Main Site | Join Robin Hood Coop | Projects | Events | Blog | Media | Forums | Mailing List | Twitter | Facebook

Free internet for third world

The debate of net neutrality and other similar topics when question is free internet for development countries is quite irrevelant. People in third world could not care less about net neutrality, censorship etc., they just want internet connection that is free to them (toll free, subscription line payment free, zero rating). If in first phase payment free internet can even be some sort of monopoly of some firm (or firms) 5-10 years, after that is time to worry about net neutrality etc. This is the same situation as people are starving in third world because they have no food. Without internet connection they are like “starving” due to lack of information without all that knowledge and connections that internet provides. When people are starving you give them food. But making announcements like “I cannot give you food just now altough I know you are starving, because I have first to make sure that this food is properly made” is perfect analogue what is happening if possible free internet connection is denied from people of third world because it is not net neutral or whatever. The food in this case is information and knowledge. Only some time later, when free net connection is available at third world, is time to worry about net neutrality, 5- or 10 years later, and made corrections then if necessery. Most important is to provide free internet connection to third world now, not postpone it to foreseeable future. Google, Twitter, Facebook, Microsoft Math, Wikipedia have all zero rating internet programs. Collecting these together and using data compression and limited internet access with data rate restrictions would make cost free internet (cost free to user) true in third world. Even few dozen kilobytes / sec or few megabytes per day, if it is payment free, is enough to start free internet in third world. Net connection is paid by adverisements. Better data rates can follow later. Now people in third world have no internet connection, so 90% advertising money (220 billion dollars, huge sum, is spend in third world at advertising in every year) goes to traditional outdoor and newspaper marketing in third world, and only 10% goes to internet. This could be otherway around if people there could have at least some kind of free internet option instead of subscription line payments, so 200 billion dollars could go to internet advertising, not just 20 like nowadays. Spotify, Pandora Radio, T-Mobile streaming service, FreeBee Data are examples of costly internet operations that are however free to internet user. They are advertisement paid. Similar structure is suitable for free internet in third world. But instead of music streaming this “Netify” service offers free internet in third world. Because data rates are restricted in free internet and quality of service is not so good, free internet will not alienate those who already pay for internet services in third world and those people does not suddenly jump to free “zeronet” with limited capacity. But people who have no money to pay will use this zeronet. Also phone can be given free to free internet user in third world, because price of electronics is so low nowadays. If one family receives one cheap smarthone, that has two or three “remote screens” that are connected with smartphone with thin wire couple of metres long, information goes with electric current to remote screen like old “electric internet” internet connection through electric cables systems. Remote screen is phone without processor or batteries, only black & white screen, electric ink or other, no touchsreen, it is “remote control unit” for smartphone. Smartphone and remote control units use roll printed electronics in almost every component except processor SoC. Smartphone colour touchscreen, batteries etc. can be made with roll printing. Also headphones, 3 or 4 that comes with phone, are made with roll printing, cheapest possible headphones. This package, cheap roll printed smartphone (with ordinary silicon SoC), two or three remote controllers with screen, couple of headphones, will cost only few dollars and most of that price comes from processor SoC. This phone is advertisement paid like net connection (sponsored). Operating cost of net connection in year is higher than few dollar phone that many people can use simultaneysly with cheap remote control units, so also this phone package can be given free to free internet user. Quadcore ARM SoC cost only 4 dollar couple of year ago, so octacore ARM phone SoC that is ordered from factory 1000 000 000 units for free net phone, will cost very little. iPhone is sort of status symbol in third world, so “iPhony” clone phone that looks like iPhone and has similar looking graphic GUI but that cost nothing, because it comes free with free net connection, would be new mark of equality of technology in third world. Wire connections between smartphone and remote controller units are made with roll printed flat ribbon electric wires. “iPhony” s operating system is however Linux or Android etc. It is also cheaply build using roll printing whenever possible. Compressing data, like graphics in ScreenPressor, audio and video compression etc. so free net if it has audio or video at all that is low quality highly compressed. Text can also be compressed, and most of free net internet traffic is propably text information, because its costs almost nothing to send or receive nowadays. Compression methods like Additive Quantization (Martinez), Anamorphic Stretch Transform compression, “Unicode table compresion in FOX”, and “ScreamingHD” video compression can be used. Also multicast transmissions, so millions of devices receive same information in same frequency and in same channel can be used, so free net is more like radio and television broadcasting where millions of people listens and watches exatly the same radio frequency channel, than normal two way internet traffic. However free net can have hundreads of (net TV) video and audio (net radio) channels on air at the same time. Because only few hundred channels, frequencies and time slots are in use that is much cheaper than usual internet traffic where millions of people occupy millions of channels. Normal two way internet traffic can be provided but it has data retrictions in free net. No voice calls (voice over internet is however possible). Radeeus, Neverthink TV etc. are services that are suitable for free net. Ecological search engines donate some amount of their profit to charity. Search engine that gives money to free internet connection instead of ecology is possible in third world. mCent / Jana and Puggle Me are web browser and search engine respectively for free net. Nobody has noticed “Puggle Me”, cheap program for free net, altough billions of dollars are spend in satellite etc. free internet projects. DataWind Ubisurfer, iSpaces cloud browser / computer are other nobody notices - projects. Ekopedia, healthcare and educational information, Free Basics can also be in free net. Speech synth in smartphone is needed because people in third word cannot read or write in many cases. Connection to public domain, freeware, abondonware etc. content in internet could be possible in free net., and connection to costly material that internet provider must pay, restricted or nonexistant in free net. In netpage Wealthdaily in 2016 is article “The cheapest internet in the world”. In similar way than using one way transmission like net radio and TV also text information could be send, as “book radio service”, continous transmission of text in some