first it's not the virus that is doing anything. it's just a scaffold. the virus just self-assembles the scaffold for you. the interior DNA / RNA is irrelevant.
that said, the design for the self assembly and display is in the virus DNA I believe. so given a host to express itself in, it could presumably reproduce this in the wild. it would not be any use to the virus. But you could imagine that some host cell might harness the virus to make hydrogen for it's own purposes.
So I suspect that if this gets loose in the wild that the virus won't keep this trait long enough for some host cell to adapt to taking advantage of it.
Basically, hogwash. Having one language and one tool chain leads to fewer methods of being able to solve problems, and reduces the utility of your computer system.
that is an absurd assertion. Look at the success of the JVM, and.Net and the CLI. look at the GNU (multi-language) compiler. These are considered marvelous innovations. look at how linker's work to allow you to call from one language into another.
What all those have in common is defining an interface layer that is standardizes things.
But those are high level. if you want to start thinking about the future of fully automated multi-processing you have to move the interface layer down to find the re-orderable sections of code.
The rest of your assertions have no techincal merit.
Does a single thing you said explain why they won't allow frameworks which compile to Objective-C?
That's where I'm having trouble. I can see a technical reason to force people to use a single language, or at least a single runtime representation, in the same way that, say, the new Windows Mobile forces you to use.NET. I don't see any technical reason for them to care what language you originally use to produce it.
I dont' have a complete answer but I can grunt out a few ideas on this.
1)
Consider the problem of python's global interpreter lock. it prevents parallel execution of python: most people are suprised when they discover that a multi-threaded python application excecutes only one thread at a time and so is usually slower than a single threaded python program. (yes there are exceptions, and yes there are non-standard native libs to fix this problem, but let's not get off topic).
If you convert a python code to C, this non-parallizable global interprer lock does not go away!
What this example shows is that even if you output something into objective-c or whatever, you can create programs that impose huge constraints on the way a code must be executed.
2) consider what happens if apple implements new features. e.g. suppose they built something like openMP into objective C. What happens if the abstraction that converts one language to objective C does not use this feature. For the case I just mentioned, openMP, I could imagine that it would be fairly difficult to convert the openMP processor directives in a machine translated way that would still make sense or work. TO be specific when you convert languages you tend to replace native commands like for x = 1..3:
y = 3 + x
There will be a lot of ink spilled here condemning apple, or explaining why apple is "alllowed" to do this. (i.e. blink! don't blink!) There will also people explaining why this a good bussiness move or good for developers or good for consumers.
So this post is here to start a thread on technical speculation on why apple might be doing this from a point of view of computer science and design strategy. For example, we recently witnesses a more nuanced form of multi-tasking that divides the use cases into ones that give better resource management hints to the OS than the carte blance form. likewise we are seeing memory management for apps being sandboxed and encrypted. clearly these are all about evolving the OS to better control it's resources.
Now what might be the resource use case improvement here?
I'll start the speculation with this thought and leave it to others to fill in more. Suppose that if you were thinking ahead to things like ubiquitous multi-processing and beeing able to carbon freeze applications for instant re-starts. It sure would be nice to have the operating system be able to make code elements re-entrant or find the finest grained atomic segements or know what non-threaded subroutines could actually be done threaded and out of order without the programmers actually having to think about this when they wrote code.
One way you could approach this is to use a single compiler or at least a single memory model and calling convention. then either the compiler itself could include hinting, or maybe the OS could even figure it out after the fact.
I note you'd not be able to do this if for example, the original code was coverted direct to assembly without obeying certain rules. (e.g. just imagine self modifying code).
I note that Objective-C is more than pure syntactic sugar on C. unlike say c++ it does not use a compile-time generaged look up table for method resolution. instead objects get messages sent to a standard receiver in the object and this receiver resolves the method at run time. (it does not even have to respond to the message!). Objective-C also bases it's function argument prototype based polymorphism in a much more strict way than C++ so there's in prinicple more information content about what is being passed. I don't want to emphasize any one thing too much. perhaps something I said is technically wrong. My point is that if you use a consistent model that takes advantage of the extra abstraction layers in how function calls occur and how memory is used then potentionally one can do finer grained parallelism automagically.
likewise, carbon freezing applications for instant restarts may become more stable as well if you have better control over what is mutable and what is shared between threads.
just a thought that this could be a purely technical goal to move operating systems and code development into the massively parallel multi-core future.
if you've followed the Grand central innovation then you can see apple is indeed thinking about making dispatch more hintable by developers.
For decades apple users have had to "enjoy" ported games. Games written either for come arcane quirks of antoher platform then ports half assed to apple or games written in some generic way. They don't support accelerations or available stuff. The linux folks know the feeling too of some round direct-X game shoved into a X11 hole.
Apple is saying now that we have a platform advantage it's going to be windows that gets the hand-me-down games. Games targeted to agnostic APIs are banned too cause it just means people are writting for the lowest common denominator.
THis is actually not entirely new behaviour. Apple always aims for highly spec-ed platfomrs-- never a stripped low end one-- and thus developers have always been able to assume a feature rich platform. for example, in the long ago bad old days apple's always had full stereo sound out when PCs had all sorts of different levels of sound outputs -- usually just a beep speaker. One an cite many such examples.
thus apple wants developers to target it's full capabilities not use generics.
I'm puzzled over the "per dive" part of this. this seems to imply it's that downward dive itself that is some how producing the energy. that is to say this energy is not continuously produced but would require another dive cycle to produce. So it' can't stay down. perhaps it can cycles dives autonomously?
And how is this power produced. I'm going to guess what is happening is that it comes up and warms up till it's core temperature is at the ambient surface temperature. Then it drops like a rock, and uses the heat differential between the core and the cold water to drive some thermo electric engine in reverse. perhaps they toss in some phase change material to extend the thermal capacity.
