HTC unveils record profits, passes Nokia http://www.t3.com/news/htc-unveils-record-profits-passes-nokia?=55253&ns_campaign=news&ns_mchannel=rss&ns_source=t3&ns_linkname=0&ns_fee=0
HTC unveils record profits, passes Nokia http://www.t3.com/news/htc-unveils-record-profits-passes-nokia?=55253&ns_campaign=news&ns_mchannel=rss&ns_source=t3&ns_linkname=0&ns_fee=0
http://www.mobilecrunch.com/2009/11/10/taptu-oneriot-realtime-mobile-search/
Sent to you via Google Reader
Great event covering a variety of Android topics at Skills Matter:
Dalvik bytecode disassembly was the highlight, mostly because it was technology for technology’s sake. As the presenter explained several times, there is no real reason to decompile a .dex file into bytecode. Nonetheless, it was fascinating and I learned a few things. Unlike all the J2ME handsets, and even stuff like SavaJE, Android isn’t Sun Java. No steaming coffee cup branding anywhere in sight. Android developers do write Java code and that Java code is compiled into bytecode in .class files but that is where the similarities end. The dx (dexer) tool changes .class files into .dex files which contain Dalvik bytecode rather than Java bytecode. This tool is hidden in the Android toolchain. Our presenter has developed a library which can take .dex files and output the proprietary Dalvik bytecode format. The tool is called dedexer and is comparable to jad in the regular Java bytecode world. The Dalvik bytecode didn’t look a million miles away from Java bytecode, but our presenter noted that based on his inspection of lots of different bytecodes over the years this implementation had “done everything right for once.” A very encouraging message for people in the Android space.
Next was the spacer project, which provides a geospatial backend for field researchers. For example, if a botanist is collecting samples in a forest and notices an unusual shrub, they would often just collect a sample and check the dataset when they get back home. This software allows a researcher to query their datasets from the field to determine if the shrub should be sampled at all. My favourite part of this presentation was the discussion of open geospatial formats. The project uses the ISO 19115 format but will be moving to the more concise FGDC format while surprisingly Open Street Maps does not provide an open format.