channels, and everybody can receive this information and read it or store it and read it later if they want, or pass it by like listening radio or watching TV. If everybody are using same time slot, same frequency, same channel, they can all receive, millions of people, the same information. That is cheap, but it is not two way communication. But sophisticated net TV, net radio, and one way net text transmissions can be used in free net together with restricted two-way communication that is more expensive than just one way transmission from link tower to all receiving devices at the same channel. I don t know why modern smarthopnes must have enormous graphic processing storage, small programs with simple graphics, like in “Tinyapps” netpage, few kilobytes is enough for computer program, megabytes is not needed in many cases. Such “tinyapps” attitude for free net computer programs is needed, not megabytes large programs with fancy graphics. Small amount of simple graphics, and that compressed like ScreenPressor and other data compression and graphics compression programs. Still pictures, audio and video is low quality and highly data compressed in free net. Text cost almost nothing for internet provider to sent, so most of information can be text-based in free net, and even text can be compressed. Instead of many smartphones one smartphone with thin wire connected to smartphone and cheap “remote control units” can be given to one family etc., so one processor SoC and one real smartphone can handle several persons internet use simultaneusly. Articles: “Africa: internet access more important for africans than net neutrality”, “Net neutrality, zero-rating & development: what s the data” 2014, “Zero-rating educational content on the internet” 2016. Normal internet link modems and links need only software patch to turn them to handle IoT traffic together with normal internet. IoT net traffic is designed to put minimal amount of strain to radio frequency communication. If free internet in third world needs communication protocol that is more cost effective than those used nowadays, that can be put to use simply adding software patch to link tower or modem. Large number of teleoperators operating in third world are either wholly or partially owned by state owned companies, from industrial countries, rich middle eastern countries or China (government owned companies). When they are state owned companies they are under parlamental rule. So parlamental decision can bring free internet in third world if there is parlamental will for it. Also private owned companies can bring free internet to third world, it can even be commercially profitable concerning billions of people without net connection (advertisement money pouring to free net in third world begin, if there just would be cost free net connection in third world). What kind of data compression to use? There is Finite State Entropy, Qdigest algorithm, Sparse Fourier Transform (SFT), Tournament coding of integers (Teuhola). Using approximate computing (Rice university), analog processing (Glenn Cowan 2005 VLSI math co-processor, with perhaps analog compression like Xampling, PAL TV quadrature compression, MUSE HD TV compression, and then extra digital compression after signal is already analog compressed and leaves analog circuit), or delta (sigma) modulation. Elettra Venosa texts about “Time interleaved ADC”, Bonavolonta “New approach based on compressive sampling” 2014, Cheung Chan (Hong Kong) “Sample by sample adaptive differential vector quantization (SADVQ)”. Cubic- or spherical vector quantization, RIQ-ADPCM, RIVQ-DPCM (recursively indexed / vector quantization), “pulse group modulation” patents by Hartmann 2002, 2003, “Quantization and greed are good” Mroueh 2013, “High order delta sigma noise shaping” Flintbox netpage, “One bit quantization without delta sigma” Takis Zourntos, “Direct-digital synthesis using delta-sigma modulated signals” Orino, Albert Weggener hardware compression patents Samplify Systems Inc. Feedback Amplitude Modulation FBAM and Vector Phaseshaping Synthesis are sound synthesis methods, but perhaps can be used in signal processing also? Octasys Comp (van den Boom) is data compression. Dithering can be used to improve low bitrate data accuracy. Delta sigma modulation uses some highly efficient noise shaping techniques. Multiple description coding with delta sigma can also be used. Other ways to bring effective digital communication: Mikogo Desktop, Dooble browser (earlier version was purely distributed form), “Flutter” Wifi by Taylor Alexander. And “Using floating point numbers as blockchain”. The price of memory is cheap also nowdays, in netpage Dramexchange NAND price is per gigabyte almost nonexistent. If slow but cheap (RAM) magnetic memory is used in ultra cheap smartphone cheap price is more important than fast memory access (fast memory has small memory capacity per dollar compared to cheap and big but slow memory). Also ROM memory can be mask ROM and that should be even more cheaper than RAM. If phone has inbuild adware in its firmware adverts can be inbuild into phone, so no adverts through net connection is needed so much, and internet data stream saved for other than adverts traffic. Ongoing “marketing campaing” is in use everytime phone is in use, if adware is build in firmware, saving internet connection data stream for “real” communication. Google cardboard virtual glasses are cheap video glasses that use telephone and small frame and nothing else. Plastic version of it, or couple of them for phone and for “remote control” screens can be given away with phone. Webinos, Google Web Intents, Mozilla Web Activities can be used in free net. Pyramid-, cubic-, and spherical vector quantization, “algebraic integer quantization” concept, jittered random sampling (Luo 2014 and Parsons 2014), “Sparse composite quantization”, “Pairwise quantization”, “Multiple base number system” Dimitrov, “multiple base composite integer” at MROB netpage, “preferred numbers” that are integers that offer better accuracy than usual integers (Renard series, E-series) etc. can be used. Perhaps preferred numbers and multiple base composite integer or Dimitrov s multiple base numbers can be combined. Texts not about internet for third world but general internet information use: "The integrated media machine " Mauri Ylä-Kotola 2000, “Querying string databases with modal logic” Matti Nykänen 1997, “Task complexity, information types and information sources” Katriina Byström 1999, “Aspects of relevance in information modelling methodololical principles and conceptual problems” Esko Marjomaa 1997 (ISBN 9514442237), “Digital media in networks 1999 perspectives to digital world” Timo Väliharju 2000, “Designing ICT for mothers user psychological approach” Hanna Parkkola 2006, “Studying user interface strategies and characteristics for developing web search interfaces” Anne Aula 2005, WEBSOM concept (Kohonen), “Adaptive informatics” concept, “Self-organizing governance” (Elinor Ostrom), “Glossasoft” (Timo Honkela), “Peace machine” (Timo Honkela). Other than internet: “Operations research” (Hamdy A. Taha), “Economics of globalisation and the enviroment” 1999 (Leena Lankoski, Jussi Lankoski), the “Triple helix association” and its publications (seven volumes). Modular phones whose innards can be replaced is sometimes presented in smartphone concepts. Because those cheap smartphones that free net uses are very cheap, even batteries perhaps cannot be changed (because they are part of phone frame in super cheap phone), perhaps some slightly expensive smartphones could have replaceable components such as processor SoC and batteries, and perhaps