Hmmm..... So if I had to guess then the symmetry is an artifact of the bounding box. That's curious. You'd think it would have residual asymmetry from the sun trajectory. I suspect there is some funny assumptions being imposed directly or indirectly. e.g. perhaps the layoput of the alelles in the genetic algorithms itself slightly favors self similar patterns
Looking at the slides more carefully, I think there's some substantially strange assumptions being made. Notice that he starts from completely random and non-symmetric shapes and these are evolved in genetic algorithms. the results he shows are all highly symmetric. some have 3 C4 rotation axes.
this makes no sense to me. the suns seasonal variation and arcs do not illuminate the ground symmetrically. So it is hard to see why it would evolve to a symmetric structure.
so there have to be some assumptions here the article is not exposing. like enforcement of symmetry.
It's an interesting, nerdy endeavor, but less practical than automated tracking systems; the expensive part of solar is the panels themselves. From TFA: His new designs are up to two and a half times more efficient per comparative length and width than traditional flat arrays.
If solar cells were free, than this would indeed be more efficient, and if there's limited space thay MAY be more practical.
Exactly!
consider the simplified case of sun's arc not moving with the seasons. then you could put down the panels in a 90 degree zig-zag. this way all light that is reflected is assured to strike a second panel. this would dramatically increase the efficiency and reduce the variation throughout the day. but it would take 1.4 times as much panels to cover the same area as a flat panel. if you go for the 3D full corner cube then it's 1.7 times as much.
If you were to spread this out you would have 40% more area. this would mean that at peak power you'd get 1.4 times as much, but at obtuse incidence angles were the reflection is high you'd take a loss. the trade off point is when the reflective loss is greater than 40% I think.
another problem with a highly faceted desing is going to be in making the nominally circular cells conform to odd shaped facets, and for mass producing these. If you look at conventional panels you see they cut they often circular cells into half-circles then put these down in a row laternating the directions. this allows them to make mass producable long sections that dont have as much dead space when the components are placed side by side. If you have facets of differenting shapes you have to make eachone differently and the chips may have to be cut differently.
The best part of this idea is the continuous power level however.
Re:Revisionist history
on
The Apple Two
·
· Score: 2, Interesting
Get your history straight. You didn't have to buy an original Apple ][ with case, keyboard, or power supply. You could simply purchase the motherboard (board on stick) and provide your own keyboard, case, and huge, hot, inefficient non-switching power supply. Such setups were never common because most people didn't want them.
Which I think is the point I was making, no?
The video management was all done in hardware. You're probably thinking of the Sinclair zx80/81 or maybe one of Don Lancaster's TV typewriter designs.
While Don lancaster's design took things to extreme, lots of the video management was done in software on the apple. for example, sprites. and at a more basic level, the colors were interleaved so you did not actually have the full resolution in color that you had in black and white. That translation of pixel position offset with color was handled in the device drivers as I recall.
The Apple ][ used an ASCII keyboard. No interesting software decoding here.
sure there was. how do you think the key presses were captured? they did not have any buffering.
I find it hard to believe you call the Apple ][ locked down when the schematics and firmware listings were in the back of every single user manual!
I think you missed my point. I was being ironic. the apple ][ was being heralded as being open. And by modern standards it is. but for the standards of it's time it was highly integrated and less flexible.
But my larger point is, to move forward we have to periodically draw a black box around some bit of technology and not worry about home brewing that part. You would not I expect want to build your own SATA replacement from scratch in algorithms and hardware (resitors and capacitors). You might want access at different levels of balck boxing. 1) a mac mini, where you don't even care if it's SATA or IDE 2) mutherboard with integrated SATA from Frys. 3) a sata controller card to plug in to a motherboard. but none of those dives very deep. it's closed at the SATA layer and that's a good thing if you want to get something done.
the Ipad is closed at the hardware layer. basically all you can do is write software or build something for it's two connector ports.
but anyhow the point was the exact same complaint was made about the apple ][ in it's day. It's only later we appreciated how freeing us from the details extended our capabilities to modify more complex aspects of the device.
Re:Meshworks, Hierarchies, and Interfaces
on
The Apple Two
·
· Score: 2, Funny
""" There are many natural scientists, and especially physicists, who continue to reject the notion that the disciplines concerned with social and cultural criticism can have anything to contribute, except perhaps peripherally, to their research. Still less are they receptive to the idea that the very foundations of their worldview must be revised or rebuilt in the light of such criticism. Rather, they cling to the dogma imposed by the long post-Enlightenment hegemony over the Western intellectual outlook, which can be summarized briefly as follows: that there exists an external world, whose properties are independent of any individual human being and indeed of humanity as a whole; that these properties are encoded in ``eternal'' physical laws; and that human beings can obtain reliable, albeit imperfect and tentative, knowledge of these laws by hewing to the ``objective'' procedures and epistemological strictures prescribed by the (so-called) scientific method.
But deep conceptual shifts within twentieth-century science have undermined this Cartesian-Newtonian metaphysics1; revisionist studies in the history and philosophy of science have cast further doubt on its credibility2; and, most recently, feminist and poststructuralist critiques have demystified the substantive content of mainstream Western scientific practice, revealing the ideology of domination concealed behind the façade of ``objectivity''.3 It has thus become increasingly apparent that physical ``reality'', no less than social ``reality'', is at bottom a social and linguistic construct; that scientific ``knowledge", far from being objective, reflects and encodes the dominant ideologies and power relations of the culture that produced it; that the truth claims of science are inherently theory-laden and self-referential; and consequently, that the discourse of the scientific community, for all its undeniable value, cannot assert a privileged epistemological status with respect to counter-hegemonic narratives emanating from dissident or marginalized communities. These themes can be traced, despite some differences of emphasis, in Aronowitz's analysis of the cultural fabric that produced quantum mechanics4; in Ross' discussion of oppositional discourses in post-quantum science5; in Irigaray's and Hayles' exegeses of gender encoding in fluid mechanics6; and in Harding's comprehensive critique of the gender ideology underlying the natural sciences in general and physics in particular.7
Here my aim is to carry these deep analyses one step farther, by taking account of recent developments in quantum gravity: the emerging branch of physics in which Heisenberg's quantum mechanics and Einstein's general relativity are at once synthesized and superseded. In quantum gravity, as we shall see, the space-time manifold ceases to exist as an objective physical reality; geometry becomes relational and contextual; and the foundational conceptual categories of prior science -- among them, existence itself -- become problematized and relativized. This conceptual revolution, I will argue, has profound implications for the content of a future postmodern and liberatory science. """
Revisionist history
on
The Apple Two
·
· Score: 5, Insightful
posting this again, since something went wrong the first time:
this is a false dichotomy forged by suggesting that this Steve is good, ergo this steve is bad, then amplifying those traits by mapping them on to perceived standards of today.
