The Hottest Phones for the Next Billion Users Aren’t Smartphones
wsj.com
wsj.com
Here are some notes I wrote to myself at the time:
- there seem to be fewer apps on the Kai store than on the Banana Hackers store, which seems odd, as all of those apps in the BH store should be useful to KaiOS users
- The official docs are kind of unclear, e.g. the API docs include some options about storage, but don't tell me whether localStorage is available. If I'm a web developer then I probably want to use the storage API with which I'm already familiar.
- another doc issue: the documentation for d-pad navigation and emulated cursor is very confusing. I can't tell whether they are two different ways of using the arrow keys, or whether they serve two different purposes. This means I need to do a bunch of experimentation to see what happens
- someone on Reddit mentioned they were waiting for app store approval, and they had waited 6 weeks so far; seems hard to get developers that way?
From the article, it seems like pre-installed apps have ample usage, no mention of installation post hardware purchase.
For that reason, I'd imagine 90% of Jiophone users in India would install at least one third party app.
However, they would get this from the JioStore, not from the KaiOS store: https://faq.whatsapp.com/en/kaios/26000183/?category=5245235
So if the majority of KaiOS users are using carrier-branded phones with carrier-branded app stores, maybe the developer story for 'vanilla' KaiOS isn't as important.
These phones will have the apps from the factory, and nothing more.
NB I'm not literally recommending Gopher but a modern day equivalent.
in fact this promise was once made by steven jobs, except that the phones at the time were not capable enough.
the question is not, what's the point of using web tech for underpowered phones, but the question is why do you expect me to cater to a minority market, and when do we reach the point that any device is powerful enough to handle web tech?
because that is where the future is going. native development is a stop gap to handle high performance requirements, but make no mistake, its days are counted. web development is the future that will replace all app development, whether we like it or not. even on the desktop, thanks to webassembly the browser is the future development target on any device out there. native apps will become a niche for only very specialized applications, and servers.
Niches have a lot of money in them, to the point that entire industries blossom just to cater to them. Things like high-end personal audio and performance automobiles are both multi-billion industries.
i have yet to hear about any apps that became popular because they performed better than others. in the whole market, that's just a blip.
things like high-end personal audio and performance automobiles are a high-end market market. you are asking me to create a high-performance app for a low-end market. the dynamics here are entirely different. and the profit margin is too.
well preforming apps for underpowered phones is a minority market.
i wouldn't even know how to promote my app to you.
my app is significantly faster than the competitors, and it even runs well on your cheap semi-smart phone
but does it have feature X?
sorry no, because that would make it slow
how much does it cost?
$20
uh, why did you think i bought a cheap phone?
apps for this market will not be sold on performance, but on solving problems that are specific to this particular group of people.
I want fast and feature X and I've no problem with $20 if it's better than the free one.
We were talking about different markets :)
- writing native apps to gain a bit of performance for a minority of users is not worth the added cost of development
- performance is not a selection criteria for most users
- low end phone users are unlikely to spend money on expensive apps
- catering to low end phones only makes sense for apps that are useful to that market. those apps have to be cheap, and that's yet another reason not to make them native.
and i was responding to a comment how it doesn't make sense to use web technology on low end phones.
my counter argument: it does not (no longer) make sense to write native apps. period.
you then responded that you would pay for better performing apps, but that puts you into a different niche demographic than the entire topic of the article.
now one point of the article is that the demographics will change and that there is going to be a huge market of users with low-end phones.
however that still doesn't make performance a selection criteria for people buying apps unless the phones are not able to run any of the popular apps which appears to be the case because kaiOS didn't even support native apps last i checked. in other words, apps on those phones are competing in an entirely different league, and the existing android market does not apply.
then again we have an entirely different dynamic of competing in a new market where there are no existing contenders, and the main question is, is the market big enough yet to make it worth it.
and since those phones run kaiOS native apps are off the table anyways
Where do you get the idea that anyone expects you to do anything? Since when are hackers opposed to more options, especially those outside the walled gardens of modern smartphones? If it's worth it, do it. If it isn't, don't. If the system sucks, it'll fail. But why rail against the idea?
