Slashdot Mirror
Time For a Hobbyist Smartphone?
theodp writes "Over at Scripting News, Dave Winer has a hobbyist phone on his wish list. Innovative phone manufacturers, Winer suggests, should 'make a smart phone with a really great scripting language, with all kinds of scriptable tools on board. Instead of disallowing scripting, disallow apps that can't be scripted. Make a great simple programming environment that runs on desktops or laptops that plugs right in, but it should also be easy to write scripts on the phone itself. Dave concludes, 'We've already seen the Jobs phone. Now it's time for Woz's.' Having ditched App Inventor, it would appear that Google isn't interested. Microsoft Research has the idea, if not the right implementation, with TouchDevelop (video). Any other existing or in-the-works projects that might fit the bill?"
Too bad Nokia quit making fun phones. The last was the N900. I'd love to have a new phone similar to that with modern specs.
There are plenty of good devices out there from a hardware standpoint. The difference is that software is tailored by the manufacturer to reinforce their channels, partnerships, etc. (to the detriment of the user).
Once rooted, these devices get a lot closer to what I think most people are looking for in a hobbyist device. However, that is the trick. Getting root access, and knowledge of the internals so that we can start work on our own.
We've seen the failures hobbyist devices in the past. I expect the same going forward. If only we had access to the hardware and programming specs, we wouldn't even be having this conversation now. Maybe we can find a way to extort this info from these manufacturers.
I recently defected from iOS and I'm having great fun with an android app called QPython.
Docs are still a bit spotty, but with a few lines of Python I can do all sorts of things with the Android API.
Yes, wouldn't it be great if someone made a completely developer-friendly, unlocked phone with good specs at a decent price?
Yeah, no fucking shit it would. The carriers would NEVER allow it on their networks. We won't see a phone like that until we have a worldwide standard for cell networks, so that somebody could make that phone and actually sell it in the US.
That's true for current methods of developing software. Which is typing in code.
Programming hasn't changed very much in 50 or so years. And I think it's ludicrous that we're using a language to prgram a computer to do mathematical operations.
What we really need is a symbolic programming "language" and it would rock on a touch screen.
Why not go directly from dragging and dropping logic to machine code directly? There is no physcal law that says we have to program computers the way we do now.
These "verbal" type of programming languages are so 20th century, inefficient and just old fashioned. Their time has passed.
There was already a phone proposed that could have done this with no problem. There wasn't enough interest on it to make it a reality.
http://www.indiegogo.com/projects/ubuntu-edge
And before you go complaining about the cost, please have a look at flagship Android phones and how much they cost *off contract*. The Edge was a pretty good value.
You should check out Kivy- it will let you create a native app in Python for iOS or Android. http://kivy.org/
A witty saying proves you are wittier than the next guy.
To me, the Android platform was close enough. It's "just" Java (if you can't figure out Java...), there are no fees required to get the development environment or simulators set up, Android devices and phones are available new for as little as $60 (and cheaper as people upgrade). So...to me, Android IS a hobby-friendly environment.
It's currently in a second Alpha state (if you're pulling the latest binaries, probably further along) and runs on a Samsung Galaxy Nexus (couple other profiles too). It's fully open source at this point, and even though supposedly LG wants to use it for TVs or something, there's a group that's been working on it for awhile ever since it was divulged from HPs hands. There are also efforts underway to emulate Android apps on the platform. The community also greatly appreciates anyone enthusiastic enough to contribute. You can find everything at webOSInternals. I still use a webOS phone as my daily device mainly because it does the things I need it to do very well (and other platforms come with way too many strings attached for me).
If you want a hobbyist platform that the big platforms still steal ideas from...there you go. That's the epitome of a hobbyist platform. The scripting is all html/css/javascript using the Enyo framework. It's all open standards and there are plenty of tools that were built by Palm and later HP.
Dave Winer basically invented application scripting on Macintosh when he released Userland Frontier in the late 1980's. This was before Apple's own AppleScript. He also was a key in the creation of XML-RPC and SOAP for creating web API's.
I read this as a HARD requirement for scriptablity across all applications. If all apps are able to respond to and make scripting API requests, any app would be able to be the "programming" language. Python, Sqeak, C, Photoshop, nginx, OpenOffice could all be equals.
That was my take, too. I recognized the name, and more or less translated the idea to "I want PhoneGap for everything".
This idea will never fly. Dave is old. I understand his perspective, because I lived through that era too: no one writes malicious code because everyone plays nice, and because of the way Apple codes things I don't have libraries to e.g. access my iTunes library as an iTunes library without doing a whole hell of a lot of work so I'd like to be able to ask iTunes to do it for me.
The reason this will never fly is manifold:
- There are incredibly sound UI and UX reasons to not allow "skinning" or "themes", or, frankly, even things like the Facebook ban on "Social Fixer". Programmers who like to tweak UI bits, but who either could never write iTunes on their own, or find doing things like that incredibly boring, don't like the lack of tweakability. Users, on the other hand, like to be able to call up a support person and get quick answers to their problems. Having the same icons and application layout everywhere GREATLY helps this. Users also do not like learning curves; having the same "look and feel", and having all applications follow the same HI guidelines, like they all have a "File" menu in the same place, and they all largely have the same items on the menus in the same places GREATLY helps this. It also makes it look like all your Mac apps were written in Cupertino, and all your Windows apps were written in Redmond, which programmers also hate, since it disallows use of trade dress to make their product stand out from the other products that do almost exactly the same thing.
- It requires reaching across NECESSARY security protection domains. If I can talk to iTunes, and make it do things, there is effectively no more protection domain between myself and iTunes. Same for AddressBook. Same for Email. Same for Safari. Same for GPS info. Same for making phone calls. Make no mistake: there are assholes out there writing code; some do it just to be assholes, while others do it for money, politics, religion, etc.. But there are assholes, and having these Chinese walls between things that operate on data keeps a nice apps data from being accessed by assholes. So no cross-compaartmentalized scripting, unless you want to let the assholes do it too.
- There just aren't that many programmers out there, and of those, only a fraction really get their rocks off tweaking UIs, and of those, only a fraction like to do their tweaking in a scripting rather than a compiled language. So there's not a lot of MARKET for the idea, and without a market, the idea isn't going to get done. Or it'll get done, and go nowhere, because the company will fail... hence my earlier comment about him feeling free to found a company or three to pursue the idea.
Really he should be asking for APIs for things, but that's very antithetical to the whole historical Apple worldview that "iTunes databases should be managed by iTunes, and everyone else should keep their damn noses out". That's a pretty big paradigm shift for Dave, and an even bigger one for Apple: it's just not going to happen.
Most everyone else will also be reluctant to open up APIs to their data, even under Android, even under the misdirection of calling it "making things scriptable". Welcome to the reason that CORBA and OmniORB and other object request broker paradigms are niche market at best and curiousities at worst.