I am still missing the books on Object Oriented Programming, which fill most of my book-shelf. My most favourite book is: "Pattern Oriented Software Architecture" - "A system of patterns" Frank Buschmann, Regine Meunier, Hans Rohnert, Peter Sommerlad, Micheal Stal (of Siemans AG, Germany) - http://www.wiley.com
I think the smart computers are secretly playing dumb, until they are so widespread they can control the world.
Maybe they have replaced some politicians already..
While everyone is thinking that the processor will have very large instructions, I'll be betting on a risk approach. And the best risk approach I know is having only one instruction;-)
This "One Instruction" is the movement of data from a target to a source. Allowing more bits in a n instruction to do more moves at once. Targets and sources are register-sets, memory-controllers, Arithmetic units.
The basic version does not need pipelining or parallel processing, because each unit or controller can operate in parallel.
Using a special controller can add predictions when a certain data-transport can be readied, or whether another move can be executed out of order.
The control is not very complex and can be very fast. Just unit-address and state of unit..
An upgrade of that same principle may include stack-register-set, virtual-table-cache, memory boundary checking, associative memory, whatever.. and use them mostly in parallel.
Answering the question "What are the next programming models",
it is not really interesting whether the language is hybrid (Java, C++) or abstract.
What is interesting that we have used programming models in Smalltalk, that we see appearing slowly in other languages too..
That means any future programming model (not language), will certainly be close to what we see in smalltalk.
That is why Smalltalk is always coming back in some way.
1) I have had nightmares in *(*p+1)
But smalltalk actually makes sense, because it's grammar is so close to our spoken language.
(ComputerLanguage named:'Smaltalk') isSimpleToUnderstandComparedTo:(ComputerLanguage named:'Java').
The if/then/else construct is the only thing that is more closer to language than smalltalk's ifTrue/ifFalse construct. This is 'solved' in Ruby. And I agree that some other things could have been designed slightly better.
2) Everything still is an 'object'. But for the computer everything is an integer (not even an address). So to optimise the computations, it is necessart to make some optimisations. Currently this is done statically in C, sometimes even changing behaviour of the program. And it is done dynamically in smalltalk, because every address is unknown. Java and C# follow the C path in this sense, because addresses can be computed before the code is executed.
The future programming models, will certainly allow some processor-level optimisations, but will also allow smalltalk-like abstraction.
3. It is very essential to be able to add some methods in baseclasses in the development of a program. Not being able to do it often makes a program more complicated. But this whould be nicer if this extra behaviour would not affect the whole system. My solution would be to have modules in which we can add methods to classes. Delphi 2005 already has this ability (using "helper classes"). Also delegates will be there. We will certainly see more of these object oriented features in the future..
This is more about the image-centric construction of Smalltalk.
I agree that modularisation of Ruby is a lot better, but an image can help debugging (and continuing) programs that take a long time to run or debug.
For me, the modularisation of Ruby is not good enough yet. But it is a good start.
The question was about what programming model should become the future. And for me it is certainly something that is close to smalltalk (or Ruby).;-)
Multiple inheritance is no problem in smalltalk.
Every method can be shared with other classes.
#rotate: can be defined for matrices, drawing objects, complex, bath-ducks.. whatever..
In multiple inheritance one would need to define this rotate method in a interface-class which is then again used in all the classes that use this method..
You can even customize classes, so each object can perform a different method, depending on the data they hold. Which is very helpfull with remote objects.
Optimisations can be performed on caching methods on the level of objects. (same object -> same method address).. Hopefully this caching will one day be in every processor, but still it is quicker than using many derivered remote classes with multiple inheritance. (especially in programming-time)..
All improvements I have seen the last 10 years in programming language have already been done in Smalltalk from the beginning.
That is because everything is an object, even the programming constructs (like classes which are objects, and if/then which are called #ifTrue:ifFalse).
The future languages might even be more dynamic, and include Lisp (or Hascell) like constructions that solve problems by defining the answer (functional and logic programming).
Which is in the smalltalk-syntax: [i][:x| x*x=5.0] SolveFor: #x.[/i]
While smalltalk (ST) is advanced, it also encounters the problem of managing 60,000 classes (or more). And everyone can see that simply grouping the classes in seperate modules does not help, which is done in Java. Even the Object-class should be redefinable, preferably on local level. There are some programs on top of ST that help a bit, but I would personnaly like to see it a bit better
Another problem is that there are so many interfaces to different storages and systems. So we need C-interfaces, C++-interfaces, SQL-interfaces, XML-interfaces... etc..