other changeable parts also. So not all new phone is necessery to buy, only new processor or screen etc. People in third world have no money to buy new smartphone in every three years or so, so changeable parts would solve this problem. But only in more expensive than super cheap smartphones. Some own energy generator, like solar cell or Peltier element generator can be used, when there is no other electricity sources. Solar cell or Peltier generator can be build cheaply using roll printing. Also is possible to make cheap portable CD/DVD players to third world market (such are manufactured in China), but even cheaper few dollar priced, where electronic logic and batteries etc. are made with roll printing, altough mechanical parts (electric motor etc.) are conventional. “Ecodisc” cheap CD/DVD/ DVD-ROM discs that are flexible single layer discs can be made at almost no cost per unit (0,1 dollar or so). Ecodisc prototypes of dual layer DVD were made but not put into production. In China ordinary CDs and DVDs are made with less than 0,1 dollar factory price, and Blu ray discs less than one dollar. If BD-ROM discs or recordable Blu ray discs are used, and cheap players for them, cheap optical memory is possible. GoTenna, Shruti Drishti, EnOcean, MyriaNed, Open Garden Inc., Skyfire browser (discontinued), wireless IP phone, SIP phone, Wifi phone, “Beyond P2P: What if drop down?” are means to make internet communication. Wifi phones are cheaper (?) than ordinary phones. Inexact computing in video and audio signal processing can significanly make signal processor simpler (Rice university). Chinese TransOS multi user OS, and shape adaptive transform (“Shape-adaptive transforms filtering” -version etc.), and Sparse Fourier transform are new concepts, and DataWind UbiSurfer browser, Infinit (file sharing), Fon Wifi, FabFi Wifi, Oluvus free internet, Android Be-Bound app, Intabank Bandwidth Bank, CrowdRoaming, Solu Machines SoluOS (discontinued Linux distro), iSpaces cloud computer / browser, Global Villages Inc. internet bus (Mobile access point MAP- principle where motorcycles, busses etc. provide internet, not fixed link towers), and Freerice. “A human right: everyone connected”, and Humanitarian free and open source software (HFOSS) are organisations. Privacy- agnostic prgramming and Policy- agnostic programming (Lifty, Jifty, Binah) and a proposed alternative to internet communication network that is much safer than internet now by Tuomas Valtonen are safer methods to make connection. To make text databases smaller compression is needed. Instead of 8 bit unicode text can use Intelligent Dictionary Based Encoding IDBE, and very large dictionary. Inglish langiuage has 170 000 words, so 18 bit dictionary is needed. 18 bits is 256 000 values so 170 000 for words and 80 000 most used sentences of english can be stored, with about 3000 unicode characters and 3000 IDBE vowels which are 2-4 letter long for example. 18 bit can be mode from 16 bit using Bounded Integer Sequence Encoding BISE or dthering 16 bit to 19-20 bit precision. So instead of 8-32 bit UTF character per letter, one 16 bit data can include complete words and even complete sentences. Using 16 bit floating point IEEE standard or Intel deep learning 16 bit FP format, and mantissa expansion (Gal s accuracy tables revisited method etc.) even more information space is available for text. 8 bit can be used if Microsoft deep learning FP format with 2 bit significand is used, and mantissa expanded with Gal s tables etc. up to 17 -18 bits, enough for english words setc. So 8 bits can store complete word of text and 16 bit not only word but complete (of most used sentences list that is stored) sentence, so text can be compressed to small space. Or use 1 bit delta-sigma modulation with 16 bit internal multibit or multirate processing, now 1 bit is enough for bit compression of 1:16 compression ratio (1 bit is 2 values, 16 bits is 65536 values, so this is huge compression ratio), and this system can be used for other data compression also. Large dictionaries can also be used with video compression, picture is divided to small screens, each small screen has CGI generatd shapes and patterns, only time necoding how shape changes over time, colour and shape pattern for each small screen is needed, video is computer image constructed, so very small bitrate video compression is possible, but it needs large dictionary. Now when even cheap phones have gigabytes of memory, one or two gigabytes can go to this video format memory to store shapes and patterns of image, now still pictures and video can be send with very low bitrate (no individual pixel information is needed, picture is build using “building blocks” of shapes and patterns, that are in phone s own memory, and no individual pixel information is need to send), so only few hundread or thousand bits can build a picture, and moving picture only needs time encoding how these shapes and patterns change over time, so video compression with minimal bitrate is possible. Because floating point numbers of 16 or 32 bits have huge range, and their (mantissa) accuracy can be increased accordingly, only few floating point numbers can be used to point to dictionary of billions values accurately. However picture is computer generated animation, so it propably would not look very convincing. Multiterminal source coding, multiple description coding, multiple base composite integer at MROB netpage, multiple base number system (Dimitrov) etc. can also be used. Posit floating point is relative to a-Law logarithmic encoding, so Gal s accuracy tables method (revisited) method should work with posit floating point format also, Quadiblock (John G. Savard) in Google Groups 29.9 2017 “Beating posits at their own game” text. S. Boldo and Malcolm mantissa expansion can work in posits also maybe. Tethering, application sharing, FreeLAN, Akamai, Diaspora social network, Sulia, Zooniverse, LeanPath, Meka (Uganda), Matibabu, Clinixon, and Moeda are ways to use internet. Broadband over power lines and its better version E-Line (Goubau line) can send internet traffic over power lines. Broadband over power lines and E-Line can perhaps be used simultaneusly in same power cable, because they use different frequencies, increasing capacity. Internet over power cables is suitable solution for third world countries, when mobile internet is not used. If ordinary 3G or 4G network can be turned to IoT communication simply using software, turning 3G or 4G network to special free net (zeronet) communication protocol that makes possible small data rates (with data compression) and economical communication costs is possible simply using suitable software patch, in third world. IoT uses cheap economical communication already. Collecting zero rating internet providers for third world (Google, Facebook, Twitter, Microsoft etc.) together, and using best available cost effective data and internet distribution and data compression techniques, cost effective, advertisement paid, free to user (limited) internet can be true in third world, even without satellites, balloons etc. obscure projects. Free here means without subscription line payments or other cost to user, line connection is advertisement paid / commercially sponsored completely. In India DataWind has in some cases 0,5 dollar or even 0,1 dollar / month payment rates for internet. What if this 0,1 dollar or 0,5 dollar cost is paid also by advertisement money, like rest of the line cost? Totally free of cost to user internet is then true. Wideband connections for free net can be sometime in the future, even small data rate for third world free internet will do if it can be started now.

