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

Free internet for developing countries: more

For people in poor countries who have no money to pay teleoperator subscriber line costs for net connection (even if internet will be free of additional data rate cost), there are different projects like Piratebox, its landline version Maidsafe, FabFI in Afghanistan, and others that use “TV white space technology”, radiochannels that are not used by commercial operators, for total cost free internet in poor countries. All these are “do it yourself- internet” - type solutions to bring net to poor people in developing countries, who dont have money to spent on net connection. Also direct connection mobile ad hoc networks like B.A.T.M.A.N in Germany and Commotation Wireless in USA exists. Motorola IDEN and WiDEN phones can connect about 10 km distances, similar standard (OMA PoC) is in Europe but no phones for it is manufactured. Gotenna is low energy bluetooth solution which connects phones directly 100 km distances theoretically. All these are DIY solutions which offer alternative connections to existing commercial solutions. Also possibility of using skywave or other athmospheric propagationmethod (there are many of them, some are in several megaherz and even gigaherz range) that are previusly used by military communications about 50 years but not anymore because of satellite communication, can use for free long range net communication study. Long range Wifi together with athmospheric propagation is already under study. Most of these athmospheric skywave anomalities are near equator, where most of poor people live, in africa/ mediterrinean is specific skywave propagation etc. Also using long wave propagation like "groundwave" etc., with very small data rates. On other hand gigahertz athmospheric propagation exits even in high 3- 5 ghz frequency. Those athmospheric methods will be real multicasting, every message is heard almost everywhere over the planet. But all these free methods means bypassing existing internet (commmercial) front end structure. There is dual commercial/free solution (teleoperator Fon in Spain) offering WIFI either free or with payment, but it is not total cost free. Those worldwide multinational teleoperator firms that have largest amount of subscribers, are almost all state-owned or partially state-owned by some rich, or very rich (Norway, middle eastern countries) country. In fact telebisnes is most tightly under governmental control because of state-owned firms, perhaps even more than oil industry etc., than any bisnes in the world. Commercial TV - bisnes is free for charge for end customer, because advertisement money covers expenses of operation and also brings profit. Commercial radiobisnes is free of end user payments, even it relies heavily on expensive base (copyrighted music with its royalty payments). Both of them use radiowaves for transmission. However, internet is using radiowaves too, using mostly content that is free for teleoperator, it gets lots of advertisement money, and still charges money from end customer. Why ? Answer has been that landline costs are high, but in development countries most teleoperators have only mobile internet. But still those firms charge money even more than they charge from rich people in rich countries. In netpage" Price rating by country of internet (10 mbps, unlimited data, cable/adsl) (monthly)" one can see that in the poorest countries of the world internet connection is also most expensive in the world, dozen times more expensive than in rich country. And still state-owned (or partially state-owned) teleoperators have businesses in almost every poor country, under different names and disguises. Because state- owned firms are owned by citizens of each individual country, only political deciscion is needed to correct this situation. If teleoperator cant afford plain commercially paid net connection, solution in 3G and 4G networks (even 4G networks exist in development countries, even poorest ones, nowadays) will be slowering data rates per free customer, for example drastic reduction that drops data down to only 1/100th or 1/30th of commercial customer, teleoperator now gets for 100 times more advertsement money than for one commercial customer compared to data transmisson rate and bandwith/ channel usage. Of course net connection will be slow, but sophisticated data compression methods (libde265, future Daala for video, ATC Labs Immersive Sound Field Retention for sound, Intelligent Dictionary Based Encoding for text, Fox Unicode text character encoding, Finite State Encoding/ HPACK for general data encoding, My Photo Zip for still pictures etc.) would help. Also free connection would be only for net usage, no normal voice calls or SMS. However VOiP calls over internet and internet messaging services are allowed. There exist today one teleoperator that uses these methods already, Globalgig, but of course it doesnt offer free net connection. One probllem is that emergency calls are not possible in VOiP. so these must be allowed for VOiP calls. Also translation of different languages, both text and video files, and voice synth translation for people who cannot read is needed. Translations can use machine translation, or other (volunteers who do text and video file translation for free etc.). Many teleoperators use "zero rating", in south africa etc. They offer net content free with only subscriber line cost, and offer TV- series etc. material. One Tv- film consumes 300mb or more data and also royalty payments are high. Isnt it more sensible to offer free net without subscriber line cost without expensive content, offering only public domain, freeware material, and restricted data rate which multiplies advertisement money for teleoperator for badwith usage. Also customer base would have explosive growth when large number of poor people are now using the net. Some kind of data restriction (max monthly usage of gigabytes etc.) must be on free net. For emergency use also partion of next months usage is allowed for emergency call etc. important if one months data usage is already spent. In Ukraine continous 10 mbit/sec dta rate line costs 3.9 dollars per custome a month, and teleoperator is making profitable business. Ukraine is an industrial country. In some poor african or south american country line costs would be even cheaper. Extra advertisement money that covers 1 - 3 dollar customer subscription fee (among advertising money teleoperator already gets) is realistic goal for teleoperator and advertisers even in poor countrie. If operator gets that, total free line cost is reality even for pure commercial basis. In netpage “Worldstream.com: Avarage cost per click by coyntry around the world” is statistics how much advertisement money is generated by internet operators by contry. There are “Tier 1” worldwide internet connection wuthout data rate cost. The cost of sending data from other side of the globe to another is dropped so now cost is 0,005 % compared to year 1998. But teleoperator charges haven`t drop to 0,005% compared to year 1998. Why? Also advertisement money for net has grown exponentially so they have money on that way too. But internet line costs are almost same level as 1998. Why? Also using sort of closed file system like Virtual File System with reduced access to “real” internet (only public domain materia is allwed and freeware) is possibility to bring costs down for free net connection. This kind of “zeronet” would have minimum server and search engine usage, and thus low costs. For example Wikipedia can be scrutinized to 3,5 gigabytes without pictures and 5 giabytes with pictures. Whole english Wikipedia. If that database is in every cell phone tower link everytime Wikipedia is used message pass only between nearest cell phone tower and user, so use of teleoperators resources is minimum. Same can be implement on GNUhealth, Ekopedia ym. Public domain material like openculture.com -netpage books and education courses etc. can also be in cell tower link “near info package” in every link tower. Bringing costs down that way and same time slowering data rate to end user (so that teleoperator gets more advertising money per bandwith), total cost free internet can be profitable solution even for commercial operator. But state-owned firms could implement them first bacause they are under governmental control (either majority or minority ownership of stock shares). In Afghanistan there is a teleoperator whose majority ownership is in charity fund, so that would be one posiibility to start total cost free mobile internet in poor countries. But state-owned teleoperators around gloge operting in poor countries must start it too. Making political lobbying group worldwide that takes this initiative to public debate is needed and people must begin action.