I built and sold homebrew computers in the era when the apple II hit the market. At the time we all laughed at the apple as a "toy" because it was so locked down and not built from components. Back then, sonny, you built a computer like an Imsai, altair, cromenco, by starting with a metal box, putting in a non-switiching power supply, choosing the largest capacitors you could fit in the box, then an s-100 (altair) buss. then you picked a cpu board from one manufacturer, some memory cards from another, a keyboard uart decoder from another, a keybaord from another, a video card, and a TV screen modded with an RF converter on channel 4.
These apples were hideously locked down. Switching powersupplies with just wires coming out of a metal box, no way to ugrade the capacity and very little excess capacity. the keyboard was integrated into the case ! and wholly shit a mother board with soldered in chips, video, meomery, and CPU.
Even the address space of the cards you plugged in was decoded on the motherboad not the cards (which allowed the cards to be smaller than the ones for the S-100 bus). THe cards even got regulated voltages not raw rectified AC.
they sucked all the flexibility out of it.
the software was essential to the operation of the hardware not separate from it: a lot of the video management was done in software. the timing one the disk drives they put out used soft sectors not hardware determined sectors (only one hole punched in the floppy instead of 20, one for each sector). Even the memory refresh was handeled on the video updates which in turn were backsided on last half of the 6502's instruction cycle (when it would not be fetching). It was one of the very first systems to successfully use dynamic memory. (Only a fool would not use static memory in an altair, since you had to do all the refresh handling on the memory card).
You had to buy apple floppy disks, and apple plug-in cards for many things cause they were not standard cards or drives.
And of course the apple II in hind sight was one of the most geniuous machines ever built. it's lock downs let hobbiest's soar in other directions. plug in cards were small and the pre-decoded addresses and regulated voltages let you put all your effort into what they did rather than barely getting them to work. the dynamic memory allowed cheaper larger address spaces and the standardization of the video (all apples had to have the same video card) meant all games written would work on all apples. the same was not true of the others' since every s-100 bus machine had some different video card standard.
the use oif software decoding of keyboards and disks and so forth introiduced an era that eventually led to the apple desk top bus in the macintosh. What a brilliant simplication. Now we of course have USB instead of different ports for keyboards, parallel printers, scsi drives, tablets, mice.... But the only reasons we went down that track was Woz's apple paved the way. by making so much of the hardware immutable, the software could rely on standard configurations in every machine and thus software timing of other events became reliable for the very first time.
so this is BS revisionism to say that Woz was all about openness and Jobs all about lock down.
What it was both. lock downs of previously unlocked down things created growth to build on. you were not constantly re-inventing the wheel from scratch. In case you have not noticed it before the thing that makes apples great is they always are expensive: this is because they spec them out at high levels using fewer but a complete set of advanced components even on base models. This means software can always count on a feature being there and thus not shoot for the lowest common denominator. think back to pre-w
this is a false dichotomy forged by suggesting that this Steve is good, ergo this steve is bad, then amplifying those traits by mapping them on to perceived standards of today.
I built and sold homebrew computers in the era when the apple II hit the market. At the time we all laughed at the apple as a "toy" because it was so locked down and not built from components. Back then, sonny, you built a computer like an Imsai, altair, cromenco, by starting with a metal box, putting in a non-switiching power supply, choosing the largest capacitors you could fit in the box, then an s-100 (altair) buss. then you picked a cpu board from one manufacturer, some memory cards from another, a keyboard uart decoder from another, a keybaord from another, a video card, and a TV screen modded with an RF converter on channel 4.
These apples were hideously locked down. Switching powersupplies with just wires coming out of a metal box, no way to ugrade the capacity and very little excess capacity. the keyboard was integrated into the case ! and wholly shit a mother board with soldered in chips, video, meomery, and CPU.
Even the address space of the cards you plugged in was decoded on the motherboad not the cards (which allowed the cards to be smaller than the ones for the S-100 bus). THe cards even got regulated voltages not raw rectified AC.
they sucked all the flexibility out of it.
the software was essential to the operation of the hardware not separate from it: a lot of the video management was done in software. the timing one the disk drives they put out used soft sectors not hardware determined sectors (only one hole punched in the floppy instead of 20, one for each sector). Even the memory refresh was handeled on the video updates which in turn were backsided on last half of the 6502's instruction cycle (when it would not be fetching). It was one of the very first systems to successfully use dynamic memory. (Only a fool would not use static memory in an altair, since you had to do all the refresh handling on the memory card).
You had to buy apple floppy disks, and apple plug-in cards for many things cause they were not standard cards or drives.
And of course the apple II in hind sight was one of the most geniuous machines ever built. it's lock downs let hobbiest's soar in other directions. plug in cards were small and the pre-decoded addresses and regulated voltages let you put all your effort into what they did rather than barely getting them to work. the dynamic memory allowed cheaper larger address spaces and the standardization of the video (all apples had to have the same video card) meant all games written would work on all apples. the same was not true of the others' since every s-100 bus machine had some different video card standard.
the use oif software decoding of keyboards and disks and so forth introiduced an era that eventually led to the apple desk top bus in the macintosh. What a brilliant simplication. Now we of course have USB instead of different ports for keyboards, parallel printers, scsi drives, tablets, mice.... But the only reasons we went down that track was Woz's apple paved the way. by making so much of the hardware immutable, the software could rely on standard configurations in every machine and thus software timing of other events became reliable for the very first time.
so this is BS revisionism to say that Woz was all about openess and Jobs all about lock down.
"If comcast could profit from bit torrent then they will be happy for it because, when done correctly, bit torrent more efficiently broadcasts across the edges of the network rather than the backbone."
So, you don't think that the average Bittorrent user pays for higher-priced plans than the average non-user? Network neutrality allows charging users for their network usage, it is meant to prevent discrimination based on type of usage. I think it is probably safe to say that people downloading lots of content on the internet (and uploading lots) pay to have faster service than people who don't use it much.