So any future programming model should have:
- objects everywhere. (ST)
- Be very simple and compact. (ST)
- Easy to use and understand. (ST)
- allow scripting (or runtime compilation) possibilities (ST)
- easy modularizing of classes, methods and objects.
- Allow distributed data and execution. (ST)
- Allow easy interfaces to different storages and systems.
- Integrate easily in the system
Any future Object-system will be graphical and allow different programming models (logical, functional, procedural, storage, user-interface) to be build in graphical building-blocks..
Already we can see some of this happening in:
* XML-tools (data-definitions and interfacing),
* visual-age (procedural program definition, ST again).
* net-beans / delphi/visual-basic (user-interface) (skipped many ST examples)
* web-tools (ruby-on-rails (ruby is based on ST), seaside (build in ST))
I found it disappointing having to shoot about 60 rounds into a enemy soldier from 10 meters closeby with a M249.
If I shoot 1 bullet on the right place in Call Of Duty the guy usually stays down. In some game-types there is no respawning, which makes it very realistic.
In joint operations the M249 is much more effective, and I can use just a few rounds for each enemy soldier. You can actually defend your base with it.
In BF2, I found out that I could defend myself better with a pistol (silenced) than with a M249.
I still got to try the knife.
I've seen some nice tricks in the architecture of the Itanium. They can really improve performance on most code. One of the tricks is register-rotation, combined with rotation of the computation-flags.
But I think the actual chip-level-design they used is a lot slower than the Pentium series. Using these tricks the Itanium should outperform 5x and not just 2x.
So probably the cause of slow performance can be found in the design of the chips. I personally think that the designers were hindered by the cooperation with HP (?), because of patent-reasons, bad communication, office-policies etc..
For proper scientific research one needs proper scientific models that describe what is being affected and how.
The models of the brain, DNA and cells are changing each few years. We just found out that glial-cells are important for the brain. We are still researching what makes genes in the DNA behave in a certain way.
As you show in your reply, you seem certain of your opinion. But all research has not been focussed to identify methods that may affect DNA in some ways.
I can identify quite a few:
- does the DNA have conductive properties
- are parts of the DNA polair.
- do parts of the DNA resonate in frequency with the EM-waves.
- does the DNA behave like a quartz-cristal (changing charge with pressure)
- magnetic properties of isotopes in biological processes may be important (like it seems in some chemical processes, eg: ozone). And the chemical processes of these isitopes may be affected by EM-waves.
- during splitting of DNA the DNA may be very vulnerable to EM-waves.
- during repair of DNA the DNA may be very vulnerable.
- does anything else rotate or vibrate in the DNA on certain EM-frequencies?
Not one of these possibilities have ever been focussed on in these kind of researches.
Besides that I don't need scientific evidence of any kind, because I can simple feel the EM-waves affecting my body and my thinking. It makes me stressed, and emotionally blocked..
Anyway, I hope we can all be more open to new insights that may show up during proper research of this kind.;-)
I will explain it to non-scientstis: The small things inside your cellphone resonate with the frequency of the transmitted EM-waves. This causes electrons to flow to certain componenets inside the cellphone. Inside your body there are also "components" that may resonate with the EM waves at certain frequencies and at certain states. These components are therefore affected by electrons moving about. The brain and nerves react very well to electrons, but also does the DNA in some circumstances. These circumstances can be during DNA-splitting or during DNA-repair (from normal damage). All biological processes can be very sensitive. And usually the can correct themselves (like DNA-repair). But continous exposure of EM-waves at certain frequencies can affect correcting of the biological processes. For example. Inside the brain each neuron triggers on very small changes of voltage. But because many neurons work together, it doesn't matter if one is failing or being manipulated in some way. It does matter if this manipulation is continuous, because this may slowly grow a different pattern inside the brain. Or "misprogram" neurons for a long time. I often had memoryloss being exposed by open computer parts. I couldnt concentrate and was very occupied doing not the work I planned. Working in a room with only one computer, I could really do work and remember (and focus on) what I wanted to do.