Quika launched its internet satellite already, and Outernet / Othernet launches its own satellite this year. Those internet satellites are not “free internet”, Quika charges high price for its satellite receiver and Othernet needs its own receiver at ground, its own Wifi hotspot, altough Othernet is aiming to free communication. But Wifi hotspot that competes with ground based teleoperators is however not cost free. Those satellite operators however may start a price competition with ground based teleoperators which is a good thing, maybe Wifi and internet connection then becomes so cheap that true cost free internet in third world countries may start. If ground based teleoperators simply would lessen their profit margins cost free to end user internet could be launched even today in third world. They however have to give internet cheaper than nowadays when those satellite operators open their “free” internet networks if they want to stay with competition. Nokia and Vodafone are building 4G network on the moon, that has nothing to do with third world internet, but moon can be used as a link base for internet communication without satellites. Also phones when offered at third world market, if they are already loaded with free apps before delivery, user had not load those apps themselves, and those useful free apps can be some of that user perhaps does not find anyhow altough he/she would need them, and preloading cheap phones with free useful apps could even lower phone s cost because apps are adware, and more adware phone gets the more phone or telephone / internet line costs can be covered by adware (apps). Several different app blundes can be made from which user could choose most suitable for him / her. App bundles collected towards male- female- and youngster user profiles can be collected, and app bundles towards muisicmaking etc. hobby or work. Several dozen different app bundles can user choose, one or two app bundles fits in phone each time, so user must choose only one or two. Those free app bundles collect useful apps that user (often illiterate african, south asian or south american) perhaps never find out without those software bundles. And cost of phone drops when part of phone s price is covered by app adware. Perhaps app bundle with adware can cover all of cheap phone s cost, or even cheap limited “zeronet” connection cost, so with app bundles when free apps fill most of phone s memory (if user wants memory he/she needs memory card) those apps are not only collection of useful apps for third world developing country user but those apps help to bring phone s cost down to perhaps nothing and make possible limited free internet connection. Perhaps hundread or more apps that fill the phone is adware enough that when user uses the phone there is constant flow of advertisements, and those advertisemens help to pay both phone and limited internet line cost. There are perhaps million free Android apps from which to choose from. Collecting balanced app bundles for third world market for each cheap phone can have large collection of preloaded useful apps, when user gets phone he/she loads the most suitable bundle for him/her in telephone shop etc., and each cheap phone has one or two of them. Third world market needs cheap PC. In China is cheap x86 processors for server (AMD license) and Rockchip smartphone SoC (Intel license). VIA has licensed its x86 chip also to chinese. But what is missing is cheap chinese x86 CPU for PC. Intel and AMD have made licensing deals with chinese manufacturers for x86 chips. If cheap x86 PC processor is made, very cheap mini-PC for third world market with price of few dollars or few dozen dollars, like smartphone, could be made, and very cheap laptops also. Mini PC has no screen or keyboard (sold separately), it is small and very cheap with chinese x86 CPU. Better is two processor mini-PC or laptop PC. One CPU or SoC is x86 and another very cheap ARM SoC / CPU. Now different Android or Linux programs can be run in same PC, or even Windows Phone OS. Switching with OS is easy, Android phone can run with Linux and Symbian OS also with ease. Even iPhone apps can be used with emulator in Android phone. Perhaps better than emulator is Darwin OS in ARM version. Phone or laptop PC or mini-PC needs multiple OSes or several emulators for multiple OSes. If free iPhone apps can be used in cheap “iPhony” clone phone that would bring more real iPhone feel to these cheap third world market phones. An attemp to build multi-OS was long ago, Linux-based E/OS or “Emulator operating system”, but not finished. Others are Cooperative Linux, Longene and Linspire for Windows programs, and then there is PearOS but that is just Linux distro that graphically looks like iOS, without any iOS functionality? Emulators for old Amiga etc. software can be used in modern phone, and old Symbian apps also. Main benefit is that they need only few dozen kilobytes for each app so enormous amount of software apps fits in modern phone s memory or in cheap PC. Cheap PCs can have same memory card extra memory like phones, and simple mass memory build using row of cheap memory cards inside. This kind of cheap PCs with low price can sell hundreads of millions of units in third world countries, but nobody has released those kind of cheap mini-PCs or laptops yet with double processors. Or even special mini-PCs for third world market. Cheap chinese x86 PC processor is what is needed, somehow chinese make server and phone x86 processors but not that what is needed most, cheap x86 PC processor, altough they have license (many manufacturers). Emulators for old 1980s software can provide in modern PC memory thousands of old Commodore 64 or Sinclair Spectrum or Amiga programs that now are freeware (abondonware). Instead of emulators simply making copies of old MOS 6500- 8500 and Motorola 6800 - 68000 processors or Zilog Z80 is possible, late 1970s - early 1980s processors can be made with roll printed electonics, transistors are made at few micron process now with roll to roll printing, other components perhaps even smaller. So 0,01 dollar copy of old early 1980s processor discontinued long time ago is possible, or complete old PC circuitry in one piece of roll printed plastic. So PC main processor does not need emulator because inside PC is roll printed real early 1980s processor or actually roll printed copy of VIC20, Commodore 64, 128, Atari ST, Amiga, Sinclair Spectrum, TRS 80, early Mac or even Apple 2. Early 1980s processors were several micron wide process made, this is about the same as roll printing technology today. Hybrid printing where modern cheap (ordinary semiconductor) memory is printed together with roll printed CPU components can be made, modern memory that is as slow as early 1980s costs almost nothing. Now cheap PC can use old 1980s games and programs without emulator, and cost is almost nothing. Also old game consoles can be build again as roll printed almost no-cost versions, and if PC has those also inside old console games can also be played in PC without emulator. Those computers and game consoles were build using few then available CPUs, mostly MOS and Motorola and Zilog. One roll printed CPU can then emulate several early 1980s computers and game consoles, if those computers and game consoles additional electronics is also build in roll printed piece of plastic together with CPU. Modern ways to make processor simple is “bespoke processor” (another version of inexact computing CPU), counterflow pipeline processor and TACO protocol processor for transport triggered architectures, and Mill processor (not released yet). Not only CPUs and old computers, but old DSPs even from early / late 1990s can be made with roll printing, those roll printed versions are almost as good as those original. Those DSPs were mainly used in music synths. Copies of early digital synths can also be made, simple roll printed plastic picture of Yamaha DX7, Casio etc. keyboard that has functionality of real keybard including touch keys (keys do not move because they are printed in plastic picture, but they respond to touch), so about one dollar price flat picture of old synth that works like that synth and produces sound when pictures of keys and buttons, dials etc are touched. Manufacturing millions (hundreads of millions) units of these is possible, it is just roll printed piece of plastic with roll printed electronics, so cost is almost nothing per unit. Old samplers of the 1980s that do not require analog components, and drum machines can also be roll printed now. DX7 with one dollar price and made of sheet of plastic will be hit product third world countries. Synclavier was all digital late 1970s / early 1980s? So most expensive musical keyboard ever (Synclavier) as one dollar plastic sheet version would be success. Or Korg M1 as roll printed plastic “playable picture”. Not only 1980s things but 4000- series or 7400 series integrated circuits can be made with roll printing. 