I'd dare say it's already happened on the desktop. I do IT for an elementary school, and pretty much none of the employees use any app other than their web browser. Right now the computer I'm sitting in front of is running Outlook, our ticketing system, WhatsApp, Discord, and Spotify all within Firefox which is the only native app I have open, and likely the only app I'll open all day.
i do realize that we are reinventing everything here. and as some said elsewhere, html is not the best of tools to create user interfaces. not by a long shot[1]. but the dream of a single unified architecture that we can all develop towards may at last become a reality.
[1] actually, thanks to things like canvas or webassembly we are not necessarily stuck with html. we'll be able to create gui toolkits in javascript (hey, js-framework developers, here is your next target: the js-gui-toolkit of the week ;-) or maybe even port existing toolkits to webassembly and then our gnome or windows apps will simply run inside the browser instead.
the day of the linux desktop (for those that are still waiting for it) will be the day of the browser desktop
Seems kind of fun in a way, the bloated JS frameworks we are used to probably won't run very well, so you would need to work at a lower level and really optimize.
However, when our sales guys talk to customers (SMBs), they are universally pleased that we don't have a native app for phones. None of their employees want to install an app just to use the software on their phones. The fact that our product has a responsive design and is sufficiently performant on android and iOS based phones is more than enough for them.
Gotta know your market.
I concede css make styling far easier than any native toolkit.
And is true that most UI toolkit are so unfriendly that make html+js look soooo good in contrast.
But native development have been far simpler, productive and easier for solo developers before: Delphi, HyperCard, Smalltalk, Visual FoxPro, Visual Basic prove it.
My experience is that most developers I know (in latin america) prefer 1000% of the time (maybe not exagerating by much) to code in that tools than html.
Html/js is dreaded. Is like "damm... I MUST do html and js because... what choice I have?".
The huge indictment of our industry is that it bet in a subpar tech (html/js) and have forgotten to invest on UI toolkits.
Look how much take Apple to get swift UI. Decades!
Decades!
And it have hypercard before!!
I have a sense that the web stack makes up a large majority of developers who can/do develop UI-centric user applications.
Yeah. I bought two FirefoxOS phones and was keen. Unfortunately it never got to critical mass. Actually mailed the last of them to someone from India I met on HN. Recently I saw a French guy in HK on my LinkedIn managing some new effort under KaiOS. I was all motivated to meet and share some ideas (used to work in mobile, most handset manufacturers ~2009-2010) but found out the company is owned by that evil mafiosi Ambani, the same guy who bulldozed thousands of people's homes in Mumbai to build an uninhabited luxury condo, and who is destroying Australia's Queensland/Barrier Reef region with a massive coal mine. Crook.
Do you have source for this claim?
> who is destroying Australia's Queensland/Barrier Reef region
That's Adani, not Ambani.
I'm no fan of crony capitalism, but such false claims on HN aren't cool.
Source:
> KaiOS is a web-based mobile operating system that enables a new category of smart feature phones. It is forked from B2G (Boot to Gecko), a successor of the discontinued Firefox OS. KaiOS brings support of 4G/LTE, GPS, and Wi-Fi, as well as HTML5-based apps and longer battery life, to non-touch devices. It has an optimized user interface for smart feature phones, needs little memory, and consumes less energy than other operating systems. It also comes with the KaiStore, which enables users to download applications in categories like social media, games, navigation, and streaming entertainment.
Just ordered a pack of µC the other day for under 1$ per unit. These come with powerful computation cores, floating point units, analog-digital converters, countless integrated serial interfaces... You don't even get a coke for that price. Completely insane world...
Smartphone production is just the same:
First the people are enslaved to get the resources out of some mines, then some Asian workers (inc. children etc.) have to assemble (996 without a loan you'd want to live with). Then it's transported via logistics which is another sector that refuses to treat people like human beings. In the end of the line we have the eager sales guys who knowingly lie to customers and themselves to get a lot of sales.