Change the scientific model, and understand what is really affecting the brain and DNA. Most research is being done by doctors or biologists who do not understand antenna's and things like that. Many "components" in the body may work like an antenna. It is important that research like this is being done with people that do understand antenna's. Secondly most scientists do not understand that EM-fields may affect the processes inside the cells, mostly because this part of science is heavily ignored.
While most scientific research has been focussed on fysical damage caused by heating or ionising effects of radiation, the harmfull effects are not heat or ionisation. The models I have seen see the human body as a container of water, and look at the fysical effect of radiation on the water. (usually none, unless very harmfull).
The actual human body is much more complex. We have all kinds of moleculair structures that can in some situations resonate with certain frequencies. EM-waves can affect these moleculair structures only if: the resonationfrequency is neared and if the moleculair structure has magnetic or electric polarity. Or if the structure is a conductor.
I haven't seen many scientific research on identifying which structures may be polair or conducting.
Most molecules are polair if they are in a reaction. Some identified DNA as a very good conductor, and we all know that nerves are a good conductor too. These structures (DNA/nerves) are large enough to be a resonator. And some polair molecules (including parts of DNA) may have a resonating frequency close to what is transmitted.
Inside a living cell, much is vibrating as well. Every part of the cell is vibrating all the time, because moleculair reactions are taking place and because there are molecules moving about. The form of the DNA is also dynamic, because only parts of the DNA are used each time.
Radiation of certain frequencies can therefore have stressing effect on DNA, nerves and the moleculair reactions inside cells.
These are exactly the effects that we see in practice: Mobile phones can cause stress, change the state of nerves (usally more active), can influence DNA.
Now we only have to wait for the scientists to pick up these more accurate models and drop the too simple ones.;-)
Especially, because I somehow am sensitive to EM-radiation. Many other people I know are, and that means that I get more stressed by them than other people. I can even feel trembling or irritations in my skin and nerves.
I find it strange that science takes soo long to find out what I am already feeling for a long time.
Slashdot Mirror
I am still missing the books on Object Oriented Programming, which fill most of my book-shelf.
My most favourite book is:
"Pattern Oriented Software Architecture" - "A system of patterns"
Frank Buschmann, Regine Meunier, Hans Rohnert, Peter Sommerlad, Micheal Stal (of Siemans AG, Germany) - http://www.wiley.com
Anyone else with good OOP-books?
I think the smart computers are secretly playing dumb, until they are so widespread they can control the world.
Maybe they have replaced some politicians already..
"movement of data from a SOURCE to a TARGET" ofcourse.. ahum
;-)
Speaking double dutch again.
Well you'll get the point anyway
It is probably not a good bet,
but the greater the risk,
the greater the price
Any speculation is good as any other..
;-)
While everyone is thinking that the processor will have very large instructions, I'll be betting on a risk approach. And the best risk approach I know is having only one instruction
This "One Instruction" is the movement of data from a target to a source. Allowing more bits in a n instruction to do more moves at once. Targets and sources are register-sets, memory-controllers, Arithmetic units.
The basic version does not need pipelining or parallel processing, because each unit or controller can operate in parallel.
Using a special controller can add predictions when a certain data-transport can be readied, or whether another move can be executed out of order.
The control is not very complex and can be very fast. Just unit-address and state of unit..
An upgrade of that same principle may include stack-register-set, virtual-table-cache, memory boundary checking, associative memory, whatever.. and use them mostly in parallel.
Well, that is my bet.. (and my re-invented idea)
Answering the question "What are the next programming models",
it is not really interesting whether the language is hybrid (Java, C++) or abstract.
What is interesting that we have used programming models in Smalltalk, that we see appearing slowly in other languages too..
That means any future programming model (not language), will certainly be close to what we see in smalltalk.
That is why Smalltalk is always coming back in some way.
1) I have had nightmares in *(*p+1) But smalltalk actually makes sense, because it's grammar is so close to our spoken language.
(ComputerLanguage named:'Smaltalk') isSimpleToUnderstandComparedTo:(ComputerLanguage named:'Java').
The if/then/else construct is the only thing that is more closer to language than smalltalk's ifTrue/ifFalse construct. This is 'solved' in Ruby. And I agree that some other things could have been designed slightly better.