In india newbuild 4G network user pays minimum 1,5 dollar a month for internet/ telephone line use, but this cheapest option has data restrictions (unlike in Ukraine where has unlimited 10 mbs/month for 3,9 dollars). If it is profitable business to offer in newbuild 4G network which has unpaid network building/capital costs 1,5 dollar per month service, for how much less money it is possible to operate old 3G or 2G network, when network building costs have paid already. And it is possible that this less than 1,5 or less than 1 dollar that is charged from customer is completeley paid py advertisement money so that for user net connection is free, in development countries. In Russia teleoperator Yota planned this kind of total cost free telephone line internet service in early 2010s, whose operating costs are paid by advertisements or other third party money, not charges for end user, but ended using ordinary subscription line costs. In some poor country where labour etc. everything is very cheap this kind of total free net service is more likely to happen than Russia, which is industrial country. In India elsewhere there are people who dont have money even for cheap 1,5 dollar / 20 dollars year internet/ telephone connection, because that money is needed for food etc., and they dont see that internet can be beneficiary for them. And those people who dont have extra money even a penny to such "trivial" as telephone and internet connection, are perhaps most of humanity, billions of people that dont have now and also in the future no extra money to pay for commercial internet, no matter how cheap commercial (money is charged from end user) internet connection is. One reason why they dont use it even they have money to cheap net connection, is also because they dont understand what opportunities net offers to them, but if net is offered for free, and the net that it is offered to them is organized the way Ecopedia offers its material (food, clothing, housing, education etc. matrial organized in logical way to topics9 that is for example directly possible to click icon that reads “food” or “education” or “health” directly in the devices screen, and information on these matters are opened this way. That kind of simple access and usability, and added translation (machine/ otherwise translation of english language material, voice synth for people who cannot read, added text/ voice synth for foreign language video material) will make people happen that internet can indeed help them it is not for "rich people only". The price of electronics is so cheap, Allwinner A33 system on a chip is priced at 4 dollars in factory order, and Intel says it sells Intel Atom quad core processors for less than 5 dollars to customer. Atom is an x86 chip so old/ new Linux freeware material is possible to transfer to telephone/ tablet PC devices. To bring costs down it is possible to use cheap magnetic mass media like micro SD card or USB memory stick or SSD card in every telephone link, so that when query is send that concerns some of the matters that are already in the divices screen as icon, that information is stored inside the 2G/3G/4G telephone link so query doest have to go any further, the message bounces back with nearest link with information. If micro SD card RAID is used that uses cheap micro SD mermory card as RAID station, for example 16 cheap chinese bulk produced 64 GB memory cards will make an 1 Terabyte RAID station. If prices for cheap bulk prodiction and large factory order is close to one dollar per 64 GB memory card, Terabyte mass memory in each telephone link is possible, and 150 SD card / 10 Terabyte in each large link tower. That mass memory could be Wikipedia (all Wikipedia can be compressed to 3,5 gigabytes, or to 5 gigabytes with pictures in "Wikipedia offline" - mode), health information, educational programs, and free e-books (perhaps million of them in data compressed text form using specialised text and text character compression). And all this in nearest telephone link. That kind of "zeronet" that uses minimal search engine and server resources and is almost cost free to use, and teleoperator still can send commercials and advertisements to end user, willw make total cost free internet (actually zeronet) true. Connection (limited restricted usage connection) from zeronet to "real" internet is through monitored and sensurized "ports" and "gateways", which makes filtering unwanted material easy. The cost of sending same amount of data from one way of the globe to another is down to 0,05 - 0,07 % compared to year 1998, if I read the statistics right this time. But data amounts that move across the net have increased also. However so is magnetic data storage capacity so information that 1998 was on the other side of the globe is now in some close server so teleopetars claim that data costs force them to charge end user high costs is not true. Telebusiness is among oil business and international weapons trade and (illegal) drug trafficing one of the words most profitable businesses. If teleoperators who are state owned and already making lots of money in rich countries but also have businesses in poor countries of the world, start offering total free net connection in those countries, and they still have advertisement and otherwise third party money generated in those countries, some way of free internet is already possible to begin immediately. Swedish Millicom is commercial firm that is specialised in developing countries only teleoperations, and is doing billion dollar business, so the amount of (advertisement etc.) money that is moving among teleoperations there is not small, even people if people in these countries does not have money to pay large sums of line fee.

Development countries have made their own useful telephone/ net apps like Daral project and mLouma, both in Senegal. One is for cattle health / trade and another is for wheat trade. Matibabu is an telephone health app that fights against malaria, and WinSenga is health app intented for pregnant women, both of them comes from Uganda, and so is mPedigree which is app for medicines. So people in development countries have already made internet applications that are close to their everyday life and designed for their needs, and are helpful for their health care and trading food products like meat and wheat. If people also would have free net connection they would have possibility to use these applications. They also would have an opportunity to make small amount of money using “Advertisement watching work”, a job that meants watching adwerts for hours a day and then clicking some questions about them or filling question form in the net, and then get some small amount of payment for just watching adverts. That would be also a great way to finance total free net connection. The money they have would perhaps go to a count for net shopping so money transfer to real cash that is expensive is not needed. This “bank account” they could use for bying products from cheap (chinese) net shops, making large factory orders (1000 to 10 000 pieces for large factory order). People from same geographical area can gather around in the net for a “buying pool” for one large product order and then get stuff directly from the (chinese) factory cheaply. When that stuff is transported to factory to buyers mathematical models that optimese “nearest path” can be used to ferrying these products to 1000 or 10 000 different customers, and these mathematical nearest path transport models can also be used to estimate traffic costs. Also Mifos X, which is a microloan platform but unlike typical western versions which transfer profit from microloans to western investors, Mifos X can be used by people of poor country themselves, those who have can little extra money can loan it to some other and get some profit too for loaning, but in this case the profit does not go western investor but stays in the country / area where loan giver lives. All this is possible in free internet connction and billions of people (5 billion or so) that does not have money to net connection are waiting. Among other massive amount of free information such as openculture.com and other netpages that internet offers, if only connection (subscription fee) would be free. It is technically and financially possible to offer some kind of free net / zeronet to poor countries. Something must be done, and the positive effects of free net are so vast and wide, and if its commercially possible, at least for state-owned teleoperators, then people must begin action that those teleoperators who are under partial or complete governmental control will start free net in poor countries. The freenet is one of most important questions of our time , and solution, even partial solution for no subscription payments line connection, would bring vast positive effects to billions of people, so this free net thing should be high on everybodys priority list, even those who are not political activists.