The difference is that companies earn a lot more on a single sale for a smartphone but the scale of the sales of these feature phones could end up being more revenue I think. On the other hand marketing costs could be much lower because the price plus features alone would beat every smartphone at once.
I think it's crazy we still didn't modularize these devices and have nice OS'es for them.
- C.S. Lewis
Everyone who isn't ignoring this based on a decision to do so knows about the practices under which people have to suffer every day to sustain our nice lifestyle here I think.
The problem is the way the whole industry works because new products need to be sold continuously for a company to survive.
We all know climate change is here but it will take us decades to react globally.
Of course they turn their heads - it's much easier that way to enjoy all the money they got for acting like this.
https://www.globalslaveryindex.org/2018/findings/country-stu...
http://news.trust.org/item/20180918134734-kqc8m
https://www.amnesty.org/en/latest/news/2017/11/industry-gian...
There is much more if you want to research further. I'm astonished you didn't know that.
To realize this led me to buy used computers and phones instead of new ones among other stuff I try to improve in my personal/professional radius.
Integrated circuits are tiny specks of material, usually encased in comparatively huge plastic rectangles which are still only a few mm on each side.
The cost is almost all capital investment. After years of having a working design and process, the marginal cost of each chip is negligible. Most of what you actually pay for is probably validation and transportation.
What's really expensive is getting an application running on one with the reliability that people expect of a consumer electronic device. Yes, modern Cortex-M cores can do just about everything, but it takes some expertise and/or time to get them to do anything in particular.
Anyways, how much do you pay for a few hundred milligrams of acetaminophen? Is that also completely insane?
Consumers are way too entitled. Look at aeronautic applications where it just has to work for a few hours until you have to reboot everything.
But anyway, even if development costs amortize with giant quantities, it still requires complex machinery and a trained workforce for manufacturing.
It is evidently possible, but the result is still amazing.
Development for the chips is costly, but in many applications the same principle seems to apply.
Porting to another processor can happen if a chip becomes available that is a few cent cheaper. I have often seen that, even though porting isn't trivial and in many cases pretty error prone in my opinion. (compiler change, specs just completely different...)
> Anyways, how much do you pay for a few hundred milligrams of acetaminophen? Is that also completely insane?
Honestly it seems that it is. I got the impression that something about the price is wrong. That is why I often have the thought that pharma might rip off their patients and chip manufacturers concentrate on exploitation of workers instead. Pharma might not have this option to that degree because counterfeiting is easier to accomplish in that domain.
Not that I want to complain about cheap and powerful hardware too loudly...
They are most probably subsidized, though not sure how much of it is considering they offer 14 GB for 1.50$ a month.
It has quite good battery time days to weeks. The phone is very sturdy.
Drawbacks, cannot purchase train tickets online, cannot use online bank.
Drawback camera.
One needs to use the brain to navigate so one forces to know the way how to navigate. Its a bit hard if you get stuck in the middle of nowhere but then the phone has a navigation map app.
Smartphone zero.
Most folks I know don't really use calling anymore. If one of my friends called me, I'd probably pick up because I assumed it was an emergency. If they just called to talk, I'd tell them to text me.
YMMV here but younger crowds seem to shun voice calling IMO.
calling you on the phone means that my time is more important than yours and you had better be available when i want you to.
that said, pure social calls are fine, they are less disruptive than needing to discuss something that is not urgent or even important. if it's anything i need to remember, or think about before responding then i really prefer messages.
The internet is 50 years old this year. I'm sorry to break this to you but you are no longer young.
Prior to that, the only way to get access to the internet (D/ARPANET) was (generally) thru educational or military connections; one of the few exceptions to this (I don't understand exactly how it worked) was being a member of The Well, which allowed some kind of internet access (not sure if it was limited to email, or if other access was allowed?).
