/ Android

AndroidTO 2013

Yesterday afternoon my boss asked if I wanted to go to the AndroidTO developer conference today, as one of our reserved spots had freed up. I said sure, why not, I like Android. I'm not a developer, but my work is developer-adjacent, and I figured I could attend some of the less developer-ish tracks that our devs might not be as interested in.

I was quite amused that the Business track was in the IMAX theatre with the big comfy seats. This only had a slight impact on my mostly attending Business track presenatations, honest. ;) I didn't attend any of the Gaming or Design track sessions, and only one Development. I'll add in links to the various presentation decks or videos if/when I find them posted.

tl;dr version -- I went to a bunch of presentations and learned interesting things from some and not much from others, and would certainly go back again next year if someone else is paying for it. :)

Venue

The conference was held at the Scotiabank Theatre, the Cineplex movie theatre at Richmond St and John St in downtown Toronto. This is conveniently one block from our office, so easy to pop back there at lunch and after work, so I didn't have to haul all my stuff with me the whole time.

I thought it was a good idea to have a conference in a movie theatre -- I'm not sure if they've done this conference there before, but it seemed to work nicely. The theatres were good for speaker presentations, there was some good central space for booths, and plenty of washrooms. It did seem unreasonably warm, though, especially for a movie theatre. I'm not sure if the conference folks intentionally asked for the AC to be turned down, or if the building just doesn't turn on the AC until later in the afternoon because the theatres are not usually open until then. Either way, I found it quite warm, far warmer even than our office.

Opening Keynote

I missed almost all of the opening keynote by Akshay Agarwal of ARM because it started at 8:45am (?!) and I didn't get there until almost 9:30am due to my normal commute schedule plus some GO train congestion. I saw the very end, and from what I saw it was pretty laid back and not overly exciting, though there may have been some interesting content in there about ARM and Android. Feel free to add a comment if you were there.

First speaker - Rob Woodbridge from Untether.tv###

(Business track)

Topic: "Don't be a Mayfly - How to lengthen the short life-span of mobile app revenue" (slides)

Rob was an energetic and engaging speaker, and his basic premise was that almost all mobile apps have a lifespan of only two or three days (like a Mayfly) before they fall into obscurity and are never heard from again, and that these few days are the cumulation of months of work. A lot of this is because it's really hard to find anything on the app stores these days, and also it's really expensive to do any sort of traditional marketing to get your app out there.

He also pointed out several times that if you're not building your app with some sort of directly associated revenue model in mind, you're not actually in business, and you should do something else. This seemed fairly specifically focused entrepreneurial app developers and not as appropriate for some other motivations behind making apps. At my work (digital marketing agency) for example, most apps we work on are focused on extending existing marketing activities of clients, so there's not usually a separate revenue stream directly associated with an app, it's just one part of a larger endeavour.

His final conclusion was that you shouldn't be chasing the "big top 6" apps that have been around for years, you should just try to be the app that makes it a few days longer than the rest -- then you can work on lasting a few days longer than that.

In all, the talk was quite entertaining, and it had some interesting ideas and examples -- Rob included lots of quotes and anecdotes from interviews on his Untether.tv channel, which I will check out -- but I'm not sure how much of the central topic of his talk was directly applicable to anything outside of the entrepreneurial app development realm.

Second speaker - Orrie Shannon - Wattpad

(Development track)

Topic: "How We Share Wattpad with the World"

When I first looked at the summary schedule I'd decided I wanted to check out this presentation because I wanted to find out more about Wattpad, because I see their name all over but I don't know that much about what they do. As it turned out, this was not only one of the more interesting presentations I attended, but unlike almost everything else at the conference it's actually directly and immediately applicable to my job. :)

Orrie's presentation was about how Wattpad uses various strategies in their release process to prevent releasing apps that suck (paraphrase) and make sure developers are focusing on the right things.

Specifically, their process uses an internal beta, an external beta, and a release version of the software. As their release cycle rolls out, at around day 7 they start cutting internal betas from their development branch, which are distributed to everyone in the company via an internal Play Store distribution feature that's available to anyone using Google Apps for Business. Internal users give feedback and report bugs, fixes go in, and a bit farther down the line (day 14?) they cut an external beta release from the development branch, and over the next week or so increasingly push it to 5-20% of users with the Play Store staged rollout feature.

Somewhere around day 21 of the cycle they merge into their release branch, and that's used for all minor bugfixing and beta and release builds for this version, and devs can start new features on the developmene branch. They do another beta with 100% of beta users (they have a force upgrade feature on the betas) that is basically a gold release candidate, then once everything's basically in line they cut the final release. The whole cycle is right around four weeks.

(note: going from somewhat tired memory on this, so some day numbers or other specifics may be misremembered, but you get the gist)

Orrie also covered in some detail how they gather analytics -- both to test potential changes in the betas and validate changes in the releases -- and Play store comments, crash data and logging to help support and devs debug issues in the app.