In Uganda there is Meka internet net shop / trade portal, “african Amazon”, and similar netsops are in almost every big african country. These trade portals are founded, owned and managed by africans themselves, so profit from these also stays in Africa and not go elsewhere (to rich country). In India there is MySmartprice etc. When these trade portals are owned by people from poor countries themselves, and their customers are also from those countries, these are most suitable shopping places for “zeronet” or cost free internet (among super-cheap direct factory order chinese netshops). And about claims that totally commercially (advertisement) paid teleoperation is not suitable for poor countries because people in them does not have money to bough things: in India government had to limit telemarketing (SMS message advertising) by special law, because SMS marketing was increased to level that it need special law to regulate it, SMS adverts filled phones. India is one of the poorest countries in the world. And yet special law had to be implemented because there was too much (!) advertising through teleoperators. In development countries, when free net / zeronet comes reality, most of advertisng and trade on the net will be food and clothes, less consuner goods like electronics. And if hardware is bought these would propably be useful tools like “multimachine open source machine tool” which is part of Open Source Ecologys (OSE) Global Village Construction Set. And 3D- printers that chinese netshops sell 100 - 150 dollar price at cheapest. When free (but limited) access to internet (or its close relative zeronet) becames possible, simple direct on device (phone) icon screen that are divided like Ekopedia`s 12 different parts (birth, education, food, lodging, clothing, arts and craft, thinking, living together, health and care, travel, play, communication). Or like Wiser.org, or Appropedia. All three of these can be compressed and added together with Wikipedia in “near info pack” in every link tower for “zeronet”. Ekopedia is now built to western user, but version of it (and Appropedia and wiser.org), whose material is towards development countries and their everyday life is suitable for free net. In net there are pure information based public domain / freeware portals (Europeana / Athena Plus, OpenGLAM) and information and entertainment together portals (openculture.com, Open Knowledge, ibiblio, Public Domain Review). Direct access to them (once again straight from pressing an icon on phone screen) would be good.

There are models for “zero rating” that offers internet free of data charge, but not perhaps free of subscription line fee. And then “toll free model”, totally free of cost. In South Africa there is Project Isizwe that aoffers 250 megabytes per day free Wifi internet to poor areas of South Africa and schools and institutes. In Bangladesh there is Grameenphone / Telenoir model that after watching one advert one can have 20 megabytes per day toll free internet. Data rates are small that only text based browser like WAP is reasonable, and cheapest feature phones dont have video processor anyway. Also very efficient data compression methods are needed, perhaps straight SoC (System on a chip) implemention, if so slow data rates are going to be standard in toll free internet. But in Ukraine has has 10 megabits/sec continous data eate for only 3,9 dollar monthly fee. That kind of continous data rate model would be suitable also for a toll free internet / internet line connection. Toll free model is offered by commercial apps and are limited to maximum of about 40 in Indias Airtel Zero. About half of worlds internet providers offer at least one toll free app, and that one is Facebook. In India there is frim Gigato whose whole business idea is to offer toll free internet connection using toll free apps. So toll free internet does not bring financial losses, but profit and new business oppurtunities in development countries. Telenor (Telenor app pass), and Opera (Opera web pass) and Mozilla are offering this kind of toll free net through commercial apps in developing countries. But access is limetited to 10 - 40 web portals only, and altough Wikipedia and Facebook and Google mail is included, this kind of commercial apps only model leaves out vast amount of public domain and freeware material that net has full of. If also together with commercial apps some kind of "best of public domain / freeware netpages / portals" are offered to toll free net user, free nets usability increases significantly. Openculture.com is one example that uses that kind “best of public domain material” already. And access to e-education, e-health, e-governance etc. pages must be added to toll free internet among commercial apps, perhaps through simple “press this icon on the phone screen” model. Text / voice translation of 92 different languages are required for text and video material, coarse machine translation even for millions of public domain / freeware netbooks. And perhaps even open source CMU Sphinx based server / phone personal digital assistant, Sirius in server and Pocketsphinx in phone working together as unit, but perhaps only more expensive devoces / phones have this capability, devoces for toll free net connection should becheapest available. Toll free apps are offered from Freeband Technologies, Zidisha / Twilio.org etc. 220 billion dollars are advertisers using each year in developing / poor countries for advertising, but only 10% of money goes to internet advertising, simply because people does not have an internet connection so advertising in that media is not used. If toll free internet is being offered large scale to everybody in those countries and using model that offers those public domain portals, then this 220 billion dollars moves from newspaper / outdoor advertising to toll free internet and began cumulative effect that brings toll free internet even more profitable option for business, and brings 5 billion people more custoner base to advertisers. But those apps like Facebook neet large megabytes of memory to work, so simplified version of Facebook and other commercial apps are needed for ultra-cheap feature phones, Diaspora or other. Or using technologies / routing like ZigZag/Xanadu that Tent protocol already uses, and ZIgZag protocol is being developed in different versions like “Gzigzag platform for internet experimets”. Xanadu / ZigZag also has (experimental) implementions like “Token Word” copyright payment model, “Tinderbox” experimental interface, and sentences.com database model. DBpedia is semantic version of Wikipedia etc. But now only limited app based toll free internet is offered that does not give access to millions of diffrent public domain e-health, e-education, e-books, public domain films, music etc. material that is offered in internet without royalty payments that teleoperator must pay for copyright owner. There are demonds that all internet should be toll free in poor countries, with its roylty payments to copyright owners included. But it is not perhaps possible now, perhaps possible never. But when these materiall that teleoperator must pay is excluded from payment free material. even Spotify, Youtube etc. is possible in toll free internet in poor countries. Perhaps first step is “best of free netpages / portals material” selection that can be stored compressed even in each cell phone link tower in zeronet form, that is offered together with toll free commercial apps.