4000 series was 10 micron or over in 1968 and 7400 series 16-20 micron wide manufacturing tech in 1965. So both can be made easily using roll printing (if they were digital, not analog). From them can be build for example Lunetta- style music synths at almost no cost. Also computers can use old Yamaha and Roland (MT32) sound chips roll printed new production as roll printed sound card. Not to mention old Crumar GDS, Atari AMY, Alpha Syntauri, Computer Music Melodean, Con Brio ADS, Strider DCS, Coupland synth, Resynator and Fairlight Voicetracker guitar synths, DMX-1000 Signal Processor Computer, Kurzweil K 250, Yamaha FX-1 synth. All these museum pieces can be recreated using roll printing one dollar price copy of electronics in plastic sheet if they were all digital. Some PPG synths (some were all digital), and Fairlight mark 3 was all digital also? Resynthesizers like Tecnos Acxel and Lyre FDSS that are old. Not only music tech but other useful tech can be roll printed, like MP3 players at 0,01 dollar manufacturing cost, wristwatches (roll printed watch with flexible plastic wristband) at 0,05-0,1 dollar cost, displays and screens (both black&white and color), electric thermometer at 0,01 dollar cost, solar cells, Nintendo Game&Watch style handheld game consoles at 0,01 dollar cost, electronic components of digiboxes old style 1990s early MPEG digital video, portable CD / DVD player electronic logic components. All these products could sell hundreads of millions of units in development countries (or billions of units). But still nobody is making them. Why? Roll printed electronics has developed to the stage that those products are possible now. When hundreads of millions of units of some product is sold it also earns money to its maker. Nobody seem to be interested in manufacturing roll printed electronics for third world market, altough it is there where largest markets for this kind of products are. Hybrid printing of conventional semiconductor memory (old tech RAM and ROM that is cheap and slow, or just cheap and slow RAM) with roll printed logic circuits in cheap products. Batteries can be roll printed also, like solar cells. Roope Raisamo: “Multimodal human-computer interaction” 1999. EmSynth miniature synth and other microcontroller based music synth projects are suitable for third world market also, but their price should be few dollars only if microcontroller costs just less than one dollar in bulk order, if those products are aimed at third world market. Other microcontroller cheap products are suitable for third world market also, not just printed electronics, microcontrollers are less than dollar and even DSP prices are just few dollars at cheapest, not to mention phone SoCs (C.H.I.P computer and Pine64) but those are PC projects. Roll printed electronic products can use modern cheap memory card as extra memory if needed. Yamaha DX7 sound chip was made at 4 micron manufacturing if I remember right (?), but I don t know if it was old or new model of DX7 (early or late 1980s), but roll printing of it is possible, like other DX series synths of 1980s or Yamaha sound card synth chips of same era. Using methods like “Fractional-delay inverse comb filters”, “Detection and lateralization of sinusoidal signal in presence of dichotic pink noise”, “Hyper-dimensional digital waveguide mesh”, “Normalized filtered correlation time-scale modification (NFC-TSM)”, “Fractional delay filter design based on truncated Lagrange interpolation”, Vector Phaseshaping Synthesis, Trendline synthesis, Discrete Summation Formulae (DSF) synthesis, “Fast automatic inharmonicity estimation algorithm”, Casio ZPI Zygotech Polynomial Interpolation, Impulse Based Substructuring / replication of impulse response frequency characteristics, spectral synthesis, Bezier curve sound synthesis, allpass filter chain synthesis, the geometrical oscillator, feedback based amplitude modulation, bandlimited waveforms using iterated polynomial interpolation, moving avarage vector quantization, n-gon wave synthesis, polynomial transition regions, and differentiated polynomial waveforms in signal processing in some electronic devices, not only sound synthesis but all kind of signal processing, may help. In article “Another internet is possible: a developing country perspective on universal internet access” 2015 is old worries of net neutrality and demands for “Universal data allowance”. Theory that everything in the internet can be proveded free to customer without payment for net connection is utopic and never gonna happen. So there must be separate net schemes for paying customers and then cost free internet. Worries about net neutrality are irrelevant, when there is no cost free internet in third world. If already now some kind of net line cost free internet in third world is possible without net neutrality and it can be reality today, it is better than no net connection at all. Only after cost free limited access to internet is reality in third world, 5 or 10 years after first no cost net line, is time to alter situation towards net neutrality, before that cost free net connection (no subscription line payments needed for net connection in telephones and tablet PCs) can be made even monopoly of some internet provider. The opportunity to have even limited cost free internet access is much more important than any net neutrality etc. question in third world countries. “Alliance for affordable internet”. Rich countries don t even need cost free internet, poor countries only. Windows is cost free now in cheap devices, or bargain priced to manufacturers. Cheapest Windows PC (Gole 1) costs 79 dollars. But Windows PC with x86 processor must be down to 20 dollar Linux mini PC prices. Costly x86 processor is now main point of “high” price, if chinese manufacturer who have x86 license would made x86 processor as cheap as cheapest ARM processors today 20 dollar Windows PC for third world market would be true. And so Windows could compete with Linux / Android in cheap devices (mini PC and tablet PC / laptops). Demands for net neurality do nothing but harm in third world, if it prevents cost free internet becoming reality. Has anybody ever asked africans, south americans and south asians what they think about net neutrality? If they can have cost free internet without it. In third world there are software / hardware firms like Champiansounds (Pakistan) which are homegrown music tech. “The COSA operating system” is about “reactive software”, new way for computer logic. VISC processor, Tachyum universal processor, and Fujitsu FPGA combinatorial optimisation problems solving system are new microprocessor techniques. In netpage Yehar (Olli Niemitalo) is “enumeration of functions”, and he has written to different netpages different DSP subject issues, in “music DSP” archive netpage, and in stackoverflow 2009 “8 bit samples to 16 bit”. If 32-16 bit integer audio is bit-width reduced to 8 bit or synth voice processed at 8 bit, then it can be expanded to 16 bit using technique represented at end of answers in “8 bit samples to 16 bit” text, perhaps with sparse Fourier transform or other fast processing. So 16 bit sound can be made 8 bit for storage and processing and expanded back to 16 bit in audio out - process. Similar ways was used in early 1990s in synth audio processing, Roland synths etc where 16 bit sound was actually 8 bit in internal storage. Also adding additional header to integer like some DACs use, and converting in the end integer to floating point (Wolfson Audio) increases accuracy artificially. Or record or make sound at 4 bit ADPCM or DDPCM (Dynamic DPCM, bitsnbites EU netpage) and convert that to 16 bit PCM in audio out.