The number of people on the internet prior to (roughly) 1992 were mostly confined to the United States, and was a relatively small percentage of the total population of the country, simply because of the exclusivity of the connections; it was open to commercial traffic. Prior to 1992, most people who had experienced some from of "online connection" did so via either multiuser private commercial networks (two of the largest in the USA were Tymnet and Telenet), or large "walled garden" shared systems like AOL. Then of course there were the myriad local and long-distance Bulletin Board Systems (BBS) - which toward the "end" (I put that in quotes because they never really fully "died") in the early-1990s started to standardize on form of "email" called FidoNet (IIRC, even the original developer thought of it as a hot mess of code that barely worked).
All of this is to say that for most people (but not all - certainly more than a few people here on HN likely had access to the internet prior to 1992) the internet is only about 30 years old.
Though you remain correct in the assessment that "we are no longer young"...sigh.
Yes but I'm specifically talking about someone who's an "early adopter"!
It is strangely hard to find information about all the research networks or competing international commercial BBSes from then, but two names I can find are the Association for Progressive Communications and CYCLADES.
The phone forces me to call people more often as it is tedious to text on its interface which is a good thing.
Before smartphones, prior 2006 people opted for cheaper SMS prepaid plans over costly Voice rates(dedicated channels & limited resources). Our choices are based on cheaper rates of communication, now we've unlimited voice calls as a norm.
WhatApp'ing(or similar on smartphones) is richer than MMS'ing. SMSes are losing out, only limited to VAS-App & OTP code notifications.
The boundary is fuzzy, sure, but the term usually means a phone where the app usage is front and center and phone usage is secondary.
Now these new phones might be somewhere in-between.
The only reference it has to how they're used talks about app usage "Now he listens to Bollywood music on the job, using Google’s built in voice assistant to search for Hindi-language tunes on YouTube. At night his family crowds around the device to watch movies."
They just lacked CPU power and RAM capacity compared to what came when the iPhone launched.
Models would be:
6600 6630 6680 N70 N73
I actually had another Symbian powered Smartphone when the iPhone launched, the Motorola Rizr Z8 which made the Nokia powered Symbian's at the time look really slow. I sort of miss that phone and it's form-factor, it even came with a pair of Stereo Bluetooth Headphones. Massively ahead of it's time, of course not as slick as the iPhone and AirPod combo I have now, but great in it's own way.
Those Symbian S60 Nokia phones you listed could install and run apps written in "C++". Emphasis on those quotes! But regardless, your very own native code, doing your bidding. Been there, done that.
It worked, but writing those apps was very tedious and hard. Everything was async to save power. No standard libraries, etc.
I'm very happy there's no need for "leave stack" and "cleanup stack" etc. in the modern world.
But damn, that architecture was really helping the battery to last for a long time. Your code simply would not be running until some async thing did a callback.
>Feature phone is a term typically used to describe a class of mobile phones that are still technically otherwise smartphones, besides their lack of highly advanced hardware and capabilities of modern ones. Feature phones tend to use an embedded operating system or real-time operating system with graphical user interface which are small and simple, unlike large and complex general-purpose mobile operating systems like Android or iOS, they typically provide voice calling and text messaging functionality in addition to basic multimedia and Internet capabilities and other services offered by the user's wireless service provider.
After that, the iPhone was clearly something qualitatively different, so people (or rather, journalists, phone stores and the like) started calling the old style “feature phones”.
I’d say it’s now misleading and even incorrect to call something a “smartphone” if it doesn’t have a full-size touchscreen.
All deserve to be called smart phones.
You don't misrecall. Even phones without a full keyboard like Nokia's N-series (e.g. the very popular N-70 https://en.wikipedia.org/wiki/Nokia_N70) were considered smartphones.
If you're looking for a word to describe phones whose interface consists of a touchscreen with icons and gestures, I'd suggest "phablet".
When the iPhone launched, people called it a smartphone even though you couldn't install apps. Multitasking was also years away.
"Smartphone" has never meant anything really. Just marketing.