If CMU Sphinx based personal digital assistant is used (like Sirius and Pocketsphinx), or other digital assistant type system, commercial or non-commercial, that opens totally new prospects to education. This digital assistant can be used as “digital teacher” or “automated machine teacher” to people in areas who cannot have education or chance to get education is rare, and improve education beyond local teachers abilities. These "digital teachers" that are found in cell phone / cheap tablet PC could teach people to read, write and calculate, and teach subjects that are found in normal schools. If those 5 billion people that are without net connection could have it free, education is the most important thing they can benefit fro toll free net. How sophisticated system this digital automated teacher is, or how sophisticated it could be at those resources that is available for toll free internet. This "teacher" can be used in other digital intelligent assistant roles in toll free net also. Having electrical power to phones can be provided using solar panel as charger, and solar panel charger can be also used for bringing electricity to small LED lamp, that can be one colour (yellow or red) if white LED is expensive, to bring light at night. Lamp canbe connected to PC or cell phone battery in its primitive form, small LED lamp with very small lumen number / lighting ability is plucket by electric wire direct to phone or tablet PC and is using its battery. Solar panel charger can have extra battery for phone / tablet PC that can be used for LED lamp also. Solar panel technology has increased in recent years, cheap solar panels that are also effective are in existence today, but they are not in mass production. Mass production of cheap solar panels should begin so that these can be used for free net phone chargers and other uses that solar panels are needed in developing countres. In net shops, particularly chinese netshops there are "scammers" that sell things that dont exist. Because of that toll free internet or zeronet must be monitored and sensurized, and system which scans this net net and filters unappropiate material out can me computer program etc., or real person who accepts for example large netshop factory orders before they are send to seller. Instead of memory card printed paper which has Colour Construct Code, or Xerox Dataglyph or Microglyph can be used as “mass memory”. Phone must then have a camera that can scan scanning code. CCC code needs special plastic reflective surface on top of the paper in order to work. Besides CCC code or Dataglypf / Microglyph versions standard QR or other scanninc code can be used, but in colour form that multiplies the information capacity. But now memory card capacity has increased and price dropped to level that even ordinary paper ark cant compete with price compared to cheap mass bulk production SD memory cards. But if there is a situation that memory cards are not in existence paper ark with sophisticated scanning code like CCC or Microglyph can store megabytes and even gigabyte (?) of information in A4 form, if paper ark size increases also amount information increases. There are two implementations of Xanadu / ZigZag routing philosophy that I know, Tent protocol, and CommnityWiki in its german peer to peer version. Other real life applications of Xanadu project I dont know. Xanadu type protocol is already in use in smart phone ad hoc networks like Firchat or Commotation Wireless, but smart phone Wifi range to other phone is only 20 to 30 metres. There are home made antennas that are made of tin can etc. to increase Wifi direct connectivity range. And also Motorola iDEN and WiDEN phones that has 10 kilometres range about, and Gotenna with even larger range. But all of these are do it yourself - internet type solutions in a time when toll free intrenet even in commercial basis is becoming true. but these techiques can be used to lighten toll free internet / zeronet server - bandwith burden etc. together with toll free internet which has contact to “real” cell phone network / mobile internet.

Another way to electronic devices to have electricity instead of using solar panel is charger generator that uses human power. There are electric bicycles that have electric motor that can also be used as a charger generator for bicycle batteries, such as “Sanyo Eneloop electric bike”, ElectroPortal Charger bike, FlyKly Smart Wheel, Hammacher Schlemmer electric bike etc. If this kind of self-charging electric bike is offered to people of development countries, and with it they can have electricity for electronic devices and light (LED lamps), no changing batteries are needed when current ran out in one battery. Cheapest electric bicycles, chinese made start about 100 dollars for direct factory order (mone than 1 piece order is needed for factory orders), and 80 dollars is lowest factory price Ive seen on chinese electric bicycle, or ordinary basic bicycle with electric motor and battery, like these cheapest models are. There is Siva Cycle USB electric device (phone) bicycle charger, model Siva Atom is cheapest. There is "UpCycle Ecocharger DIY bicycle generator kit" for home electricity. These kind of methods for providing electrical power are needed in poor countries, less in industrial countries where electricity is everywhere available. There is also Elf velomobile half bicycle half electric car. Chinese factories produce electric tricycles / transports like "electric rikshaws", cheapest are 200 dollar Baoge BG-6007 transport tricycle and Baoge BG-P006 (480 dollars) which looks like electric car / riksha (factory order prices). Also "Tyrone TMK-08" conversion kit to turn riksha to electric power at 80 dollar exist. If these products would have ability turn pedal power to energy and use electric motor as generator, these vehicles could be used not only transport vehicles but as moving electric generators also. 4 wheel all terrain vehicles (ATV) are in fact at same price or even cheaper in china than aforementioned but those use gasoline engine and gasoline costs money. Electric power could be used not only for electronics but for cheap LED lamps also bringing light at night. Simple LED, perhaps yellow or green colour if white LEDs are expensive (these lamps must be as chap as possible) with plastic mirror surface paper thin reflector and a wire long enough that it can be raised to ceiling in a small hut. About electronics: feature phones are becoming cheap, 4 dollars is cheapest factory price in China (no video playback, WAP browser is "optional" perhaps for extra price. Cheapest Ive seen is couple a year ago when one factory sold “wristwatch phone” which looked like plastic phone on a wrist, small black and white low pixel count screen, no touchscreen, small minimal keyboard on the phone, no internal memory. The price was 1,35 dollars / piece for large factory order (1000 - 5000 pieces or even more). It had Symbian OS, and WAP browser was optional. So it was words cheapest Symbian "smartphone". It has now been discontinued. But if feature phones can be so simple and cheap, it is even possible that those ultra-cheap entry level phones they can be offered for free to customers that are using toll free internet or zeronet. The advertisers that are paying for toll free apps can also pay the minimal price of 1,5 - 4 dollars of "entry level" super cheap phone with WAP, advertisers / toll free app providers already pay together more than one dollar a month per one customer for toll free internet apps. So Toll free internet user dont have to pay even for a phone, altough these “feature phones” have minimal capabilities. But when the user have money they can then buy better phone for internet connection. WAP is still used in poor countries because cheap feature phones lack mali graphics and video processor. One solution to bring costs down would be using one “master device” (smartphone or tablet) that belongs to the head of the family and other family members would have simple printed electronics made PDA - style (personal digital assistant, or device which has no internet / radiowave connection but simply memory and screen) device. Printed electronics is new and untested technology, but electronics can be manufactured almost at no cost per device price. These ultracheap “assistant devices” simply connect directly to smartphone, through usb port or otherwise, or simply take SD memory card from smartphone, and inbuilt is only screen and battery (built on printed electronics also) and sort of simple MP4 player with simplified audio / video compression (like TICO for video), and conversion from high efficient audio / video compression to simple and less efficient compression on printed electronics devices must be done in the smartphone. When education is spread using toll free internet, right to use cost free net could be connected to education. That means that toll free internet user must go through some educational course in order to contue use of cost free internet. Like learn to read and write etc. some kind of educational course. So education can be tied up with cost free internnet, and make education efficient, user now has motivation to learn (otherwise free internet connection will cut off, at least for a short time, and user could then later start the course at point where he / she left out and continue education / free net using again). If in education is used “personal intelligent assistant” as “automated machine teacher” learning results would improve significantly. In china chaep “Android stick” TV- sticks are sold at 7 - 8 dollar factoruy price cheapest. Android TV stick is an Android computer with wifi- or other connection. Even american manufacturer is now releasing 10 dollar “minicomputer board”. These have no screen or keyboard. If cheap chinese video glasses like Elekworld EGS or VGA 260 (Elekworld production already discontinued, but other chinese factories make it with almost same designation) are connected with Android stick or mini PC board no screen is needed. The device can have simple phone type simplified cheap keyboard, and all information is seen on the video glasses. Elekworld 260 clones are sold at a only few dollar price at chinese netshops (large factory orders). So this about 10 dollar videoglass / Android computer combination is cheap and offer better way to watch videos than phone screen, so poor people don`t need to buy television anymore at developing countries. If the device itself have display it would be simple black and white LCD icons only perhaps, only basic functions can be seen on device display, all other information goes to video glasses.