There is also “world s smallest computer” done by IBM and university of Michigan, which both cost less than 0,1 dollars to make. IBM version has several hundreads of thousands or one million transistors, so it is more powerful and cheaper than usual microcontroller. Those cheap microcontroller-like chips can be used in cheap consumer electronics made for third world market. Super cheap electric thermometers, DVD and MP3 players, video glasses and and screens (TV and other), handheld game gonsoles etc. can include those chips. Also in telephone and computer circuit boards that now have microcontrollers for keyboard and screen scan etc. duties can use those microcontroller- like cheap chips instead of ordinary microcontrollers (if they have any advantage over microcontrollers of same price class). More computing power would be several chips / cores grouped in one. But if quad core ARM telephone SoC cost only 4 dollar and x86 Intel Atom quad core 5 dollar already few years ago cost benefits for several 0,1 -0,2 dollar microcontroller processors is not much compared to “real” and efficient telephone / tablet PC processor. But in computing that needs microcontroller - like speed and functionality those may be used. Intel Quark was something similar. There is also USB- stick sized mini-PCs and Android stick TV boxes. If telephone / tablet PC is just screen, battery and radio frequency comunication SoC inside, and actual processor is in USB stick or other modular replaceable type, when processor becomes old it can be replaced but telephone / Tablet PC remains the same. So telephone and tablet PC is not need to be bought every two year or so, only processor and perhaps memory also (new faster and bigger memory), if both CPU and memory is in USB stick or replaceable module. Memory can also be in the phone / tablet PC in non-replaced form. In third world market only CPU and perhaps memory is needed to be new in telephone and tablet PC, not all new device. Laptop PCs can also have replaceable CPU in module that can be taken away and replaced. In third world countries people have no money to buy all new telephone in every two year or so, so replaceable CPU and memory solves that problem. This replaced CPU can be included to very cheap tablet PCs and phones also, not just expensive devices, because in third world people have no money to buy expensive devices. There is “Endless mini PC”, cheap PC that is designed for low bitrate internet connection or net connection that has data rate restrictions, in third world. It has inbuild content so PC can be used offline without internet connection also. Similar design philosophy is suitable for third world internet devices. Really cheap PC would be microcontroller- style (and microcontroller priced) chip with Intel Quark, IBM RISC, ARM, or then Motorola, Zilog or MOS internal achitecture (such microcontrollers exist). Those microcontrollers can run 1980s / 1990s computer programs, so “PC” is 0,2 -0,3 dollar priced microcontroller that can run old PC / Apple / Amiga / Sinclair Spectrum programs and games. There is also CoreOS, which is cloud-based Linux distro? (or something similar, CoreOS is something new in operating systems and how they work). Microsoft EDGE E2 is new computer architecture, expanded RISC, made together with Qualcomm, but not going to production, altough it can offer 100-times more speed in processors, in newspaper The Register is article about it. Such things as Mill processor, VISC processor, EDGE processor, and the COSA operating system, and transputer (XMOS) can help to make roll printed processors more effective, they have now several microns manufacturing tech, which is hundread times less than ordinary silicon semiconductor. Because they are big and slow anything that makes them faster and more effective is welcomed. So if they cannot conquer established processor markets perhaps roll printed processor can use new architectures. Making processor simpler and transistor count smaller is possible using “bespoke processor” and “inexact computing” (Rice university) principles, without changing processor achitecture. But those principles work with new processor types also. When roll printed microcontroller etc. processors are being made. The John G. Savard text mentioned in earlier text (about floating point) is in Google groups msg comp arch “Re: beating posits at their own game” 11.8 2017 by Quadibloc, if it does not show in Google browser it is shown in Microsoft browser. Auli Keskinen: “Towards user empowerement - on development of utilisation of information and communications technology in decisionmaking of adminstrations” 1999. Jana discontinued its mCent app 2017, different ways to make internet cheaper like DataWind UbiSurfer and Jana / mCent should brought together and not trying to several separate projects for cost free internet. If “Spotify for subscription payment free internet” that offers cost free internet services (not music) is made, new blockchain based music streaming services can offer model for gathering money to support free internet. Musicoin, Maestro (from Korea), Choon, Voise, BitSong, Myceliaformusic, Resonate (from Iceland?), Peertracks, Bittunes, Ujo Music and Inmusik are all blockchain based music streaming and selling. Similar system, using for example Moeda as cryptocurrency can offer way to internet service in third world that is cost free to end user and advertisement paid where adverts cover user costs, using UbiSurfer or mCent or something cost effective and blockchain / cryptocurrency. Free intrenet / zeronet will offer music only if its freeware or public domain, because line costs take all advertising money, but if music is distributed “user centric royalty distribution” like “the ethical pool solution” can be used, and similar solutions for zeronet advertisers, if it helps to bring more advertisement base to zeronet. Webinos is forgotten project for web standard unification. Other EU financed are CORDIS projects RIFE, WiSFUL and UNIFY, where RIFE was specially for affordable internet, and iKaaS, another project. Can knowhow of these projects bring cost free internet in third world with other tech like Jana and UbiSurfer and data compression to maximum? If combining all these and other together makes zeronet in third world possible why it is not started yet? Xbox music / Groove offered unsurpassed royalties to music artists and was however free in 15 countries. New Zed (music player?) offers 11 times more money per stream compared to Spotify. This Microsoft Groove or Zed commercial system, not for music streaming but for free internet in third world, “netify” service where instead of music streaming “internet streaming” (surfing on the net) is cost free, at least in limited and lightened version of internet (zeronet) which is specially “build” for cheap as possible use. Because teleoperators charge extra for virus protection in their lines, cost free internet can be actually be normal-priced mobile phone line, because in India cheapest 3G or 4G line costs are 2 dollar about per month, with data rate restrictions. So “free internet” is normal cell phone line with voluntary 2 dollar cost for virus / malware protection. Those who not pay that 2 dollar get line for free. Large part of line users choose that 2 dollar protection, so teleoperator gets actually paid for “cost free internet”, Those who not pay 24 dollar / year have free line, without phone call option (they have to use voice over internet solutions), and they must use restricted version of internet, zeronet, where costly content is removed and data use has restrictions and internet content is “slimmed down” to become as cheap as possible to teleoperator. This is advertisement paid, like majority of that 2 dollar line cost, and teleoperator costs in every phone line nowdays. But user does not have to pay that 2 dollar price if he/she do not want. The 2 dollar / month loss is covered by slimming down internet costs to teleoperator (and removing phone calls completely). That is for third world only, because in there many people have no money to pay 24 dollar / year, but those who have money will pay that 2 dollar / month for normal cell phone line and malware protection. This is for first phase of free internet, if advertisement money is generated enough, more comprehensive cost free internet can be offered, but this “cost free internet” that is just normal cell phone line at different name and charged with voluntary “extra” malware protection that is actually normal line cost, would be first version of cost free internet in third world. There methods how to move money payments transferred to future, like “celebrity bond” style payments, that are for example studied as means to pay USA university students university costs. If some similar payment method in third world free intrenet, altough if not cost effective now, if those costs can be postponed to future using something similar arrangement, in future there is monetary base enough to make free net commercially profitable, and then those payments from earlier time when cost free internet to end user was not so profitable, can be paid. So if cost free internet is not profitable today costs for using it can be transferred to future when it becomes profitable, and cost free internet in third world can be started now. Spotify uses similar logic for music streaming, it has not made profit, but it is huge success because it is expected to make profit in the future. Similar “netify” service to zeronet / free internet for third world internet can be done, altough it some years make no profit it becames profitable in the future, and costs are transferred to future using similar payment models like celebrity bond etc. or something other model.