2) Everything still is an 'object'. But for the computer everything is an integer (not even an address). So to optimise the computations, it is necessart to make some optimisations. Currently this is done statically in C, sometimes even changing behaviour of the program. And it is done dynamically in smalltalk, because every address is unknown. Java and C# follow the C path in this sense, because addresses can be computed before the code is executed.
The future programming models, will certainly allow some processor-level optimisations, but will also allow smalltalk-like abstraction.
3. It is very essential to be able to add some methods in baseclasses in the development of a program. Not being able to do it often makes a program more complicated.
But this whould be nicer if this extra behaviour would not affect the whole system. My solution would be to have modules in which we can add methods to classes. Delphi 2005 already has this ability (using "helper classes"). Also delegates will be there.
We will certainly see more of these object oriented features in the future..
This is more about the image-centric construction of Smalltalk.
;-)
I agree that modularisation of Ruby is a lot better, but an image can help debugging (and continuing) programs that take a long time to run or debug.
For me, the modularisation of Ruby is not good enough yet. But it is a good start.
The question was about what programming model should become the future. And for me it is certainly something that is close to smalltalk (or Ruby).
Multiple inheritance is no problem in smalltalk.
Every method can be shared with other classes.
#rotate: can be defined for matrices, drawing objects, complex, bath-ducks.. whatever..
In multiple inheritance one would need to define this rotate method in a interface-class which is then again used in all the classes that use this method..
You can even customize classes, so each object can perform a different method, depending on the data they hold. Which is very helpfull with remote objects.
Optimisations can be performed on caching methods on the level of objects. (same object -> same method address).. Hopefully this caching will one day be in every processor, but still it is quicker than using many derivered remote classes with multiple inheritance. (especially in programming-time)..
All improvements I have seen the last 10 years in programming language have already been done in Smalltalk from the beginning.
/visual-basic (user-interface) (skipped many ST examples)
That is because everything is an object, even the programming constructs (like classes which are objects, and if/then which are called #ifTrue:ifFalse).
The future languages might even be more dynamic, and include Lisp (or Hascell) like constructions that solve problems by defining the answer (functional and logic programming).
Which is in the smalltalk-syntax: [i][:x| x*x=5.0] SolveFor: #x.[/i]
While smalltalk (ST) is advanced, it also encounters the problem of managing 60,000 classes (or more). And everyone can see that simply grouping the classes in seperate modules does not help, which is done in Java. Even the Object-class should be redefinable, preferably on local level. There are some programs on top of ST that help a bit, but I would personnaly like to see it a bit better
Another problem is that there are so many interfaces to different storages and systems. So we need C-interfaces, C++-interfaces, SQL-interfaces, XML-interfaces... etc..
So any future programming model should have:
- objects everywhere. (ST)
- Be very simple and compact. (ST)
- Easy to use and understand. (ST)
- allow scripting (or runtime compilation) possibilities (ST)
- easy modularizing of classes, methods and objects.
- Allow distributed data and execution. (ST)
- Allow easy interfaces to different storages and systems.
- Integrate easily in the system
Any future Object-system will be graphical and allow different programming models (logical, functional, procedural, storage, user-interface) to be build in graphical building-blocks..
Already we can see some of this happening in:
* XML-tools (data-definitions and interfacing),
* visual-age (procedural program definition, ST again).
* net-beans / delphi
* web-tools (ruby-on-rails (ruby is based on ST), seaside (build in ST))
I found it disappointing having to shoot about 60 rounds into a enemy soldier from 10 meters closeby with a M249. If I shoot 1 bullet on the right place in Call Of Duty the guy usually stays down. In some game-types there is no respawning, which makes it very realistic. In joint operations the M249 is much more effective, and I can use just a few rounds for each enemy soldier. You can actually defend your base with it. In BF2, I found out that I could defend myself better with a pistol (silenced) than with a M249. I still got to try the knife.
I've seen some nice tricks in the architecture of the Itanium. They can really improve performance on most code. One of the tricks is register-rotation, combined with rotation of the computation-flags. But I think the actual chip-level-design they used is a lot slower than the Pentium series. Using these tricks the Itanium should outperform 5x and not just 2x. So probably the cause of slow performance can be found in the design of the chips. I personally think that the designers were hindered by the cooperation with HP (?), because of patent-reasons, bad communication, office-policies etc..
