lua-users home
lua-l archive

[Date Prev][Date Next][Thread Prev][Thread Next] [Date Index] [Thread Index]


Why not "luaverse" ? (akin to "universe", "versatility", and the ability or "verse" to express many things, i.e. transcribing it in its own "words" or language that "others" can understand, and possibly reproduce).

Aren't people on this list all "luaverse" (hopefully), as if they were versed from a spreaded chaotic world into a new container organizing it locally to create some order and being able to describe it, and reuse that order to create new useful, or just beautiful things ?

Isn't "luaverse" less negative (in perception) than "luantic" or "lunatic" (where imagination and creativity does not create any useful order, but just is an additional source of chaos, and an unrecoverable loss of energy)?

Note that Lua is still not "omniverse" (and IMHO, it's quite futile to pretend that it will help organize and solve all kinds of problems; notably because it is still restricted as living in a Turing machine with finite binary states; the next language we'll see in a near future will no longer be Turing-machines, i.e. fully representable with binary arithmetic, but will work with probabilities and new kinds of "fully logic gates", and with new exceptional capabilities, notably by fully destroying the need of any clock with incremental time, and implemented in an incremental space which for now is still mostly a 2D space even if new chips have a so-called "3D" topology, which is just in fact the superposition of a few incremental planes.)

Even before we get to a full fuzzy logic machine, we'll have another intermediate step working with crystalline structures in real 3D patterns, which will be useful to bypass the current Moore law which has almost reached its limit.

For now we have tried to avoid that limit using networking, but at large scales of spaces, we are still not at the scale of crystalline structures, where the computing units would not longer be small gates but resonators in crystals, and a fully distributed, i.e. delocalized, information with extreme levels of parallelism (this could be the real start of "artificial intelligence", really mimicking what our brains does. Locally that 3D crystal would delocalize the information, so that it will be only localized with some probability (think about Shannon theory of information: information unit can be smaller than one bit stored on one place, and it is usable at any precision level provided that the network connecting the nodes is large enough, and this also grows with the number of connections that nodes can make together. In a 2D space with nodes the size of atoms, there's a highest limit of 4 connections, but todays chips need two of them to bring energy to the system; now go 3D and you add another axis with 2 additional connections per node and this dramatically/exponentially increases the "computing" capability of these units so that the unit of information can now be any fraction of bit, depending only on the total size of the crystal; that system will also be highly resilient to "errors" or local defects in the crystal).

How will we create programs for these new kind of computers? certainly not with languages designed for Turing machines. We will no longer depend on the imperative order of events with a clock, instead there will be equations driving us and converging hopefully to a new state of stability (including reasonating states, which are definitely not "stable" i.e. fixed to a finite state, but in a state that exhibits measures in timely and spacial dimensions, which can be observed directly from anywhere on the crystal, e.g. only on its tiny surface, meaning that the effect of measurement will be neglectable on the state of the crystal: that crystla will then have strong intrinsic capabilities for BOTH storage and computation).

What will be the computing limit of speed for suth computer ? It will be limited of course the speed of light in the crystal, so the smaller the crystal the best it will be (but then to preserve the computing capability, the crustal will have to be denser, meaning that it will also materially slow down the speed of light (with a higher indice of refraction).

In which material will these crystals made? Hopefull they should be isotropic and (not uniaxial or biaxial) to maximize the propagation of signals thoughout the crystal with the best distribution, and will have a high indice of refraction. This brings us to what we already use today: silicium (or better with high indices of refraction, but a bit more expensive: germanium, gallium, indium); diamonds (carbon) are excluded here due to their cost (so they are too small in size) and their refraction indice is lower than silicium.

We could also imagine using common glass (low cost) because it can be easily used to build very large crystals (provided that we don't break them: common glass is very fragile and small defects generally propagate to very large planes of fractures; they can break not just because of chocks but also only by differences of temperature under the internal stress of pression caused by dilation or retraction).

Will have have real 3D crystalline chips using silicium (possibly doped by arsenic or similar) acting as a single giant gate, and programmed by feeding signals on one side and collecting the results on the opposite side after they've generated some resonating patterns inside the crystal without burning or breaking it? Can we imagine a programming language to work with such computing chip (we'll just have to take into account the speed of propagation of light signals in the crystal, but it is still extremely low compared to existing binary gates that we need to stabilize, and will use MUCH less energy: stabilization of binary gates requires feeding or extracting lot of energy to satisfy requirements set by the clock limit)?

Get ready, we are just seeing new researches resurrecting the concept of "analog computers": the "digital computer" was successful but has reached its limits. Tomorrow computers will work in dramatical different ways if they want to go faster and be more powerful and more universal! This means the end of Turing machines and reducing the use of imperative languages (including Lua...) only to interfaces (including, but not necessarily for long distance communication over a network: may be there will be a much better worlwide network than today's Internet, using new kinds of non-binary protocols, and working really like a giant crystal with its resonators and natural resilience to local defects; that future Internet will necessarily be fully peer-to-peer; today's Internet is only partially 2P2 but with wellknown central points of controls, so it is highly hierarchical and organized with purely binary decisions and cannot replicate what we'll have in tomorrow's computers: today's Internet will then be reducved to become only a light interface on the surface of a much more powerful fully-P2P network, which will also need its own non-binary protocols, and its own programming language compatible with what tomorrow's computers will do internally, so that tomorrow's new Internet will be the natural way to scale up tomorrow's non-binary computers).

But are non-binary computers only for the future? Don't they already exist somewhere that we are not aware of? May be yes: look at the nature all around us. May be we don't even need to use new materials with high efficiency (i.e. high indices of refraction), given the gigantic size of our environment. May be we can program that nature for our interests. Or nature is just already running a program (in its own programming language: the natural laws of physics, that we still cannot fully decipher, even if can already make some use of it).



Le mar. 28 déc. 2021 à 23:56, Lorenzo Donati <lorenzodonatibz@tiscali.it> a écrit :
On 28/12/2021 23:26, Marc Balmer wrote:
> lunatic?
>

Although a nice joke, I don't think it has the same impact as "pythonic".

In English it means either "an insane person" or "reckless,
irresponsible person". Not a nice definition.

Even in other languages it has no positive meaning: in Italian
"lunatico" means a crazy, unreliable, person and I guess in Spanish and
Portoguese it means the same (i.e. "loco").

Although the origin of the term was something like "influenced by the
moon", and so it could have been viewed as somewhat positive in some
context, nowadays it has a rather negative connotation.

I wouldn't really like to associate that with Lua.


>> Am 28.12.2021 um 23:15 schrieb Lars Müller <appgurulars@gmx.de>:
>>
>> I'm searching for an adjective for "in the spirit of Lua", like
>> "pythonic" for Python. I came up with "lua-esque" but that doesn't sound
>> too well. Suggestions?
>>
>> -- Lars
>

-- Lorenzo