In a nutshell I agree that metered service would be attractive. But I disagree that Comcast is making substantial profits because the average bit torrent user is buying higher priced plans than they would otherwise.
there are several reasons for this. First I suspect bit totrrent users are the sort that pay for fast connections no matter what. second, if you have noticed comcast's pricing has a large fixed cost and then small marginal increases for more bandwidth. They are banking on people not really using it.
third, and perhaps most important of all, in most cases the last mile of comcast's network is a shared line. They can't just increase and decrease capacity at will because some customer wants more now and less later. They have to count on shared behaviour, and burstiness in comms to make the service seem fast to the users. Steady state high band width users are 1) not being charged the cost they impose in signal degredation on the others 2) comcast can't just build out the worst case scenario for all neighborhood usages.
that might change when capacity at the edges exceeds demand cheaply: e.g. fiber netowrks to the home. this pushes the bottleneck up stream where it will become practical to meter usage.
SO no I don't believe comcast sees bit torrent as a good thing right now. But they could if they were getting a cut from the content delivery folks.
That's cute, but the FCC is subject to the Supreme Law of the land, just like every other part of the central U.S. government. And the Supreme Law says:
"To regulate Commerce with foreign Nations, and among the several States..." and "The powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people."
So you see the FCC has the power to regulate "communications" AMONG the states, not inside the states, and many ISPs operate within state lines, therefore the court reached the decision that Comcast is outside the central government's jurisdiction.
You are armed with an insufficient education on constituional law interpretation. Wiccard V. Filburn says why the FCC and other agencies can regulate within states. Since it's the basis for nearly every federal law, it's not going to get overturned! so you are simply not correct.
>>>Notice the word Communications. So it seems like they might have some authority here.
Yes the FCC has authority over commerce/communications AMONG the States. Not inside. Not over a local ISP that operates inside a town or county and does not cross the border. (Try reading the Supreme Law sometime.)
I think you are pretty confused over what the government can regulate within a state. Read the case of Wiccard's Wheat sometime. In a nutshell, if an action within a state can affect commerce outside the state or the products of an action (e.g. Agriculture) could eneter or affect a market that goes across the state then the commerce clause of the constitution allows the feds to make a law about it.
Frankly I don't agree with that at all. But there's no any controversy at all over this being the operative interpretation. In the case of Wiccard a farmer grew some wheat on his own land for his own personal consumption. This was at a time when the fed's were directing what crops could or could not be grown. The feds argued that if the wheat had not been grown, the farmer would have had to purchase wheat for his use on the market and this affected commerce.
Pretty ridiculous right? yes. ambiguous? no. ever since then the feds have used this and a few other precedents to make almost all federal laws.
For example, why is murder in a school zone a federal crime? where do they get the authority to federalize hate crimes? Why can't people raise marijuana for their own use?
Disagree all you want. But there is zero doubt the FCC has jurisdiction within states.
indeed all the posts on slashdot crossed state lines as did the advertising on this web page.
Is there a proposed ultimate limit for lithography before one has to jump to molecular electronics? 25nm is well below what anyone though practical a decade ago (since it's so many times smaller than easily produced optical wavelengths). Now it's closing in on the limit of easily produced x-rays.
while the resolution of the smallest resolvable element is shrinking, is the utilization of area increasing proportionally. That is are we densely filling the area with 25nm structures or is that simply the finest linear element and these are well separated?
A 1cm chip would have 1E15 resolvable points at 0.025 micron resolution. And then there is the vertical resolution to multiply that. I should think it would become prohibitively difficult to design something with so many possibilities.
Running software not designed for it is not the real issue here. The issue is what is a stake and how appropriate the lockdown is.
apple tends to argue for "speed bump" DRM. basically make something difficult enough or a sufficient game of cat and mouse on the one hand (the speed bump) and and offer an express lane you can pay for. So for example, itunes. you can break the audio files if you want to. they impose some speed bumps to make it not worth your effort. then they offer enhanced value for staying in the itunes eco-system: the seemless updates to the ipod, cover art access, organization of the meta data, safe store, etc... All things you could do on your own but would most people would prefer it to happen magically rather farting around with bit torrent or sending your credit card to some russian mafia website.
Now for somethings like the iphone they have taken a much more agressive lockdown. I rationalize that by thinking about what they are protecting. You don't want crazy shit happening on a cell phone. so you make it hard to install anything not vetted by the mothership. Even the android market has this vetting. It's not that you can't do it. they just make it even harder. people will get much more enraged if their cell phones crash or the cell network itself starts malfunctioning.
SO that makes sense.
the Ipad is sort of in the middle regime so it's going to be more of a test of how apple wants to go. it's really more of a general purpose computer. some units don't have cell phones in them, and even those are not there primarily for voice but for data.
so you could see them going either way on this one. My suspiscion is they will try to maintain the lockdown. that's what they did with the ipod touch (which is not a cell phone).
The new rationale will be that this is an appliance not something you are supposed to mess with. protection for content owners will be seen as paramount over make-like access to the internals. Byt treating it as a appliance that protects content owners they will be able to more freely provide content without onerous access modalities. This will keep the device behaving more like an appliance than a computer.
COntrary to cory doctrow I see this as good. why? well it's not an either or situation, it's an all of the above. If I want true access to my computer then I should buy a computer that allows this. it's called a laptop. I can put linux on it. it's mine to mod. But I should expect that I'll also run into access restrictions from content owners. I might find that less user freindly. On the other hand if I want easy access then there's this appliance I can use for that. I can't modify it. that's the trade, but it's a trade that gives me a value I want.
you could wish for both in one device and if this were easy to provide then someone will do it.
but because both devices, laptop and appliance exist, I have not lost anything.
Slashdot Mirror
And by the way, which one is Pink?
--- squirted from my Zune.
now that's funny.
first it's not the virus that is doing anything. it's just a scaffold. the virus just self-assembles the scaffold for you. the interior DNA / RNA is irrelevant.
that said, the design for the self assembly and display is in the virus DNA I believe. so given a host to express itself in, it could presumably reproduce this in the wild. it would not be any use to the virus. But you could imagine that some host cell might harness the virus to make hydrogen for it's own purposes.