Another substitute for television would be something like “Google cardboard virtual reality kit”. That one is made from cardboard but chinese factories already made plastic versions of it. Simple plastic DIY self assembled kit for cheap feature phone, no stereo “3D” screen is even needed, main point is that screen is closer to eyes than just arm length away, if screen is small (like cheap feature phones often do). That plastic kit can be send along with cheap feature phone, and the kit costs almost nothing, but however can offer something like “TV watching experience”. Real virtual video glasses need optics in the glass, but if the principle is just 2D screen not 3D it probably don´t need optics, and Google cardboard kit doesnt need it anyway. If Allwinner A33 SoC (system on chip) costs 4 dollars to phone manufacturer, how much costs a phone built around it? Cheapest available chinese phone batteries, Cheapest OEM built phone with small (1,8 inch is smallest) screen, bad quality color screen, no touchscreen at all perhaps. Cheapest feature phones have only 16 - 64 megabytes of ROM and RAM, so small memory is a problem. Android first version had minimum theoretical requirements 200 mhz processor, 32 MB RAM and 32MB external memory. Later versions of Android could run on ARMv5 core and even Android 4 in some versions could operate at ARMv6. Brillo is new OS for internet of things that needs 32 - 64 MB memory, and is using Google Weave and Google OnHub routing protocols, and is Android in small form. Another solution is that part of the operatimg system (core, or more) is in this very limited phone internal memory and rest of the OS is in SD memory card, if it doesnt fit in the phones memory. Android, Chrome OS, Firefox OS and now Brillo OS use (almost) same Linux core. Chrome OS and Firefox OS are cloud based systems, and need very little memory for initial core, altough their memory reqirements are quite high (both are based on newer versions of Android libraries etc.). Early feature phones used cloud based browsers like Opera Mini in its first versions and Skyfire (now owned by Opera and discontinued) with browser based data compression. It is still used in mobile browsers like DataWind Ubisurfer in India & elsewhere as a method for offering one year toll free internet, and chinese mobile browsers use cloud browser method also. Efficient browser based compression like Ubisurfer and Skyfire is needed in free internet / zeronet. If Android could be scrutinized to such a small memory amount that are typical to cheapest feature phones, or split in phone / memory card combination, or phone / cloud browser combination (Chrome OS and Firefox OS are already using it) that enables sophisticated OS using in cheap feature phones with low internal memory. Combining together Chrome OS and Firefox OS, who use different search engines, is perhaps possible in same phone. Older Nokia Symbian phones used Nokia browser that was half between Firefox, using perhaps partly its Gecko search engine, and half Webkit. There are also modifications of Firefox that use Webkit. Or perhaps is best to use new innovations in browser technology like "iSpaces Cloud Browser / Computer", which is patented technology. Anyway, how much would phone cost using A33 SoC, using cheapest available components and cheapest OEM manufacturers? If the factory price is 7 - 8 dollars, now imagine that when toll free net / zeronet begins the advertisers and commercial sponsors offer this phone free to everyone 5 billion people that are today without net connection, because they dont have a money to buy it. 5 billion x 7 - 8 dollars is 35 - 40 billion dollars. Each year 220 billion dollars is spent in development / poor countries in advertising. Only 10% of this sum is spent on internet marketing and advertising, because 85 - 90% population is without net connection. If net connection is available to them free, advertisng money is now spent 90% to net and 10% otherwise. That means 200 billion dollars per each year. For one year 200 billion minus 35 - 40 billion is 165 - 160 billion still vailable to free net operating costs to teleoperators, So this 5 billion commercially sponsored free phones to free net usage is in fact win-win situation and both the advertisers and teleoperators (increased advertisement money) have profit. Altough only 20 billion dollars, about one tenth of total advertisement money is today spent on internet marketing, it is enough even today to launch toll free apps and other commercial cost free internet applications like Airtel Zero. What is the case if toll free internet is provided to all in poor countries with free phones and commercial (advertisement) money now increases 10 times as large as it today at development countries internet. At 200 billion dollars sophisticated net services can be provided in every poor country for free. But nowadays advertsers are forced to use this sum to newspaper and street marketing because people does not have internet connection. Other than plain advertisements, such as adware (computer games that are in fact advertisements, soap opera style “webisodes” or “mobisodes” that are in fact long advertisements etc.) new advertisng techniques like “shoppable video” and “timesync” in video stream etc. If people of poor countries have access to NeoBux and bux.to “watch adverts and make money” type websites that would be another win-win situation, poor people make money only for watching adverts and advertisers make money too. Also access to Samasource -style microwork sites would help people in poor countries. Because the toll free net opens enormous business opportunities in development countries (and rich countries too, when their advertisers and marketing people get their hands on to 5 billion people that are now without net connection) more quickly the toll free net / zeronet opens to everyone the better. And besides information free net could offer entertainment like openculture.com and mubi.com public domain film collections, altough these films are free and public domain in free net they would have “commercial breaks” like commercial television, in order to get money for operating toll free net. Also Webinos, the new unified standard for net could bring solution to different operating systems dilemma, but in toll free net it must perhaps be in simplified form which doesn`t need much memory space.