By your definition the original iPhone (3.5” screen) was a phablet.
It was, of course, absolutely bloody awful and didn't have even enough RAM to do the most basic things. But it was 3G and I had an unlimited plan that I tried to make the most of by watching videos etc on the go. I even used to use it as an MP3 player.
Exactly, "feature phones" were called that because of how the cell phone companies had to try to sell previous-generation phones in the same market as the iPhone. When other cell phone companies saw the iPhone, they realized the only way to present their phones as competitive was to frame the iPhone as a bundle of features (music player, web browser, games) and say hey, that's not so special, our phones do that stuff too. So you had flip or candy-bar phones that had an app for playing music, a terrible, crippled mobile web browser, maybe email, maybe a weather app, and some built-in games.
Calling these phones feature phones is a pretty awful knock, in other words. It means they aren't comparable to smart phones.
"Feature phones" were at one point the high end of a carrier's phone lineup, as they offered features, such as music playback, a camera, or WAP internet access, that a basic phone did not. Usually they could only run very limited applications, written in things like Java ME, that were available for sale through your cell carrier. When the iPhone was released, feature phones became the new low end. (Pre-iPhone smartphones were quite rare, usually being carried by on the go execs and the like.)
It did not make a big splash even among Linux enthusiasts, and vanished tracelessly as Android took over.
Only now, the type of a device you can call "an Internet pager" is truly finding its use again.
The irony is as the Internet becoming near ubiquitous, the less is the marginal value in having the latest and greatest device for using it. Most of what people use them now are IMs and digital wallets.
A lot of my coworkers are now settling with something even dumber for phone use, and use 7-8 inch tablet as the main mobile device: https://www.aliexpress.com/item/32824383761.html?spm=a2g0o.p...
Or some go for things like this, a bluetooth remote for the main device that can access your address book and act as handsfree: https://www.aliexpress.com/item/32839686717.html?spm=a2g0o.p...
Over the past decade, the term smartphone typically refers to a general purpose computing device that fits in your pocket that also acts as a phone. There is also more recently an implied requirement of a touch screen. A feature phone is more of an appliance device, where the primary feature is a phone, but it also has other limited features, some of which are internet enabled.
Unlike a smartphone, where you have a lot of control over what software you can put on it (similar to a PC), a feature phone only has a limited set of preinstalled features and optionally has a limited app store to add some more limited features.
By those definitions, the phones in the article are more feature phone than smartphone.
I used to develop apps and games on feature phones using Java MIDP, so 15 ago phones were already smartphones, there was just no app ecosystem back then nor real economy, that's all.
But it was possible to install 3rd party apps and even browsers like Opera Mobile on flip phones.
The difference is in the CPU power. Today some phones are more powerful than some cheap laptops.
Some of those devices (such as the Jio phone shown in the article) can be jailbroken to support a lot more applications.
[1]: https://sites.google.com/view/bananahackers/home [2]: https://twitter.com/BananaHackers
- Google Authenticator
- Password Safe
(- E-Mail would be nice, though)
Actually I find appeal in only being able to check messages (except for urgent messages via SMS) when I'm home at my computer. But I need a device for the 2nd factor and an auxiliary organizer for passwords.
I think the idea is supposed to be that internet access, even of such a limited form, I supposed to inspire at least some people to upgrade, up-skill, want more for themselves and their families, etc.
i don't want to upgrade, and i expect that someone who does want that will be better served with a cheap laptop alongside that phone than a more expensive phone.
of course that makes me one of the nay-sayers who is rejecting the smartphone revolution. but i can dream, can't i?
But I meant to frame my previous comment in the context of the article, where it's talking about those who can't yet afford a smartphone but probably want one.
no matter which though, speculating from the sidelines isn't going to help.
unfortunately i fear that a big drive to more expensive phones is because they are being used as a status symbol.
but that too is a sideline comment and is not nice to the people we talk about
Sorry. Article on KaiOS came up a few months back. This is a rewrite.