So I suspect that if this gets loose in the wild that the virus won't keep this trait long enough for some host cell to adapt to taking advantage of it.
Basically, hogwash. Having one language and one tool chain leads to fewer methods of being able to solve problems, and reduces the utility of your computer system.
that is an absurd assertion. Look at the success of the JVM, and .Net and the CLI. look at the GNU (multi-language) compiler. These are considered marvelous innovations. look at how linker's work to allow you to call from one language into another.
What all those have in common is defining an interface layer that is standardizes things.
But those are high level. if you want to start thinking about the future of fully automated multi-processing you have to move the interface layer down to find the re-orderable sections of code.
The rest of your assertions have no techincal merit.
Does a single thing you said explain why they won't allow frameworks which compile to Objective-C?
That's where I'm having trouble. I can see a technical reason to force people to use a single language, or at least a single runtime representation, in the same way that, say, the new Windows Mobile forces you to use .NET. I don't see any technical reason for them to care what language you originally use to produce it.
I dont' have a complete answer but I can grunt out a few ideas on this.
1)
Consider the problem of python's global interpreter lock. it prevents parallel execution of python: most people are suprised when they discover that a multi-threaded python application excecutes only one thread at a time and so is usually slower than a single threaded python program. (yes there are exceptions, and yes there are non-standard native libs to fix this problem, but let's not get off topic).
If you convert a python code to C, this non-parallizable global interprer lock does not go away!
What this example shows is that even if you output something into objective-c or whatever, you can create programs that impose huge constraints on the way a code must be executed.
2)
consider what happens if apple implements new features. e.g. suppose they built something like openMP into objective C. What happens if the abstraction that converts one language to objective C does not use this feature. For the case I just mentioned, openMP, I could imagine that it would be fairly difficult to convert the openMP processor directives in a machine translated way that would still make sense or work. TO be specific when you convert languages you tend to replace native commands like
for x = 1..3:
y = 3 + x
into function calls in the new language:
for_loop_method(1,3, store( y*, add_objC( x*, int_long(3)) ) )
while the simple for loop would respond to openMP very well, the second form of nested function calls would not work at all in openMP.
thus reducing something to objective-c is not the same as natively programming in Objective-c
There will be a lot of ink spilled here condemning apple, or explaining why apple is "alllowed" to do this. (i.e. blink! don't blink!) There will also people explaining why this a good bussiness move or good for developers or good for consumers.
So this post is here to start a thread on technical speculation on why apple might be doing this from a point of view of computer science and design strategy. For example, we recently witnesses a more nuanced form of multi-tasking that divides the use cases into ones that give better resource management hints to the OS than the carte blance form. likewise we are seeing memory management for apps being sandboxed and encrypted. clearly these are all about evolving the OS to better control it's resources.
Now what might be the resource use case improvement here?
I'll start the speculation with this thought and leave it to others to fill in more. Suppose that if you were thinking ahead to things like ubiquitous multi-processing and beeing able to carbon freeze applications for instant re-starts. It sure would be nice to have the operating system be able to make code elements re-entrant or find the finest grained atomic segements or know what non-threaded subroutines could actually be done threaded and out of order without the programmers actually having to think about this when they wrote code.
One way you could approach this is to use a single compiler or at least a single memory model and calling convention. then either the compiler itself could include hinting, or maybe the OS could even figure it out after the fact.
I note you'd not be able to do this if for example, the original code was coverted direct to assembly without obeying certain rules. (e.g. just imagine self modifying code).
I note that Objective-C is more than pure syntactic sugar on C. unlike say c++ it does not use a compile-time generaged look up table for method resolution. instead objects get messages sent to a standard receiver in the object and this receiver resolves the method at run time. (it does not even have to respond to the message!). Objective-C also bases it's function argument prototype based polymorphism in a much more strict way than C++ so there's in prinicple more information content about what is being passed. I don't want to emphasize any one thing too much. perhaps something I said is technically wrong. My point is that if you use a consistent model that takes advantage of the extra abstraction layers in how function calls occur and how memory is used then potentionally one can do finer grained parallelism automagically.
likewise, carbon freezing applications for instant restarts may become more stable as well if you have better control over what is mutable and what is shared between threads.
just a thought that this could be a purely technical goal to move operating systems and code development into the massively parallel multi-core future.
if you've followed the Grand central innovation then you can see apple is indeed thinking about making dispatch more hintable by developers.
For decades apple users have had to "enjoy" ported games. Games written either for come arcane quirks of antoher platform then ports half assed to apple or games written in some generic way. They don't support accelerations or available stuff. The linux folks know the feeling too of some round direct-X game shoved into a X11 hole.
Apple is saying now that we have a platform advantage it's going to be windows that gets the hand-me-down games. Games targeted to agnostic APIs are banned too cause it just means people are writting for the lowest common denominator.
THis is actually not entirely new behaviour. Apple always aims for highly spec-ed platfomrs-- never a stripped low end one-- and thus developers have always been able to assume a feature rich platform. for example, in the long ago bad old days apple's always had full stereo sound out when PCs had all sorts of different levels of sound outputs -- usually just a beep speaker. One an cite many such examples.
thus apple wants developers to target it's full capabilities not use generics.
it's their choice.
I'm puzzled over the "per dive" part of this. this seems to imply it's that downward dive itself that is some how producing the energy. that is to say this energy is not continuously produced but would require another dive cycle to produce. So it' can't stay down. perhaps it can cycles dives autonomously?
And how is this power produced. I'm going to guess what is happening is that it comes up and warms up till it's core temperature is at the ambient surface temperature. Then it drops like a rock, and uses the heat differential between the core and the cold water to drive some thermo electric engine in reverse. perhaps they toss in some phase change material to extend the thermal capacity.
or is it something different?
