems are built on fundamentally different concepts than my PC is...
I do agree that the current system is too expensive to keep on track, and that would be a good reason to redesign it. That does not mean it's design is inherently wrong...
Actually, if you were to compile C code to the 128 bit VLIW of the Crusoe, you *could* be benefiting from the runtime aliasing, as you would also have access to the ldp (load-and-protect) and stam (store-under-alias-mask) instructions. The problem however is in the fact that Transmeta reserves the right to change the instructionset underlying the CodeMorphing software with each and every revision of the processor. And that would mean rewriting the compiler...
As the article states, the UCITA can, given certain circumstances, be applied to the sale of computers. Now, I could easily see Microsoft[1] buying up hardware manufacturers and sell the computers with a license clause effectively stating that one cannot run anything but Microsoft software on the hardware.
Wouldn't this be *terrible* for open source?
[1] Microsoft is just taken as shorthand for `big evil software empire', and you can supply your own variant any time.
You are of course completely right in saying the measurement is part of the science involved, and without measurement any attempts at improvement are based on guesses at best. However, the problem with most management methods is that they do not put the measurement tools in to the work flow of the workers but only graft them on to the side.
In that case updating the measurements *is* auxiliary to the actual job at hand, and will not be done.
As an example: an editorial board can benefit greatly from knowing how long diffent stages of editing an article take, and how well a certain editor performs at each stage. Maintaining this information is a major hassle, however, as it has nothing to do whatsoever with editing articles. But, if one incorporates maintaining this information in the meta-work of finding the next article to work on, it suddenly is incorporated in the normal work to do, and the information will be kept up to date.
The court *did* rule however that ISP's *would* commit copyright infringement if a page on their machines contains a link to a copyrighted page *and* the owner of the copyrights on said page made it clear to the provider that it considered the link to be a copyright infringement *and* the ISP will not remove the link.
Basically, said ruling makes either the ISP or the copyright owners judges insofar that *they* must now decide what does and does not constitute copyright infringement. And that is *really* scary stuff.
Now this can be extrapolated to mean that *any* company or person that complains to a service provider about a link to works that company or person has copyrights for can force the service provider to remove the link under penalty of law... I'd personally love to see (say) Universal take on (say) Altavista like this...
(Oh yeah... CST is Scientology... ;))