Symbian is from the start built as minimal operating system, usingng as small amount of memory and power usage of the phone as possible. So when small memory and small power usage OS for cheap feature phone is needed Symbian, altough its aged, is an option. It has 100 000 apps too. When Android is using Alien davlik to operate Symbian apps, isnt it more sensible in very cheap phones use Symbian with Alien Davlik to run Android apps? There are sophisticated new OS:es, like Xcerion XIOS (Xcerion Internet OS), which is cloud based but also remain functinality in offline, unlike Google Chrome which needs always net connection to operate, so XIOS is half cloud half ordinary on the hardware OS. It has some multiuser innovations too. Another swedish innovation is Enea LINX routing protocol. But cloud based OS:es consume net bandwith, so when small amount of bandwith is available at toll free net perhaps cloud OS.es are not suitable for it, if OS is too large to fit in small memory of the phone perhaps putting large part of OS to SD memory card instead of large internal phone memory or cloud etc. is a solution. Old and new small Linux distros like Maemo (Harmattan, Meego) or X-Pud Linux or Slitaz Linux (which earlier had its own minimal browser Slitaz) are also an option. Linux-Tiny or TinyLinux is Linux distro that has scrutinized Linux to 296 kilobytes minimum, altough its functionality is in its smallest form is doubtful as OS for feature phone etc. There are peer to peer solutions like french infinit.io -messaging protocol and Radeeus music streaming service, both using new patented technologies. For peer to peer net television / video straming there are many solutions, some open source some private business, and australian GT TV peer to peer internet TV is using new half- peer to peer half- centralized model for streaming video. Using part peer to peer type structure to reduce costs is perhaps possible in toll free net. And when free net user wants to buy a phone, something like Google project Ara, Puzzlephone and Vsenn modular phones are good for poor people, changing ageing part like processor but still use others in the without need to change whole phone and buy another one. Those modular phones are now designed for people in rich countries (like Vsenn modular phone), but it would be more useful in poor countries that poor people have modular phones, laptops and tablet PC:s. And old laptops and tablet PC:s thrown away in rich countries can have usage in development countries if only cost of using net would be free. For example a couple a years ago expensive Symbian phones are now discarded only because they have Symbian operating system. These and tablet PC:s and laptobs that are perhaps old fashioned in rich countries but still useful in development countries could have useful new usage if net line costs in development countries doesnt prevent that like they now do. And using Puzzlephones “Puzzlecluster” solution old parts of old electronics could be put in good use in developing countries, when user changes his modular phone processor etc. to new one and puts old in “Puzzlecluster”.

If it is possible to build very cheap feature phone, non-modular (because it must be as cheap as possible), and give it for free at toll free net users (because manufacturing costs are being paid by commercial sponsors and advertisers, the cost of the phone per one customer is much less than available advertisement money per one customer, even in development countries, and the cost of the phone is in fact cheaper than toll free internet is for advertisers & other commercial sponsors who are using toll free phone apps to reach poor people, if a whole years costs for toll free internet are counted, and if the cost of the phone is compared to many years toll free cost, phone that is given without cost to toll free net user is even more affordable to commercial sponsors and marketing & advertisement purposes), that free phone can come with cheap free accessoriers. Like cheap one colour LED lamp with thin plastic membrane mirror-like lampshade which is wrapped in conical shape when it is taken out of package. Lamp can be blue, green or yellow colour, or yellow colour low pressure sodium lamp, or other cheap lamp. LED lamps are being manufacured at very low price in chine, like other stuff. If that small low lumen lamp uses phones or solar panels battery (solar panel, cheapest as possible new technology mass production cheap solar panel) can come as free with phone too. And cheapest headphones, price is about 0,1 dollars for cheapest mass production large order headphones. And plastic “google virtual reality kit”, but in this case it only 2D and only for purpose of bringing telephone nearer to eye, not exactly in order to offer 3D experience. These cheap extras that come with phone for free (cheapest possible available are chosen for lamp, solar panel, phone head kit and headphones) would`nt rise cost of phone too much, some (like headphones and plastic head kit) are almost no cost, some like LED lamps are very cheap as chinese mass production, and electric wire for lamp, new advances at solar panel technology are becoming turn solar panels also as a cheap mass production products.

Teleopertaors have more money to spent than advertisers because they get about 4 - 5 times more income from teleoperator line user than advertisers, and they get all of the profit that telephone line operation brings to them. So it is at teleoperators interest to find new customers for their telecommunication lines. Therefore it is more likely that it is teleoperators who give free ultra-cheap feature phones to customers in order to widen their customer base, and these cheap feature phones are paid and sponsored by teleoperator companies, not advertisers. Giving away a very cheap feature phone with internet connection to apeople in developming countries is efficient way to increase customer base for commercial teleoperator firms if 85 - 90% in poor countries are without internet (toll free internet / zeronet) connection. When those commercially sponsored ultra-cheap free phones have given to the people their operating costs can be paid by commercial advertisers or line costs will be shared between teleoperator and marketing & advertising money. When customer base expands 10 times more large than it is today at developing countries, teleoperators can accept slight fiancial loss which is acceptable because they now have 10 times more customers, who maybe in the future change from restricted toll free internet / zeronet to “real” internet and start to pay subscription line payments to teleoperator. For that chance it is commercially reasonable that teleoperators accept a financial loss perhaps for a first few years of free toll free internet operation, because growth in advertisement money that goes in internet marketing increases each year and finally cover all telephone line cost that teleoperator firm must pay. So it is sensible commercial solution for teleoperator to give a free feature phone and telecommunication line to a people in poor countries, because it will substantially increase teleoperators customer base and in the future will pay itself so in the future teleoperators don`t get financial loss of a toll free line. If the line costs are shared in some way between advertisers and teleoperator companies (perhaps 1 -2 years when toll free internet opens) and after that line operation costs are responsibility of commercial advertisers completely, or shared between teleoperator firm and advertisers completely, because teleoperator firm can expect that part of those free net users (perhaps most part) will at some day be paying customers at “real” internet.

For feature phone or smartphone one feature that uses most of phones resources is Graphical User Interface or GUI. If the GUI is changed to some simply form like Google Brillo has done with Android GUI, rest of processor`s resources can be saved to real operating system use. So if phone is done with cheapest components available and cheapest memory capacity, in order to save memory perhaps very simply GUI is needed so that memory can be saved, and rest of the operating system remains operational even if the memory capacity is low. For ultra-low cost feature phones.