There were of course more things covered, but that's what sticks out at the moment. This one is another that I'd like to grab the presentation slides for, because it had a lot of good operational stuff in there, which is my kind of thing. :)

Third speaker - Dave Robinson - Rogers

(Business track)

Topic: NFC and the world of m-commerce in Canada

I am not a huge Rogers fan as an end-user of their products, for various reasons. That being said, this was an extremely fascinating presentation for a few reasons.

First, the talk was not framed within the end-user realm, it was all about telco and B2B platform infrastructure. I haven't experienced Rogers at this level before, and while still ruthlessly financially driven, they also seem extremely efficient, focused, and competent -- none of which would describe most of my experiences with Rogers as an end-consumer.

Next, Dave Robinson obviously really, really knows what's going on in the NFC/contactless technology world, from the low-level technology through the global standards and up to all the Canadian legal and regulatory requirements. He was certainly the right person to be giving this talk.

Dave started with a background on the history of the NFC/SIM technologies and how Rogers has been active in integrating technologies and working with other companies (particularly banks) on various mobile payment systems for many years, how they've succeeded in defining standards for Canadian banking to use, and how they're now involved in defining platform standards at the global level.

He then moved on to technical details (at an architectural level) of how the current and upcoming NFC/contactless devices and various e-wallets work, and how they operate with the banks, with merchants, and end users, and also how apps can tie in to the system (with appropriate user authorizations). He also gave some details of how they anticipate these technologies will impact other non-financial areas such as healthcare -- apparently authenticating the recipients of healthcare in many provinces is a huge issue that everyone just ignores right now, but with your ID in a SIM that can positively identify you with an authentication step (PIN, etc) that could change.

One interesting point was that Apple has opted out of the NFC features in their devices, so all work so far has been on Android. The expectation is that once the NFC ecosystem starts rolling out Apple will add the features to their phones, or some other contactless technology will need to be used (Bluetooth, etc), or else they will be left behind. Because they're highly US-centric and the US has not adopted NFC/SIM use as much as the rest of the world, they are at higher risk of not catching up.

The talk was extremely informative, quick moving, for the most part entirely understandable (a bit of telco jargon, but not critical), and also quite candid. I learned more about how NFC/contactless payment systems and their potential impacts in an hour than I probably could've in a day of researching. The only unanswered question I have is when will all this roll out in a widespread fashion -- their timetable seems to be sometime in the next couple of years, and I will certainly be watching for it.

And it also significantly raised my optinion of at least some parts of Rogers, which I really hadn't expected. :)

Fourth speaker - Scott Dyer - Corus

(Business track)

Topic: Mobile Apps in the Canadian Broadcast Ecosystem

I'm interested in broadcast media, a lingering remnant from my ~5 years at CBC, so thought I'd see what Corus had to say. Most of the presentation was background about Corus, their programming, and how the Canadian content creation/broadcast world works, most of which I was already aware of and wasn't super interesting to Android developers. There was a bit of Corus-specific information that was interesting and new to me, and some details about decrease in broadcast TV viewership that I hadn't seen before, but not much about mobile apps until the last 15min of the talk.

To boil down that last portion, Corus has tried in the past to do their own app creation, but it hasn't been particularly successful or efficient and it's not their core business, so their current approach is partnering with or investing in app creation specialists. The focus is primarily on games and video for kids, especially for that content where they own all the rights. They try to keep a hands-off approach and let the app developers do their thing.

Overall not a bad talk, but would've been better to cut the time spent on background and left more time for questions at the end to expand on mobile app specifics.

Closing Keynote

The closing keynote by Chris Haseman of Tumblr was quite good, I'll link to it once he's put it online (should be tomorrow). Most of the talk centred around paradigm shifts he encountered during the move from being a developer to a dev manager.

One example is that developers should as much as possible be kept in an environment where they poll for information -- they need long stretches of time to dig into coding, but will come up for air periodically, so send them emails to address then instead of tapping them on the shoulder, which shatters the "stack of code they're holding in their head". In contrast, dev managers are almost entirely interrupt driven, constantly dealing with incoming requests.

Another example is that developers constantly want to have fixed and final requirements before they start to build, and constantly complain when requirements change during development. He pointed out that requirements will always shift, and are only fixed when whatever you're working on ships. Up until that point it's a negotiation between design and development to make the best possible app (or site, or widget) that possibly can be made without totally derailing development and without totally sacrificing design. If all parties are focused on the final product and there's communication between them (key) you can make great things.

There were far too many topics to cover here, it was an action and energy packed 45min presentation. I'd say it's well worth seeing Chris speak if you have the chance.

Conclusion

The conference in general was well organized and setup. The talks I attended were at least interesting, and a couple quite useful. I'd certainly be happy to attend again next year.

Bryan Fullerton

Bryan Fullerton

Just a middle-aged dude doing his best to survive. Geeky Internet things from remote professionally.

Read More
AndroidTO 2013
Share this