What is Possible with Mobile in 2010?

From DevSummit
Jump to: navigation, search

Facilitated by Tomas Krag, Refugees United

Session Description

Tomas will share learnings from his work developing mobile applications in a number of international contexts in Africa and Europe. He'll demo the Refugees International mobile application, and discuss the limitations/advantages of various mobile technologies

Session Notes

Note-takers, please paste in your notes here

Also, please check out the Session Note-Taking Guidelines


Session notes by Amy Gahran

NPdev summit what's up with Mobile 2010

11/15 1pm

Tomas Crag - facilitator


Fair distribution of ppl who work w/ mobile, ppl who are just interested.

Oddly most people are into feature phones. - key pt is whatever works -- and where does it work


Someone here interested in mobile transactions, so is very interested in Frontline SMS credit (there's a rep of Frontline here)


Range of technologies when you talk about mobile: - SMS-based applications (work on every single phone), fairly simple to implement BUT very primitive interface. Can be difficult to deploy bcd you have to do it country by country. Security, translation is difficult. - WAP: Misnomer, not really WAP anymore, based on WML. But what people cal WAP phones now are actually primitive HTML browser, no support for tags or javascript. That has a lot of solutions of web (SSL encryption & security), can use same dev tools as web sites, but very limited. - Full web experience on smartphones - although term smartphone is a bit weird. Crag's view: If you have browser on phone with full web experience, that's smart enough. But others say benchmark is ability to run native apps. You can write full HTML apps, bigger screen, sometimes touch interfaces.


HTML5: Will people still need to write apps for phones? - Apps are critical for smartphones for deployment reasons, getting to as many people. They get you awareness when you're in the app store. - But there's a maintenance issue, you have to choose. So you have to integrate some tough choices. - For most nonprofits, multi-platform app development is a non-starter


Apps: If you do need to devel on a bunch of diff phones -- appcelerator is a company that can help with that - PhoneGap is an open source solution. Not as elegant though. But it works with HTML5. You build a web site and deploy it as an app

HTML5 and CS3: It makes web sites act like applications. That is going to be amazing - But apps still can offer more value. On a phone, Google Voice is far more accurate than when you access Google Voice via the web.


Apps are good at accessing other parts of phone - accelerator, GPS, camera, etc. - you will be able to access some of these features from HTML5


In Japan, over 72% of all web pages viewed from mobile device right now - That will be the case in US by 2014 too.

By next year in the US - half of mobile user will have some kind of smartphone.

You need to understand your audience, and what kind of phones you need to reach. And you need to compromise.

Frontline SMS: Interesting to think about mobile from the user side. Usually when we dev software, it's about developing software to send text message. That does allow us to be very versatile from situations here to Sierra Leone - We will be working to initiate mobile payments

USSD and Sim Toolkit - tools you use to deploy applications to the SIM card of feature phones, get built into the menu of the phones. Often operators have their own menus. They use SMS as the data transport. - Standards are open, so easy to build the tools, but you have the get the carrier's approval to do this.

How hard is it to go from the SMS to the web? - Larger corp solution is to get cooperation from the carrier, integrate into their system. - or frontline SMS approach that essentially gets you a number with a SIM card, then you can send and receive SMS message.

Tool comparison: MobileActive.org has great resources.

USSD and SIM toolkit are built into GSM standard, but not sure if they work for CDMA.

For those kinds of limitations, that's where browser-based solutions are applicable. So many cheap phones with browsers

Resources for getting started with mobile design - NEED TO LIST ON WIKI: Design pattern sites for mobile


GPS and maps + mobile - How to track actions that are happening at a particular place and time?

HTML5 and GPS/locative information - What are the phone features you can reach through the browser? Varies, but generally camera, GPS chip - You can do Google map mashups, Google Latitude, Ushahidi


Good platform for coordinating mobile content contributions: Ushahidi.org - Free, open source


Is shtere a sourceforge for mobile apps? - If I see a good app, why should I have to replicate that work - There's a lot of Android app development hosted on Google code - For mobile web, there are a lot of scripts that are available. - Hosting your service on amazon web services and using web service calls to pull in massive amounts of data, you can make it so it scales to infinity


keep your web sites very lean -- pages under 25K - In US, capacity crunch means don't bet that mobile users will have easy reliable access to infinitely large pipes.


Appcelerator and Jquery mobile are great for mobile app development and integration. - If you're developing a new site, you might want to consider using jquery library since it's the farthest ahead on mobile for now

PhoneGap: Build an HTML5 site for a generic mobile platform, it generates code for different platforms. - But if your site scales gracefully to webkit, it will probably play nice on most phones.


If you inherently need integration with camera and GPS on your site, you might as well focus on developing for webkit browsers.


Paypal: x.com and look up their mobile signup, look at their mobile payment APi, SDK toolkit - resource for mobile payments - Their goal is to replace your credit card with your phone


How big a bastard is the mobile phone operator - In West Africa, corruption is a problem - Operators in most countries are fairly big bastards - Crag: We tend to not go to the leading carriers, with smaller players it's easier to negotiate contracts


Next year ot two -- what's changing? - Tablets - 300M new mobile users worldwide - variety of smartphones keep growing - that will put the impetus on standers (deFeacto and real) -- that will make dev harder - More and more users at both ends of the spectrum - really different form factors

If you're a nonprofit and you what to reach the most people with least resources, mobile, what's gets strategy - Low-res browsers to reach larger audiences? - Or maybe trying to engage donors more - SMS has a huge use case, hundred of millions of users globally. If that's who you're trying to reach, you have to go there.


Having an optimized web site is middle of the road, that's probably a good place to start, to be accessible and have a baseline. - If your site is in Drupal


Summary

3 classes of user experience, have to make decisions for each. - smartphone users - feature phone users - SMS


Standards - defacto vs. official - Both technical standards are imp but not quite ready. So committing to interim standards has its own risk


END