Dev Days: Mobile

Let's revisit Dev Days San Francisco. Last time, I talked about the Microsoft talk, and how I was able to incorporate auto sanitization into Rails. This time, I would like to discuss the three mobile talks that were presented. Just for full disclosure, I am a big fan of Android, so you will likely see that bias come through. I know I felt the bias while taking in the three talks.

iPhone

First up was Rory Blythe talking about iPhone development. Of the three presenters, he was the most at ease speaking in front of a crowd. I'm sure Rory is a nice guy, but he had a swagger that... felt very Apple-esque. He was the hipster Apple guy smugly looking at us and telling us why we should cow down to the almighty Steve Jobs, and why we should be happy to do so. He made the audience laugh the most, most often with flaws in the iPhone development platform that we all know is an issue. Cheekily telling us three or four times that you cannot develop for the iPhone on anything but an Apple machine sure doesn't sit well with an Ubuntu fan like me. I don't think he was trying to sell the platform, though, just give us a taste for what it was like.

Most of the talk was him walking through the steps to develop a simple application in Xcode. It was very visual. Click here and drag over there to hook up a click handler of some kind. Drag this widgit in line with these others. Ultimately, we got to peek at some actual Objective C. If you weren't at Dev Days and you have never seen Objective C, be thankful you didn't have to, because it ain't pretty. He joked about how many places you have to adjust the code just to set up a simple property. As a Rails web developer, it made me want to claw my eyes out. I would never want to develop in such a redundant, cumbersome language.

There was good news, though. He introduced us to MonoTouch, a development environment for producing iPhone applications written in C# with the .NET platform. I'll take Ruby any day, but between the choices of C# and Objective C, it is a no brainer (C# of course... from someone who generally despises Microsoft technology). Too bad I can't use Ruby to develop an iPhone app on an Ubuntu machine. I might consider iPhone development then, though not seriously. If you wonder what is wrong with me... well, I choose an open platform, because I feel in the long run it will win out to a closed platform. I am convinced all the bad press towards the App Store approval process fiasco, and Apple's draconian attitude towards the 3rd party development community will ultimately be the iPhone's undoing.

Maemo/QT

The next mobile talk came from Daniel Rocha of Nokia. He talked about Qt, specifically in the context of Maemo. He was very dry. Dryer than a desert when compared to Rory. Some people just aren't built for speaking... I'm sure I would freeze up and be just as bad, but thankfully I don't seek such torture for myself.

It was a running theme to do some actual development on stage to demo the platform for the audience. The core message of his talk was "Look! Cross platform! Same code for all the major platforms, including mobile!" Alright, C++ cross platform I guess is kinda cool, except it's not that impressive in today's context. Java has been highly cross platform from the beginning, and all the primary scripting languages are quite cross platform as well.

The most impressive aspect of his cross platform demo is that the UI is among this cross-platformness, which I suppose is the whole point. However, he admitted about still needing macros to segregate the platforms now and then. And I really don't believe that a mobile UI should be the same as a desktop UI. With the state of mobile devices, I think it's wiser to develop a UI with a small screen (and touch screen capabilities) in mind. Negative points also for running Linux in a VM inside Windows instead of the other way around.

The Qt IDE he was showing off was also quite unimpressive. It looked like Visual Studio circa 2000. The UI designer looked exactly like an old Visual Studio UI designer, with that grid of dots and all. I got the feeling that anyone used to a good IDE like Eclipse would feel shackled in this thing. Anyone used to an awesome editor like Emacs will... well, an Emacs user would never stoop to a lesser editor of any kind.

The funny thing about this talk is that it was followed by an Android presenter who pointed out how most major manufacturers, except Nokia, were coming out with devices with Android. All I can say is, Nokia is dropping the ball if they stick with this platform over hopping on Android. I saw no compelling reason to develop for Qt/Maemo.

Android

James Yum finished the Mobile hat trick with an Android talk. I wanted to be blown away by this talk, but unfortunately Rory was the only awesome speaker of the three. To be fair, James was up front that he was asked to do this talk last minute, replacing the original Google speaker, who probably would have crushed it.

He started off by showing the "Droid Does" Verizon commercial, which was an awesome commercial (despite not actually showing the device, and despite coming from a company I'm not a fan of). He then read some of the YouTube comments for the commercial, to humorous effect. I really hope T-Mobile comes out with a phone this compelling. According to James and the commercial, it has an 800x480 screen with a physical keyboard and a powerful chipset.

Great start, lousy finish. He followed this up with an uninspiring demo of how to deal with threads in Android, with the specific goal of making a snappier UI. He went from a simplistic but completely incorrect way to do threading to an object oriented way using the Android API. It made developing with Java look almost as bad as Objective C. I'm not exactly a fan of Java (though I work in it almost every day), but it really isn't this bad to work with Java, and specifically Android. I think he would have better shown off the platform by taking a real (though small) application idea, and implementing it on stage. This is what Rory did, and it had a much greater effect than slogging through the most complicated aspect of modern programming you could possibly think of.

Overall, James was far too green for the speech. He was (understandably) clearly nervous, and he was unable to answer most of the questions at the end of his talk. I was silently rooting for him, hoping he could show all the would-be iPhone developers what they are missing, but I was disappointed. Maybe Google will learn from this and keep a good speaker on hand as a backup, should the primary speaker have to drop out last minute.