This guy sells mangoes on the street and can afford a data plan that lets him stream music all day and watch movies on his cell phone in the evening.
This is unthinkable in Canada (except for those socialists in Sask.).
I of course know it's not about the developers only but the practices of the markets, delivering hundreds of ad trackers on a single page etc. But it's still for me on some level hard to accept the fact that a single web page delivering a few kilobytes of some news needs more than 1 GB of RAM.
Of course you can: run a virtual machine, give it e.g. 500 MB, don't allow the guest OS to have the swap file, install there your favorite browsers and try (and weep).
It will be hard and unpleasant for you, but that's exactly what the users with less money are faced with. To ease your initial pain (or make the installation of the browsers or the OS even possible?), you can first install the OS and the browsers with more virtual RAM on the guest and only then reduce it.
And of course, you'd have to clear the browser cache every time before you try your pages. Also ideally, limit the network speed and increase the request latency!
Also to get the "realistic" CPU speed, run that all on some less powered CPU, like: https://www.amazon.com/Intel-Compute-Computer-Processor-BOXS... with Atom processor.
The question is more the will and not the technical aspect. And explaining all that to the managers, if you are in a company which doesn't have awareness of the potential users around the world.
To give you an idea: I can read still HN site on the underpowered devices, but there are enough sites today which just don't load or crash (i.e. I can't read the linked articles). It's that bad.
that virtual machine does not compare at all to the actual performance on a phone where the app runs besides other apps and native services, deals with different latencies for device access, internet speed etc.
yes, i can test my app that way, and it will probably good to do so to catch the most blatant performance issues. but then i could just test on my own phone for the same effect. in no way though i am going to be able to test across a multitude of devices, all of which have different performance characteristics.
If I may guess, your phone has 2 GB RAM minimum and at least 4 core CPU and therefore is much stronger than what I talk about? Like I’ve said, using Apple iPad 2 (which probably are easy to find and have 500 MB RAM) is a good start. You can’t dismiss those as being not polished enough. They are just specced less powerful than the current first world phones.
My VM suggestion is that the tests can be done on every notebook. If your page fails on that setup it’s obviously not suited for the less powered devices.
A bit of Google-fu shows that a lot of these phones run something called KaiOS, and it looks like native apps are actually build using HTML/CSS/JS anyway - https://developer.kaiostech.com/
A basic HTML document with CSS styling will adapt beautifully to any screen and requires very little CPU or memory. Write once, view on every device with a browser. Publishing content is also part of the original vision of the web, and luckily HTML forms are also accessible to any device.
Yes, you want to write apps, and perhaps that requires more from a device, and maybe you don't want to cater to poor Indians. But lots of sites are sending down 3 megabytes of JS to display an article and run 48 ad trackers. That's a crappy experience even on my smartphone and decent WiFi.
I'd also wager that there are many, many "apps" you could write which could be 100% server-side rendering and simple markup, and which would be super useful in the developing world. Think Craiglist and Ebay - the kind of applications that help people share their location and agree on prices of goods.
The "use anywhere" promise of the web is easier to keep if you lean on the web's fundamentals instead of writing a browser to run in the browser and shipping it to every user.
It's cool that a (potentially) billion more computers will be running linux: https://en.wikipedia.org/wiki/KaiOS
And for a large majority even that feature phone is the first kind of computing device they ever own.
These are, nevertheless, smartphones. Just smartphones that squeeze out every drop of cost. They will have access to most Android apps, and a full featured browser.
There may be a market niche for phones without a color screen, no installable apps, etc. It will be challenging to eke out significantly lower costs than lo-end hardware suitable for lo-end Android configurations, due to volume efficiency.
Would it have been impossible to use Linux on these phones under GPL v3? I assume that it would be very difficult to provide "installation information" required by the license.
A cheap phone does not have to be lacking in features. Mostly the reason we don't see them a lot in the west is because manufacturers prefer to sell you a more expensive phone. It's that simple. Operators and manufacturers conspire to keep these thing out of the market because they want to sell more expensive products.