Hmmm..... So if I had to guess then the symmetry is an artifact of the bounding box. That's curious. You'd think it would have residual asymmetry from the sun trajectory. I suspect there is some funny assumptions being imposed directly or indirectly. e.g. perhaps the layoput of the alelles in the genetic algorithms itself slightly favors self similar patterns
Looking at the slides more carefully, I think there's some substantially strange assumptions being made. Notice that he starts from completely random and non-symmetric shapes and these are evolved in genetic algorithms. the results he shows are all highly symmetric. some have 3 C4 rotation axes.
this makes no sense to me. the suns seasonal variation and arcs do not illuminate the ground symmetrically. So it is hard to see why it would evolve to a symmetric structure.
so there have to be some assumptions here the article is not exposing. like enforcement of symmetry.
It's an interesting, nerdy endeavor, but less practical than automated tracking systems; the expensive part of solar is the panels themselves. From TFA: His new designs are up to two and a half times more efficient per comparative length and width than traditional flat arrays.
If solar cells were free, than this would indeed be more efficient, and if there's limited space thay MAY be more practical.
Exactly!
consider the simplified case of sun's arc not moving with the seasons. then you could put down the panels in a 90 degree zig-zag. this way all light that is reflected is assured to strike a second panel. this would dramatically increase the efficiency and reduce the variation throughout the day. but it would take 1.4 times as much panels to cover the same area as a flat panel. if you go for the 3D full corner cube then it's 1.7 times as much.
If you were to spread this out you would have 40% more area. this would mean that at peak power you'd get 1.4 times as much, but at obtuse incidence angles were the reflection is high you'd take a loss. the trade off point is when the reflective loss is greater than 40% I think.
another problem with a highly faceted desing is going to be in making the nominally circular cells conform to odd shaped facets, and for mass producing these. If you look at conventional panels you see they cut they often circular cells into half-circles then put these down in a row laternating the directions. this allows them to make mass producable long sections that dont have as much dead space when the components are placed side by side. If you have facets of differenting shapes you have to make eachone differently and the chips may have to be cut differently.
The best part of this idea is the continuous power level however.
Get your history straight. You didn't have to buy an original Apple ][ with case, keyboard, or power supply. You could simply purchase the motherboard (board on stick) and provide your own keyboard, case, and huge, hot, inefficient non-switching power supply. Such setups were never common because most people didn't want them.
Which I think is the point I was making, no?
The video management was all done in hardware. You're probably thinking of the Sinclair zx80/81 or maybe one of Don Lancaster's TV typewriter designs.
While Don lancaster's design took things to extreme, lots of the video management was done in software on the apple. for example, sprites. and at a more basic level, the colors were interleaved so you did not actually have the full resolution in color that you had in black and white. That translation of pixel position offset with color was handled in the device drivers as I recall.
The Apple ][ used an ASCII keyboard. No interesting software decoding here.
sure there was. how do you think the key presses were captured? they did not have any buffering.
I find it hard to believe you call the Apple ][ locked down when the schematics and firmware listings were in the back of every single user manual!
I think you missed my point. I was being ironic. the apple ][ was being heralded as being open. And by modern standards it is. but for the standards of it's time it was highly integrated and less flexible.
But my larger point is, to move forward we have to periodically draw a black box around some bit of technology and not worry about home brewing that part. You would not I expect want to build your own SATA replacement from scratch in algorithms and hardware (resitors and capacitors). You might want access at different levels of balck boxing. 1) a mac mini, where you don't even care if it's SATA or IDE 2) mutherboard with integrated SATA from Frys. 3) a sata controller card to plug in to a motherboard. but none of those dives very deep. it's closed at the SATA layer and that's a good thing if you want to get something done.
the Ipad is closed at the hardware layer. basically all you can do is write software or build something for it's two connector ports.
but anyhow the point was the exact same complaint was made about the apple ][ in it's day. It's only later we appreciated how freeing us from the details extended our capabilities to modify more complex aspects of the device.
Thanks!, and your post Reminds me of Transgressing the Boundaries: Towards a Transformative Hermeneutics of Quantum Gravity by Alan D. Sokal.
"""
There are many natural scientists, and especially physicists, who continue to reject the notion that the disciplines concerned with social and cultural criticism can have anything to contribute, except perhaps peripherally, to their research. Still less are they receptive to the idea that the very foundations of their worldview must be revised or rebuilt in the light of such criticism. Rather, they cling to the dogma imposed by the long post-Enlightenment hegemony over the Western intellectual outlook, which can be summarized briefly as follows: that there exists an external world, whose properties are independent of any individual human being and indeed of humanity as a whole; that these properties are encoded in ``eternal'' physical laws; and that human beings can obtain reliable, albeit imperfect and tentative, knowledge of these laws by hewing to the ``objective'' procedures and epistemological strictures prescribed by the (so-called) scientific method.
But deep conceptual shifts within twentieth-century science have undermined this Cartesian-Newtonian metaphysics1; revisionist studies in the history and philosophy of science have cast further doubt on its credibility2; and, most recently, feminist and poststructuralist critiques have demystified the substantive content of mainstream Western scientific practice, revealing the ideology of domination concealed behind the façade of ``objectivity''.3 It has thus become increasingly apparent that physical ``reality'', no less than social ``reality'', is at bottom a social and linguistic construct; that scientific ``knowledge", far from being objective, reflects and encodes the dominant ideologies and power relations of the culture that produced it; that the truth claims of science are inherently theory-laden and self-referential; and consequently, that the discourse of the scientific community, for all its undeniable value, cannot assert a privileged epistemological status with respect to counter-hegemonic narratives emanating from dissident or marginalized communities. These themes can be traced, despite some differences of emphasis, in Aronowitz's analysis of the cultural fabric that produced quantum mechanics4; in Ross' discussion of oppositional discourses in post-quantum science5; in Irigaray's and Hayles' exegeses of gender encoding in fluid mechanics6; and in Harding's comprehensive critique of the gender ideology underlying the natural sciences in general and physics in particular.7
Here my aim is to carry these deep analyses one step farther, by taking account of recent developments in quantum gravity: the emerging branch of physics in which Heisenberg's quantum mechanics and Einstein's general relativity are at once synthesized and superseded. In quantum gravity, as we shall see, the space-time manifold ceases to exist as an objective physical reality; geometry becomes relational and contextual; and the foundational conceptual categories of prior science -- among them, existence itself -- become problematized and relativized. This conceptual revolution, I will argue, has profound implications for the content of a future postmodern and liberatory science.