There has been news that information that Facebook etc. has been given to commercial use about their customers has been used in election propaganda etc. Could this be the key to make (limited) internet free in the third world? Instead of preventing that kind of usage in political means of internet user s behaviour data that Facebook has been charged, Facebook and other internet giants (Google etc.) could sell freely, and with good price, that information to political and other parties in third world. So instead of blocking this kind of commercial (or political) data usage, they could sell this information to those who pay most. And then money generated for this kind of political / commercial data sale (of internet user s behaviour) can be used to finance free internet in third world. That means that when in third world user uses cost free internet (cost free to end user), he or she accepts that any data collected from him/her can be used for example in means to manipulate his/hers political behaviour in elections etc. This kind of data seems to be valuable to political parties and other political forces. So those forces, if they pay enough, have access of all this kind of people s behavioral data that is collected using their internet usage. Of course any data collection method that can be used is in use, for example microphone is open all time in phone and collecting words and phrases that are valuable to commercial marketing / political control. That seems to be a way to George Orwell s 1984 control society, but point is that free internet must be in third world, and usage costs must be paid somehow, with advertsing money or some other money must be generated from using this free internet. So if poltical forces are ready to pay good price for this kind of “marketing study” (actually studying people s behaviour to attain political manipulation) and this money can make free internet in third world true, this kind of money gathering model can be used. So not only data from internet user s is sold to commerial usage like advertisers, this data can be sold to political forces too in third world. It can perhaps double the amount of money generated when compared to selling data to commercial usage only? Or generate money even moe? So if this kind of maximizing of money inflow is used, perhaps cost free (lmited) internet to end user can be opened in third world. There are many poltical parties in third wold that are willing to pay large amount of money for this kind of politically valueble data. So instead of blocking its usage like Facebook and Google have done, it can be used to generate money in third world for free internet line. Access to things like Moeda (United Nations sponsored cryptocurrency) and microloan systems like Mifos, OpenCBS and others like Musoni, Oradian, Mambu, Tenemos, Cloud Loaning Solution, Caft Silicon, Nucleus Software etc. are then open to all people in third world. Politacally interesting data can be for sale not only for political parties of that third world country, but for foreign forces also that are interested in to political influence in that third world country. For example Rusdsia has shown interest in Venezuelan politics, and China has interests in Africa etc. So instead of camouflage operations like Cambridge Analytics polically intersting information that Facebook, Google etc. collects can be sold freely, paid in cash, and that money can be used to make free internet in third world, complete with cheap free phone / tablet PC with cheap data gloves and cheap video glasses. Cheap 5-8 dollar factory price large order (billions of units?) phone / tablet PC plus very cheap video glasses and data gloves. For billions of people without internet connection. Because internet line cost in yearly basis or compared to many years of free internet line use, line cost is higher than those few dollar equipment costs per free internet customer, so this additional cost of phone, gloves and glasses is not so much. There can be auction in every year or 6 month period where politically interesting data is sold first to political parties of that third country, for highest bid, and then to foreign countries, again for highest bid, or simply to everyone who is interested. This auction can be country specific (third world country) or simple “everything must go” sale where every third world country data for political (manipulation) from all around the world is for sale at one “package”. Money generated with these auctions / sales can be then used to make cost free intrenet in third world true. Also information is available for commercial use like advertises at same time, sold to them. Combination of sale of information to commercial and political use can collect enough money for free internet connection in third world. Sale to political use can even surpass money generated for commercial use, and surpass the commercial information sale with many times (amount of money generated). Because information suitable for political manipulation can be valuable to political forces of the world, so information can be priced accordingly, not sold cheaply, and for highest bid. Is this a way to finally collect enough money to make internet free of cost to end user in third world? Later when commercial money makes completely possible to maintain free internet in third world this kind of political information auctions can be lessened or stopped completely. But in starting phase of free internet this politically motivated information sale can be a key to make free internet possible. Special cases when political information is needed immediately can also be used in “special sale” that is not in 6 or 12 month auction sale. Collecting politically interesting information about free internet users in third world and then selling this information to political forces is just market economy isn t it? Everything is for sale, including people. Capitalism rulez. But positive consequences of cost free internet in third world far outweight negative aspects like possible political manipulation through internet. There are so many things that people can do with internet in third world and internet would be tremendous help for them. Problem is that people there have no money pay even that small 2,5 dollar month (30 dollar a year) or more line cost that is in India cheapest net/phone line connection price. Line can be restricted to internet use only, without voice calls and text messages (in phone) altough cheap smartphone is the connecting device. Voice over internet solutions must be then used and sending text over internet. Even that internet connection can be restricted to save costs. But maximizing commercial potential of user data that can be gathered from user and used to commercial or political uses, in other words advertising and political manipulation, can cover that restricted internet line cost, so in third world (limited) free intrenet can be true. Of course no net neutrality is possible if free internet is offered, at least in beginning. Some years later after free net has been opened is possible to re-evaluate situation and add net neutrality or remove data sales for political usage. Another way to increase amount of money for cost free internet is if metasearch engines are used. They can gather information from 50 - 70 netpages (LIMMZ and Searx). Those pages are other search engines, browsers, and netpages. Often those metasearch engines block adverts. But instead of blocking them, metasearch engine should collect them all from 50 - 70 netpages and send them to free internet user. Now amount of advertisement money is increased to 50 - 70 times compared if user is using just one search engine / browser. Ads should be rendered to text only, no pictures or constant flow of adverts completely block phone screen. Now this huge increase in adverts can be used to fund cost free net in third world. There is Brave browser by Mozilla founder, it uses 20 search engines, uses its own