What's happening elsewhere is that the competition elsewhere is heating up for the rest of the world. For the last decade that market was dominated by landfill Android and a multitude of cheap rip offs and legacy products. E.g. Nokia kept milking S30 well into this decade and there might still be a few variants that they are selling. The original version of that dates back to the early nineties.
Before they sold their phone business, Nokia had big plans for bringing linux to feature phones and they had a platform in advanced stages of development before they layed off the entire unit. They actually had two different linux platforms, meego for smartphones, which did launch publically before killing it off and Meltemi (internal code name) which they never launched. Meltemi was intended for cheap phones (replacing the S40 platform, not Symbian in case you confused that with S60) and based on QT/QML and linux running on cheap hardware. S40 was actually a quite capable system and effectively the money maker for Nokia.
The Meltemi strategy is basically what is happening now almost a decade later in the form of SailOS, KaiOS and other linux based platform trying to fill the gap left by a mostly indifferent Google and Apple when it comes to low end platforrms. Apple simply does not sell to poor people and Google isn't much better. They like talking about it but it's an after thought in their platform strategy.
Samsung tried pushing Bada (basically an evolved Meego) for a while. LG had WebOS and there have been a few more. The problem with all of these (and Nokia) is that electronics manufacturers are not great at developing software. Apple is a lone exception and arguably they are a software developer that happens to produce hardware instead.
In other words, they are maybe like Japanese phones two decades ago.
4G capability and maps are a plus but neither essential.
Their lower-end offerings would be ‘lowcost’ thin-clients.
IMHO, that's a feature, not a bug ... if all you really want is a phone.
Perhaps it's worth remembering that these "smart feature phones" aren't meant for people who just doesn't want a smart phone. They are meant for people who can't afford a real smart phone, so the feature set will creep towards being a full featured smart phone.
For those of us who genuinely do not want smartphone, there's currently not much on offer, unless we truly do not need a single smart phone feature.
I have a small set of features I'd like to see in a phone:
* Hotspot support, so both WiFi and 4G
* A TOTP application
* BlueTooth, for wireless headset and writing SMSs and managing the contacts from an application on my computer
Other than that the screen can be a black and white LCD for all I care.
Sure it won't be very stylish - nor will it be slim - but all the parts are out there with more than enough example code to hack together exactly the phone you want.
I'm seriously considering going this route for my next phone, since 4G modules recently became more widely available and much lower in price. But even 3G would suffice for my needs. I'm just wanting a "smart phone" which I have complete control over - or close to it.
I understand not everyone has the option or skills to build such a phone, but if you do (or are interested) - google around for DIY phones based on the Raspberry Pi (especially the Zero W), or the Arduino (using a variety of processors, though the original Arduino Phone was made with an ATMega328P iirc).
For your use-case, either one of the more powerful "standard" cores could be used (ATMega2560, ATMega644, SAM3X8E) - or maybe something based off the ESP line (ESP-12 or 32 would be my choice). A small 3G GSM module, plus a GPS and maybe an IMU, some kind of keypad, and a small OLED screen or something - keep everything at a 3.7 volt level and run it off a LiPo cell(s). With a bit of luck, you could probably fit everything into something about the size of an Altoids tin.
I have tried using a "dumb phone", which works for 90% of what I want to do with it, but it's inconvenient when meeting a friend at a new place or exploring a new city. Similarly, I have tried using a smartphone, but I find myself in cycles where I am inevitably checking it every 5 minutes to see what new stimulus there is to hold my attention.
I want one device that can do the following and only the following:
* gps
* browser
* text/mms
* camera
Nothing else.
Make it connect via data to a family console. It logs everything. Give these to all your family/kids and have them be a connected cluster.
It all logs back to a server on the home network. Backs up your shit auto.
Oh one other feature:
* walkie talkie only amongst the family group.
All data sims.... no cell.
Cell would be sip routed through the home router only.