There is also a american internet firm Jana, with its mCent internet app, which makes possible for people in poor countries to use internet without line costs in poor countries. It not only makes toll free internet possible, but even gives extra internet time credit which is possible to spend on other internet use that requires payment. It has a some sort of own toll-free “app store” too. Jana / mCent has been developed in MIT Media Lab.

There are so called ecological search engines that give almost all profit they get from their net service to enviromental purposes. They can be found for example “10 search engines to help the envoroment” and “5 best search engines to search green” lists. There are others like Ecosia, Beluga, Ecoogler, Ecogine etc. too. For example Ecosia has about 100 000 users, but it still manages to donate 80% of all income that it has (advertising money) to enviromental purposes. If it is possible to build and maintain search engine that even for such a small amount of users (less than 100 000 search engine users) is so profitable that it can donate 80 - 90% of its income to other uses, such a simple search engine (opposite to Google which runs billon-dollar business with its supercomputers etc.) could be used in zeronet / toll free internet to reduce its search engine / server costs. And if a search engine can donate 80 -90% of its income to other purposes, in this case it`s for maintaining a toll free internet / zeronet. And if toll-free internet has about 5 billion users that means that same kind of cheap search engine system can donate about 99 - 95% of its total income to maintaining zeronet infrastructure. If a distributed search engine is used all 100% of income from search engine will be donated to zeronet, but distributed search engines has usability etc. issues so it is unclear can those be used in toll free net, altough some like Dooble (in its first, distributed form) are very sophisticated.

And other use for regular cheap feature phone: a musical instrument. There is a text " Design of a scalable-polyphony MIDI synthesizer for a low cost DSP" by Antti Huovilainen. He uses VLSI Solutions VS 1003 chip. Another design in netpage instructables.com " VS 1103 MIDI synthesizer" uses similar chip, but design is more complicated. Those chips are ordinary MP3 player logic circuits that can be found in every cell phone, even in cheapest models. Using only one (?) extra filter Huovilainens design achieves to build hardware synth on a MP3 player circuit. Because design is so simple and ads very little to existing circuitry those kind of hardware synths for music making can be added very simply to cheap feature phone. Huovilainen writes that perhaps ADPCM-modulation that brings reduction in bitrate is suitable for the design, and in netpage bitsnbites.eu is a "High quality DPCM" or Dynamic DPCM, simple and efficient modulation technique for sound that perhaps down to 3 bits (8 numerical values) or less, is suitable for synthetised sound. Yamaha manufactures sound chips for phones, like does french factory called Dream. Yamaha makes vocaloid sound chips and vocaloid technology is developed by Yamaha. Vocaloid could be used as good quality speech synth in phones for those who cant read or write. Unfortunately only most comprehensive and expensive sound chip comes with vocaloid. If Yamaha could offer vocaloid also in cheapest sound chip that can be installed in cheap feature phone that good quality speech synth can be put to good use. Vocaloid is used to imitate singing voice but can be used as speech synth too. And Yamahas sound chips can be used as musical isntruments (music hardware synth) also. Music synths are being offered in Android but they are software synths. If synth is in hardware, processor and memory usage can be saved. If only slight alteretation in phones MP3 player is needed to turn it as a hardware synth, or a cheap sound chip with vocaloid technology added to cheap feature phone, both good quality speech synth and music synth can be added to phones functions, almost without extra costs. The sound chips in phones are normally used as ringtones etc. but they can be used as musical instruments (multiple polyphony voices musical synth) etc.

Altough Android is de facto standard, different web apps must be developed in different platforms anyway. And for the same platform there are multiple bookmarking tools etc. that are unnecessary as only one is needed etc. To overcome this different methods that have remained in theoretical level has been developed, Google Web Intents, and Mozilla Web Actions & Web Activities. There is Phonegap Build for app development and distribution, proposed Webinos standard, and Docker, for unified web framework. Those unified structures instead of multiple formats that are now blooming in internet may offer simplicity and reduce bandwith, if data is going to be transferred cheaply.

The so-called internet of things has many different standards that have date rates up to several megabits / second and some even 100 Mb/sec “burst rates” for peak performance. Also ranges vary from metres to over 10 km (most IoT standards achieve this 10 km or more range, several kilometres is normal at IoT). 10 km is normal GSM range in non-rural enviroment from link to phone. So IoT achieves GSM- class range, but is optimized for cheap data transmission (cost for teleoperaor is minimal per IoT device than cost to connect telephone, tablet PC etc.), and cheap chipsets. Data rates also are similiar to older GSM standards. Longest range (50km) is in Sigfox IoT standard but data rates are also lowest (1 Kb/ second). Standards such as 6Lowpan / LoRaWAN, new Narrow Band IoT, LTE-M, NB-LTE, NB-LT-M, and EC-GSM rel.13 (from 2016 onwards), are coming, and NB-IoT is the chosen standard for IoT devices. But which is the best of these myriad IoT standards for toll free net communication for development countries? Bluetooth is in its long range low energy / Bluetooth Smart version capable of 100 km or more range (Gotenna). That toll-free telecommunication should allow people in poor countries have cost free net connection with some usable bandwith but without cost to end user. If range and data rates that allow people to have almost normal telecommunication like nowadays low end GSM phone standards or GSM standards about 10 years ago, and internet connection but only fraction of cost that would allow advertsement paid free net in developing countries. About 20 years a go there was less than 20 kilobits/sec avarage GSM data rate but still for example streaming media with sound and video picture together was possible, using that 20 years old technology. Sound codecs optimised for low bitrates like QDesign Audio codec which had minimum 5 kilobits/sec for usable sound (music) and video codecs optimised for low bitrates. And that was already 20 years ago, at least video codecs have been developed further what they were then. New audio codecs such as MPEG 4 Structured Audio, MPEG 7, or Seer Systems Seer Music codec (1998) are very sophisticated and require much of hardware, such as own music synth directly on the receiving device (phone etc.), but they compress musical content down to a few kilobit/sec level.Standard telephone audio codec AMR is also best audio codec for music at very low bitrates (about 20 kilobits/sec) if AMR Wide Band is used, studies of low-bitrate audio codecs have shown that. Also a new Linux distribution Solu OS is aimed for multiuser (shared) net connection but is also same time a cloud-based OS which enables it to be in cheap devices. Its shared nature is perhaps suitable for mobile ad hoc networks etc. but at same time its centralized cloud based nature requires central server, which is opposite to distributed networks. And Dooble browser in its original form was distributed search engine / browser also, but is not anymore. If any technology like this makes net connection to low costs these can be used in toll free net in development countries.