cryptocurrency, and replaces advertiser s adverts of its own and pays advertisers 55% of advert income, pays browser user 15% for using browser (pay to surf model) and collects for itself 30% advertisement money that would otherwise go to advertiser. So this browser is ideal for third world cost free internet, user is paid for using (free) internet a litle sum. IOTA or Moeda are best cryptocurrencies for third world. Half central server / user and half distributed internet that uses distributed search engine / browser like “Dweb page” / IOTA and distributed filesystems at same time it uses central servers is perhaps perhaps for free net. Central servers have less load if part of their jobs is done by distributed serverless model in free net. LTE Direct can be used. Almost every significant browser in world uses Webkit as search engine base, so altough infornmation is collected from diffrenet search engines / browsers, allmost all of them use Webkit or its Google version as base search engine, so that makes free internet simple. Only Firefox uses different search engine base but Firefox for iOS is using Webkit. There was browser Bingle that showed both Google and Bing in same screen, and Favitt showed 6-7 search engines in one screen (split screen), so if they show their adverts too adverts shown simultaneusly is 6-7 times compared just one browser. But that split screen style is too much for phone s small screen so metasearch engine which collects information from 50 - 70 netpages and collects their adverts also and sends them to one user who had to watch 50-70 times more adverts than normally, adverts rendered to text without pictures because there is no room for advert pictures due to advert overflow. This can be the way to fund free internet in third world. Data compression can be used to maximum in free net to save cost. If MyPhotoZip image compression is used that compress still image to 1 / 1500 bitrate, that is really small, but because it is fractal compression it takes very long time. But even video compression using MyPhotoZip can be used, video is first encoded using nornal video compression, then “real” frames of video are compressed using MyPhotoZip, interpolating frames are created by video codec and don t need fractal compression. Also still image (photo) can use MyPhotoZip, first photo is processed using ordinary fast image codec, when picture is in phone s memory it is further compressed using MyPhotoZip, which takes time. Fractal compression is slow to encode but fast to decode so it can be even suitable for video codec. There are Ecosia style search engines that give part of their income to charity, altough they have limited user base. How much can such search engine that has billions of users donate to charity? If it uses money collecting methods mentioned above. Metasearch engine is called search aggregator. It can be used as adverts aggregator also, it collects adverts from 50-70 netpages and then sends them to user that uses this metasearch engine. Collected advertisement money is used to maintain cost free internet in third world. Pay to surf, pay per click and pay to click are ways how to advertsiers can get money. Wikipedia has articles about those. If advertisers pay 0,005 dollar per click (according to Wikipedia) and cost of net phone subscription line with internet is between 3,5 dollar to 0,5 dollar (UbiSurfer browser) in India, 100 to 700 clicks per month is needed to maintain free net connection. That was situation already in 1997. So if even every minute user must click advert and see / hear short advert, that is acceptable if cost free intrenet is then true. And other adverts roll in phone screen while user is using that one minute also. Why those advertisement techniques are not used to make possible cost free internet in third world? Earlier chinese manufacturers made cheap wristwatch phones, 1.3 - 2,5 dollar price, small screen, and Symbian OS in some. But still screen was big enough to watch video etc. Similar wristwatch phone, made at cheapest possible components, can be given away with free net connection, Big screen is not needed if also roll printed video glasses and cheap data gloves are given with phone, and roll printed solarcell charger or Peltier element charger that charges phone using body heat etc. Bill T. Gross has made many innovations how to use intrenet to collect advertisement money. Why those internet commercialisation techniques are not used to make free (limited) internet in third world? That has been possiblle since 1997, so already 22 years ago free net in third could have been true. What is the reason for this resistance to not offer free internet in third world? Billions of people have no money to pay for smartphone and after that pay 3,5 dollar per month (40 dollar per year) for internet / subscription line connection. However ultra cheap phone (manufacturing price in the range of one dollar, manufactured billions of units, rolll to roll printed components used if suitable) can be given free with free net connection, and net connection can be commercially paid, user “pays” when he/she clicks adverts or hear and see adverts etc. That was possible already 22 years ago. If user wants bigger screen he/she must pay money for larger than wristwatch- size feature phone that has also other extra features compared to that small wristwatch phone. “Tabletop PC” was early 2010s fad, screen size was up to 27 inches. Super cheap tabletop PC can be used as television, desktop PC alternative, game console etc., so similar cheap devices have propably large markets in third world. That 1 advert per minute is avarage time, if 30 minutes is needed user can watch 30 ads in row, like Youtube shows ads, 7 seconds and then user can skiop ad, so it is 3,5 minutes for wtching ads and then 30 minutes net time. Other ads are running on the screen when user uses net time, but user does not have to click them. There is free apps that are downloaded free, but are advertised, so they have advertising budget. There is millions of apps. So advertising free apps in free zeronet can bring more money to net provider than money from adverts of food, clothing and cheap consumer products to free net users in third world. Phone hat is given free with free net connection can have inbuild ads, so thos ads are not needed to transfer through intrenet. In phone s memory is ads that pop up in screen etc. Phone has inbuild “advertising campaign”. So net capacity is not so often used in ads. Net provider collects money from both inbuild ads in phone nad ads transferred through internet. Wifi phones are cheap and simply phones, so suitable for zeronet. Some free Wifi hotspots can be privately owned, in cafes etc, and not owned by zeronet provider, altough those private Wifi links offer zeronet. In countryside is plans to use cars, busses and motorcycles as moving intrenet links. Also sneakernet or “USB dead drop” can be used, but USB and other links that work as information loading stations, that use physical connection to phone, not wireless communication, are owned either by zeronet provider or state-owned. Those techiniques can ease zeronet provider s costs so that zeronet (limited internet) is possible. If data amount of data transferred is lessened, somewhere is that point where money collected from adverts and commercial sponsors meets cost of data transfer, and that meeting point is the starting point of cost free zeronet. It has been possible to do since early 2010s when chinese phone manufacturers begin offer super cheap phones so price of electronics is not problem. Neither is net connection cost if lessening amount of data transferred saves cost and there is point when advertisement money covers cost of data transferred. However cost free internet in third world has not been offered altough it is possible. Why? Using data compression can compensate low data rates, and even just text without pictures or sound can transfer valuable information.