"""
That's a very old joke.
posting this again, since something went wrong the first time:
this is a false dichotomy forged by suggesting that this Steve is good, ergo this steve is bad, then amplifying those traits by mapping them on to perceived standards of today.
I built and sold homebrew computers in the era when the apple II hit the market. At the time we all laughed at the apple as a "toy" because it was so locked down and not built from components. Back then, sonny, you built a computer like an Imsai, altair, cromenco, by starting with a metal box, putting in a non-switiching power supply, choosing the largest capacitors you could fit in the box, then an s-100 (altair) buss. then you picked a cpu board from one manufacturer, some memory cards from another, a keyboard uart decoder from another, a keybaord from another, a video card, and a TV screen modded with an RF converter on channel 4.
These apples were hideously locked down. Switching powersupplies with just wires coming out of a metal box, no way to ugrade the capacity and very little excess capacity. the keyboard was integrated into the case ! and wholly shit a mother board with soldered in chips, video, meomery, and CPU.
Even the address space of the cards you plugged in was decoded on the motherboad not the cards (which allowed the cards to be smaller than the ones for the S-100 bus). THe cards even got regulated voltages not raw rectified AC.
they sucked all the flexibility out of it.
the software was essential to the operation of the hardware not separate from it: a lot of the video management was done in software. the timing one the disk drives they put out used soft sectors not hardware determined sectors (only one hole punched in the floppy instead of 20, one for each sector). Even the memory refresh was handeled on the video updates which in turn were backsided on last half of the 6502's instruction cycle (when it would not be fetching). It was one of the very first systems to successfully use dynamic memory. (Only a fool would not use static memory in an altair, since you had to do all the refresh handling on the memory card).
You had to buy apple floppy disks, and apple plug-in cards for many things cause they were not standard cards or drives.
And of course the apple II in hind sight was one of the most geniuous machines ever built. it's lock downs let hobbiest's soar in other directions. plug in cards were small and the pre-decoded addresses and regulated voltages let you put all your effort into what they did rather than barely getting them to work. the dynamic memory allowed cheaper larger address spaces and the standardization of the video (all apples had to have the same video card) meant all games written would work on all apples. the same was not true of the others' since every s-100 bus machine had some different video card standard.
the use oif software decoding of keyboards and disks and so forth introiduced an era that eventually led to the apple desk top bus in the macintosh. What a brilliant simplication. Now we of course have USB instead of different ports for keyboards, parallel printers, scsi drives, tablets, mice.... But the only reasons we went down that track was Woz's apple paved the way. by making so much of the hardware immutable, the software could rely on standard configurations in every machine and thus software timing of other events became reliable for the very first time.
so this is BS revisionism to say that Woz was all about openness and Jobs all about lock down.
What it was both. lock downs of previously unlocked down things created growth to build on. you were not constantly re-inventing the wheel from scratch. In case you have not noticed it before the thing that makes apples great is they always are expensive: this is because they spec them out at high levels using fewer but a complete set of advanced components even on base models. This means software can always count on a feature being there and thus not shoot for the lowest common denominator. think back to pre-w
this is a false dichotomy forged by suggesting that this Steve is good, ergo this steve is bad, then amplifying those traits by mapping them on to perceived standards of today.
I built and sold homebrew computers in the era when the apple II hit the market. At the time we all laughed at the apple as a "toy" because it was so locked down and not built from components. Back then, sonny, you built a computer like an Imsai, altair, cromenco, by starting with a metal box, putting in a non-switiching power supply, choosing the largest capacitors you could fit in the box, then an s-100 (altair) buss. then you picked a cpu board from one manufacturer, some memory cards from another, a keyboard uart decoder from another, a keybaord from another, a video card, and a TV screen modded with an RF converter on channel 4.
These apples were hideously locked down. Switching powersupplies with just wires coming out of a metal box, no way to ugrade the capacity and very little excess capacity. the keyboard was integrated into the case ! and wholly shit a mother board with soldered in chips, video, meomery, and CPU.
Even the address space of the cards you plugged in was decoded on the motherboad not the cards (which allowed the cards to be smaller than the ones for the S-100 bus). THe cards even got regulated voltages not raw rectified AC.
they sucked all the flexibility out of it.
the software was essential to the operation of the hardware not separate from it: a lot of the video management was done in software. the timing one the disk drives they put out used soft sectors not hardware determined sectors (only one hole punched in the floppy instead of 20, one for each sector). Even the memory refresh was handeled on the video updates which in turn were backsided on last half of the 6502's instruction cycle (when it would not be fetching). It was one of the very first systems to successfully use dynamic memory. (Only a fool would not use static memory in an altair, since you had to do all the refresh handling on the memory card).
You had to buy apple floppy disks, and apple plug-in cards for many things cause they were not standard cards or drives.
And of course the apple II in hind sight was one of the most geniuous machines ever built. it's lock downs let hobbiest's soar in other directions. plug in cards were small and the pre-decoded addresses and regulated voltages let you put all your effort into what they did rather than barely getting them to work. the dynamic memory allowed cheaper larger address spaces and the standardization of the video (all apples had to have the same video card) meant all games written would work on all apples. the same was not true of the others' since every s-100 bus machine had some different video card standard.
the use oif software decoding of keyboards and disks and so forth introiduced an era that eventually led to the apple desk top bus in the macintosh. What a brilliant simplication. Now we of course have USB instead of different ports for keyboards, parallel printers, scsi drives, tablets, mice.... But the only reasons we went down that track was Woz's apple paved the way. by making so much of the hardware immutable, the software could rely on standard configurations in every machine and thus software timing of other events became reliable for the very first time.
so this is BS revisionism to say that Woz was all about openess and Jobs all about lock down.
"If comcast could profit from bit torrent then they will be happy for it because, when done correctly, bit torrent more efficiently broadcasts across the edges of the network rather than the backbone."