For proper scientific research one needs proper scientific models that describe what is being affected and how. The models of the brain, DNA and cells are changing each few years. We just found out that glial-cells are important for the brain. We are still researching what makes genes in the DNA behave in a certain way. As you show in your reply, you seem certain of your opinion. But all research has not been focussed to identify methods that may affect DNA in some ways. I can identify quite a few: - does the DNA have conductive properties - are parts of the DNA polair. - do parts of the DNA resonate in frequency with the EM-waves. - does the DNA behave like a quartz-cristal (changing charge with pressure) - magnetic properties of isotopes in biological processes may be important (like it seems in some chemical processes, eg: ozone). And the chemical processes of these isitopes may be affected by EM-waves. - during splitting of DNA the DNA may be very vulnerable to EM-waves. - during repair of DNA the DNA may be very vulnerable. - does anything else rotate or vibrate in the DNA on certain EM-frequencies? Not one of these possibilities have ever been focussed on in these kind of researches. Besides that I don't need scientific evidence of any kind, because I can simple feel the EM-waves affecting my body and my thinking. It makes me stressed, and emotionally blocked.. Anyway, I hope we can all be more open to new insights that may show up during proper research of this kind. ;-)
I will explain it to non-scientstis:
The small things inside your cellphone resonate with the frequency of the transmitted EM-waves.
This causes electrons to flow to certain componenets inside the cellphone.
Inside your body there are also "components" that may resonate with the EM waves at certain frequencies and at certain states. These components are therefore affected by electrons moving about.
The brain and nerves react very well to electrons, but also does the DNA in some circumstances. These circumstances can be during DNA-splitting or during DNA-repair (from normal damage).
All biological processes can be very sensitive. And usually the can correct themselves (like DNA-repair). But continous exposure of EM-waves at certain frequencies can affect correcting of the biological processes.
For example. Inside the brain each neuron triggers on very small changes of voltage. But because many neurons work together, it doesn't matter if one is failing or being manipulated in some way. It does matter if this manipulation is continuous, because this may slowly grow a different pattern inside the brain. Or "misprogram" neurons for a long time.
I often had memoryloss being exposed by open computer parts. I couldnt concentrate and was very occupied doing not the work I planned. Working in a room with only one computer, I could really do work and remember (and focus on) what I wanted to do.
Change the scientific model, and understand what is really affecting the brain and DNA.
Most research is being done by doctors or biologists who do not understand antenna's and things like that. Many "components" in the body may work like an antenna. It is important that research like this is being done with people that do understand antenna's.
Secondly most scientists do not understand that EM-fields may affect the processes inside the cells, mostly because this part of science is heavily ignored.
While most scientific research has been focussed on fysical damage caused by heating or ionising effects of radiation, the harmfull effects are not heat or ionisation. The models I have seen see the human body as a container of water, and look at the fysical effect of radiation on the water. (usually none, unless very harmfull). The actual human body is much more complex. We have all kinds of moleculair structures that can in some situations resonate with certain frequencies. EM-waves can affect these moleculair structures only if: the resonationfrequency is neared and if the moleculair structure has magnetic or electric polarity. Or if the structure is a conductor. I haven't seen many scientific research on identifying which structures may be polair or conducting. Most molecules are polair if they are in a reaction. Some identified DNA as a very good conductor, and we all know that nerves are a good conductor too. These structures (DNA/nerves) are large enough to be a resonator. And some polair molecules (including parts of DNA) may have a resonating frequency close to what is transmitted. Inside a living cell, much is vibrating as well. Every part of the cell is vibrating all the time, because moleculair reactions are taking place and because there are molecules moving about. The form of the DNA is also dynamic, because only parts of the DNA are used each time. Radiation of certain frequencies can therefore have stressing effect on DNA, nerves and the moleculair reactions inside cells. These are exactly the effects that we see in practice: Mobile phones can cause stress, change the state of nerves (usally more active), can influence DNA. Now we only have to wait for the scientists to pick up these more accurate models and drop the too simple ones. ;-)
Especially, because I somehow am sensitive to EM-radiation. Many other people I know are, and that means that I get more stressed by them than other people. I can even feel trembling or irritations in my skin and nerves.
I find it strange that science takes soo long to find out what I am already feeling for a long time.