And printed electronics, which is becoming suitable for internet of things ceap devices, RFID technology etc, anything that allows simple electronics for IoT, is also possible in toll free net devices if they use same data transmission standards and chipsets. So using IoT technology for data transmission cheapens data transmission costs, and also cheapens cost of electronics (close to zero per device if printed electronics is used). But at the same time ordinary telecommunication techniques are making toll free net possible also, because advertisement money has increased each year etc. Also ordinary electronics is today so cheap, only few dollars for a feature phone chipset system-on- a chip. Altough toll free net with its restricted data rates etc. is not perhaps suitable in industrial countries, in developing countries it has great demand, using IoT or other cheap data technology.

The cost of electronics is not anymore manufacturing costs, they are minimal for integrated circuit now, but the cost of the manufacturing licence per device. Electronics are full of many licence fees, and that makes the cost of a device, not the cost of manufacturing. There are open source electronic projects, but to use them some kind of licence and royalty payment, if its gonna be used in commercial purposes, exists and must be paid. Examples can be found in opencircuits.com, open-electronics.com, opencores.org and openfpga.org -pages. There are system on a chip open source projects, and radio frequency component design and other like DSP and DDS (direct digital synthesis) for open source versions of radio frequency components. There is a complete phone (Tuxphone/Openmoko) that is open source. But question is not is it open source, but is it cheap or free from licence/royalty payments. The SuperH (Hitachi) processor core is in public domain and is now called J-2, J-3 or J-4, and is licence free, and has attanained a lot of interest, one processor costs only 0.03 dollars to manufacture, so for super-cheap phones/devices it is a candidate. Consider this to 5 - 10 dollar price of commercial processor so 0,03 dollars is huge improvement. Chip foundry takes its own profit however anyway, so price would be higher than 0,03 dollars for processor. The “thumb instruction set” for cheap ARM- processor is similar to SuperH. When suitable electronic components are not found, and device needs lots of different components, it often requires FPGA (field programmable gate array implementation), but they are closed source and expensive. Some small FPGA companies offer open source for their design but they are not licence payment free. There is Papilio board -project and Cubic Board for more or less open FPGA. Also Archipelago project for independent FPGA that is not build by commercial company but open source community exists. Building a super cheap device requires almost no-cost licence/royalty payments. There is netpage “Softcores for FPGAs: the open source alternatives”. So public domain electronics (electronic circuit design has a 20 year time before it enters public domain?) are needed or something with very small licence costs like Beyond Semi`s ARM processor with one payment / licence payment free model. Device should be built using these licence free / minimal license fee components for manufacturing complete super cheap feature phone or tablet PC, or using licence free / almost license free FPGA in the SoC if suitable components are not found as licence free / minimal licence fee versions. This will make one dollar feature phone / tablet PC for developing countries and for toll free internet / zeronet.

One way to exploit old electronic “legacy” integrated circuit designs of 20 years or more old, is for example in the case of CPU grouping several dozens them in a grid in one silicon chip. J2 (Super H) CPU is manufactured at 180 nanometer process, but if 64 (8x8) of them are grouped in a grid with connections between, in one 22 - 28 nanometer process silicon chip that compensates the low transistor count and slowness of old design, and silicon chip die area is about the same. Making chips smaller increases speed and reduces electric consumption, 20 year old designs have low transistor count but combining many old processors together in one multi-processor chip compensates this. These 20 years old processor designs are in public domain (?). So no licenses must be paid. Smaller 4x4 (16 processor) and larger than 64 processor grids in one monolithic silicon chip can be also used. Also when licences are elimininated other ways to cheapen end product can be used. For example if is possible to manufacture batteries and displays using printed electronics (prototypes are already being made) the rest of the phone / tablet PC electronics could be in typical silicon chip SoC with processor and radio frequency electronics, but battery and display is made by printed electronics. Display could be small, low resolution low quality, perhaps black and white or two-colour system like old colour negative film prints (instead of three colour). Even magnetic memory can be manufactured using printed electronics, altough only prototypes have been made. Super H (J2, J3, J4) is the processor that has caused interest, but russian 512 bit VLIW processor Elbrus has been manufactured since 1986, and it can run Linux and Windows. When Elbrus 2000 appeared it was claimed to be faster than Intel Itanium, but at very low transistor count ( I remember 90 million transistors ? was the claim), and chip was manufactured already old fashioned manufacturing process. Elbrus team worked with Sun systems and latel Intel, and design owes something that is common to SunSPARC, Transmeta Crusoe, and Intel Xeon. SPARC is open source and Elbrus team was developing it also. Only other VLIW processor in manufacture is Fujitsu FR-V, from 1999 but based on 1980s design, and FR-V is now discontinued (?) and replaced by ARM processor version. VLIW technique makes very efficient processor possible, enormous amount of theoretical research work have been done in decades but only those two processor have made it to mass production. Elbrus 2000 is from 2005 but basic design is from 1986 and in between these years some improvoments in basic design were made. So minimal transistor count but still fast and effective processor is VLIWs advantage. New and perhaps better conceps than VLIW have appeared but these two are old (20 years or older) designs and were in mass production. Cheap phone can be accompanied by solar panel, made by printed electrincs and charger for phone. If it is not possible to build feature phone without licence fee for electronics, perhaps then it is possible to have lower license fee for such a cheap “charity phone” that is paid by advertisement money like toll free net connection that comes with the phone, and phone is distrubuted without cost to people in developing countries who have not yet internet connection. Because phone is so cheap it can be given for free and paid by adverstisng money. But lower licence costs for electronic component or license free components are required. ARM Holdings Plc. has Linaro concept and collaboration for open source, perhaps that is the way to obtain cheaper ARM licenses for free “charity phones” with free internet connection. Any cheap technology will do, ARM processor, Intel Atom or even Quark, operation system Symbian (early version) or Maemo Linux or simplified Android like Google Brillo or Android 1 (only 32 + 32 MB memory) and simplified GUI. Internet of things internet protocol etc. that is data efficent and fast for transmission etc. If cheap licenses for processors are not obtained, then any FPGA, CPLD, ASIC will do as processor or SoC of the phone / tablet that is license free or minimal license.