So, you don't think that the average Bittorrent user pays for higher-priced plans than the average non-user? Network neutrality allows charging users for their network usage, it is meant to prevent discrimination based on type of usage. I think it is probably safe to say that people downloading lots of content on the internet (and uploading lots) pay to have faster service than people who don't use it much.
In a nutshell I agree that metered service would be attractive. But I disagree that Comcast is making substantial profits because the average bit torrent user is buying higher priced plans than they would otherwise.
there are several reasons for this. First I suspect bit totrrent users are the sort that pay for fast connections no matter what. second, if you have noticed comcast's pricing has a large fixed cost and then small marginal increases for more bandwidth. They are banking on people not really using it.
third, and perhaps most important of all, in most cases the last mile of comcast's network is a shared line. They can't just increase and decrease capacity at will because some customer wants more now and less later. They have to count on shared behaviour, and burstiness in comms to make the service seem fast to the users. Steady state high band width users are 1) not being charged the cost they impose in signal degredation on the others 2) comcast can't just build out the worst case scenario for all neighborhood usages.
that might change when capacity at the edges exceeds demand cheaply: e.g. fiber netowrks to the home. this pushes the bottleneck up stream where it will become practical to meter usage.
SO no I don't believe comcast sees bit torrent as a good thing right now. But they could if they were getting a cut from the content delivery folks.
That's cute, but the FCC is subject to the Supreme Law of the land, just like every other part of the central U.S. government. And the Supreme Law says:
"To regulate Commerce with foreign Nations, and among the several States..." and "The powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people."
So you see the FCC has the power to regulate "communications" AMONG the states, not inside the states, and many ISPs operate within state lines, therefore the court reached the decision that Comcast is outside the central government's jurisdiction.
You are armed with an insufficient education on constituional law interpretation. Wiccard V. Filburn says why the FCC and other agencies can regulate within states. Since it's the basis for nearly every federal law, it's not going to get overturned! so you are simply not correct.
No you are incorrect. the FCC does have the authority to regulate inside states.
Comodoe64_love, you are wrong. see this post and the one below it to learn about constitutional and federal law
Here is the case law that lets the FCC regulate this
>>>Notice the word Communications. So it seems like they might have some authority here.
Yes the FCC has authority over commerce/communications AMONG the States. Not inside. Not over a local ISP that operates inside a town or county and does not cross the border. (Try reading the Supreme Law sometime.)
I think you are pretty confused over what the government can regulate within a state. Read the case of Wiccard's Wheat sometime. In a nutshell, if an action within a state can affect commerce outside the state or the products of an action (e.g. Agriculture) could eneter or affect a market that goes across the state then the commerce clause of the constitution allows the feds to make a law about it.
Frankly I don't agree with that at all. But there's no any controversy at all over this being the operative interpretation. In the case of Wiccard a farmer grew some wheat on his own land for his own personal consumption. This was at a time when the fed's were directing what crops could or could not be grown. The feds argued that if the wheat had not been grown, the farmer would have had to purchase wheat for his use on the market and this affected commerce.
Pretty ridiculous right? yes. ambiguous? no. ever since then the feds have used this and a few other precedents to make almost all federal laws.
For example, why is murder in a school zone a federal crime? where do they get the authority to federalize hate crimes? Why can't people raise marijuana for their own use?
Disagree all you want. But there is zero doubt the FCC has jurisdiction within states.
indeed all the posts on slashdot crossed state lines as did the advertising on this web page.
oops my bad: 1.6E11 resolvable points on a 1cm chip. still a lot to design for.
Is there a proposed ultimate limit for lithography before one has to jump to molecular electronics? 25nm is well below what anyone though practical a decade ago (since it's so many times smaller than easily produced optical wavelengths). Now it's closing in on the limit of easily produced x-rays.
while the resolution of the smallest resolvable element is shrinking, is the utilization of area increasing proportionally. That is are we densely filling the area with 25nm structures or is that simply the finest linear element and these are well separated?
A 1cm chip would have 1E15 resolvable points at 0.025 micron resolution. And then there is the vertical resolution to multiply that. I should think it would become prohibitively difficult to design something with so many possibilities.
Running software not designed for it is not the real issue here. The issue is what is a stake and how appropriate the lockdown is.
apple tends to argue for "speed bump" DRM. basically make something difficult enough or a sufficient game of cat and mouse on the one hand (the speed bump) and and offer an express lane you can pay for. So for example, itunes. you can break the audio files if you want to. they impose some speed bumps to make it not worth your effort. then they offer enhanced value for staying in the itunes eco-system: the seemless updates to the ipod, cover art access, organization of the meta data, safe store, etc... All things you could do on your own but would most people would prefer it to happen magically rather farting around with bit torrent or sending your credit card to some russian mafia website.
Now for somethings like the iphone they have taken a much more agressive lockdown. I rationalize that by thinking about what they are protecting. You don't want crazy shit happening on a cell phone. so you make it hard to install anything not vetted by the mothership. Even the android market has this vetting. It's not that you can't do it. they just make it even harder. people will get much more enraged if their cell phones crash or the cell network itself starts malfunctioning.
SO that makes sense.
the Ipad is sort of in the middle regime so it's going to be more of a test of how apple wants to go. it's really more of a general purpose computer. some units don't have cell phones in them, and even those are not there primarily for voice but for data.
so you could see them going either way on this one. My suspiscion is they will try to maintain the lockdown. that's what they did with the ipod touch (which is not a cell phone).
The new rationale will be that this is an appliance not something you are supposed to mess with. protection for content owners will be seen as paramount over make-like access to the internals. Byt treating it as a appliance that protects content owners they will be able to more freely provide content without onerous access modalities. This will keep the device behaving more like an appliance than a computer.
COntrary to cory doctrow I see this as good. why? well it's not an either or situation, it's an all of the above. If I want true access to my computer then I should buy a computer that allows this. it's called a laptop. I can put linux on it. it's mine to mod. But I should expect that I'll also run into access restrictions from content owners. I might find that less user freindly. On the other hand if I want easy access then there's this appliance I can use for that. I can't modify it. that's the trade, but it's a trade that gives me a value I want.
you could wish for both in one device and if this were easy to provide then someone will do it.
but because both devices, laptop and appliance exist, I have not lost anything.