Jump to content

  • Log In with Google      Sign In   
  • Create Account


Member Since 10 Aug 2000
Offline Last Active Aug 19 2015 05:09 AM

Posts I've Made

In Topic: The best way to build your Android app?

08 July 2015 - 07:25 AM

Horses for courses, I suppose; of the many development platforms I've worked with, Android is certainly one of the easier ones. Dialing back the hyperbole is a good idea, if you're asking for help.


If you are intending to work with native Android, then there is really only one answer: Android Studio.


I assume you have a good reason for working in the NDK, in which case, you should of course use the 1.3 beta, perhaps even look to the canary build to see if there is some useful new stuff in it, though I suspect that at this point they've ceased adding new features to it. If Android Studio does not include the specific debugging tool that you want; use the ones that it does have: logging of everything and use ndk-gdb and adb for more advanced needs (getting ndk-gdb to work with gradle requires jumping a few hoops, but isn't really all that difficult: google/stack overflow it). The terminal is there for a reason.


Google no longer supports Eclipse, so while I love what Eclipse offers (I use it extensively in my day job), I wouldn't recommend it for Android development anymore. At this point, AS is the superior solution, and the functionality gap is just going to continue to grow. If you're working on anything that you plan to support for a while, you don't want to be stuck on an old IDE for an ecosystem that changes as rapidly as Android still does.


The only reason I can see for going with Visual Studio is if you're working at a Microsoft-only tools studio.

In Topic: Developing in native for mobiles question

08 July 2015 - 06:02 AM

Olof and frob have already answered your question, I think.


Just want to mention that for an existing Android Java project, I would look towards RoboVM + libGDX. You won't get away with still having to reimplement the UI and stuff like that using RoboVM/libGDX bindings, but - depending on your code, of course - you should be able to reuse most of the actual game engine.

In Topic: Is Google Play Game Services or Apple Game Center worth it?

08 July 2015 - 05:57 AM

Take everything Oluseyi just wrote, and frame it on a wall. So much truth that it hurts.

In Topic: Easy source control for educational uses?

03 July 2015 - 02:36 PM

Definitely go for Mercurial and use Bitbucket. It's a DVCS like Git, but with an interface that is (by design) almost as simple as subversion.


Subversion really isn't worth it, IMO, at this point.


I use Git for everything these days because of Github - simply easier to standardize on one VCS, rather than maintaining separate VCS for public and private code (which is where Bitbucket shines -> free private repos). But Mercurial was my gateway to DVCS's back when, and IMO it is really not all that much more complicated to explain and use than SVN, as long as you don't insist on doing overly complicated stuff. If you think your people can handle SVN, then they should be able to handle Mercurial as well.


That being said, we've recently started moving the researchers at my place of work over to Github, and that has actually gone pretty well. Git isn't that hard to use, if you stick to a simple workflow, and Github has pretty excellent documentation.

In Topic: How do I make Android backward compactible

03 July 2015 - 12:57 PM

Unless you're doing some advanced stuff, version incompatibilities should usually not be the first thing to suspect. In 5+ years on Google Play, I think I can count on my hands the number of issues caused by new versions of Android - as opposed to simply caused by my own stupidity (i.e., bugs).


Also, I'd strongly recommend moving to Android Studio. I stuck with Eclipse for a long time myself because it's what I use at work, but at this point AS is the clearly superior IDE for Android. After Google ceased support for the Eclipse SDK, I don't think sticking with Eclipse is worthwhile for Android development. Potential cause of issues, with few upsides.