Pogue’s Posts Blog: A Better Google Maps App for Apple and Android Devices

Google Map’s new directory buttons.

Our story so far: Last September, Apple decided to dump the Google Maps app that had been on the iPhone for years. Apple replaced it with its own Maps app — software with so many problems that Apple’s chief executive, Tim Cook, apologized and even recommended that people use other apps until Apple could fix its own one.

In December — incredibly quickly — Google responded by introducing its own Maps app for iPhone. It’s a spectacular app, among the best apps ever written. It’s fast, beautiful and so good at guessing what you mean when you start typing a destination, it’s almost mind reading. You can read the details here.

Today, that delightful news gets even better. Not only has Google improved Google Maps for iPhone, it’s also brought that same free app to three machines that never had it: the iPad, Android phones and Android tablets. (The Android versions are available for download today; it requires the Ice Cream Sandwich or Jelly Bean version of Android — recent versions, in other words. The iOS versions will be available shortly.)

For Androidians, the biggest news is the design of the app itself. It’s modeled on the iPhone app, the one that’s simple and fast and elegant. It’s also uncluttered by the morass of menus that have always plagued the existing Maps app for Android.

But for practitioners of all religions — tablet, phone, iOS, Android — the other news is the new features that today’s new version brings. They include:

* Greater speed. All app versions are faster than before.

* Better place information. Half the time, you don’t even need navigation instructions; you just use Google Maps as the world’s smartest Yellow Pages, to find a nearby restaurant, movie theater, drugstore or whatever.

The details for found places now include a one-line description (“Chinese restaurant famous for dim sum”); a five-star rating system (including a decimal — “4.3,” for example — because, let’s face it, almost everything these days winds up with a four-star rating); the ability to upload your own photos of a place; and a more complete integration of the Zagat guides, which Google bought.

* Greater emphasis on exploration. Google Maps has always excelled at getting you to a known destination. But Google now wants the app to help you choose a restaurant, bar, store, recreation center or hotel, at least in major United States and European cities.

If you tap in the Search box without typing anything, new, photographic buttons appear: Eat, Drink, Shop, Play, Sleep. Each opens lists of corresponding facilities, sorted by criteria like Local Favorites, Popular with Tourists and so on. (Google says that these recommendations are never paid placement.)

* Traffic incidents and auto-rerouting. At last: Google Maps shows more than colored lines indicating current traffic speeds on major roads. Now it also displays tiny icons that represent accidents and construction. Tap one to read the details: “Right lane blocked on 680,” for example. (In case you were wondering, the information on traffic incidents doesn’t come from Waze, the traffic-incident app that Google recently bought. That data has yet to be incorporated into Maps.)

Better yet: Maps now looks ahead for traffic jams on your route, and interrupts your drive with a dialog box that offers to route you around it (if the new path would be quicker, of course). On its own.

* Offline maps. This feature is something of an Easter egg. It’s undocumented, a feature inserted by Google engineers simply because they wanted it. You can access it only if you know the secret. But wow, is it worth it.

This feature memorizes the map data for whatever area is displayed on your screen right now (up to a whole city in size). That way, you can use Google Maps even when you’re overseas and don’t want to turn on data roaming (because that’s insanely expensive), or when you’re in an area where there’s no cell reception. It’s very handy.

To capture a map snapshot like this, tap in the Search box. Use the speech-recognition button and say, “OK Maps.” (It’s a riff on the command “OK Glass” that prepares Google Glass, the company’s “smart headband,” for voice commands.)

A message quietly lets you know you’ve successfully stored the displayed area.

*Nice tablet layouts. On a tablet, Maps really shines. The app smartly reformats itself to take best advantage of whatever screen shape you have: two or three columns of place listings, for example, and luxuriously displayed photos and reviews for each business.

This new, improved Maps app works identically on both major flavors of phone and tablet. You know what? I don’t care how much you distrust Google and its motives. This is crazy good software, some the best work Google has ever done.

Bits Blog: 300,000 Android 'Phones' Roam a Virtual Town

David F. Gallagher/The New York Times A cluster of 500 PCs at Sandia National Laboratories in Livermore, Calif., is running a complex simulation of cellphones in the wild.

An air-conditioned data center at Sandia National Laboratories in Livermore, Calif., has the ambience of a Best Buy showroom, with 500 inexpensive desktop computers lined up on racks. But the PCs here are doing something special.

A small team of cybersecurity researchers have lashed the computers together to form a homebrew computing cluster. They are stacked five levels high, braced in impromptu fashion with metal girders in the event of earthquakes, and woven together with colorful Ethernet cables. They serve as a cyber-Potemkin village — a distributed software simulation that is intended to mimic the behavior of an entire city’s worth of Android smartphones.

So far, the researchers have emulated the behavior of a network of 300,000 Android phones, down to details like sending text messages and transmitting wireless data, and even the behavior of the phone’s radios or sensors.

In the future, they hope that their software, which will be freely available through an open-source license, will be used by groups designing social network applications like Foursquare, military battle planners, disaster relief workers and American hackers who are building mesh wireless networks intended to make the Internet available inside nations controlled by dictators, among others.

“We wanted to create something that would approximate the many, many phones in a real network,” said John Floren, a Sandia computer scientist who is part of the group of researchers that built the system, which is called MegaDroid.

The idea is that while such simulators have in the past been developed by the engineers of large data networks, they have not been generally accessible to small groups of designers who might be working on a some new type of distributed application.

To date, wireless smartphone networks have not suffered malicious software infections at the same rate as the standard wired network, which is now plagued by a vast array of malware. Many computer security specialists believe that this will soon change as personal computing goes mobile, much as Willie Sutton robbed banks because, as he put it, that’s where the money was.

A wireless network simulation like Megadroid would make it possible for security researchers to gain new insights into “attack vectors,” according to the Sandia engineers. Their computing cluster is in an unclassified area of Sandia, which historically has been one of the two engineering and design centers that has built and maintained the nation’s nuclear weapons.

On a large monitor outside the data center, it appears as if the researchers have turned loose a frenetic army of virtual Charlie Sheens on the city of Livermore, Calif. (The use of the Charlie Sheen icons is apparently an inside joke among the researchers.) The scene is actually a visualization tool used by the MegaDroid simulator, which includes fake GPS data and gives the appearance of a beehive of cellular activity as the Android phone owners drive and walk around a map of the city.

The researchers stressed that simulating a wireless network is a much more complex undertaking than simulating a standard desktop-based network.

“In a desktop PC network, the biggest problem you have is someone kicking a cord out,” said David Fritz, one of the MegaDroid designers.

In contrast, modern wireless networks are composed of shifting herds of tens of thousands of smartphone users. The simulation is made even more complex by the fact that each phone contains multiple radios and sensors.

The danger of not having a deep understanding of the behavior of a computer network was made clear several decades ago at the very beginning of the Internet era. In 1988 a Cornell graduate student brought much of the Internet to a standstill — it was composed of about 50,000 computers at the time — when he unleashed a program known as a worm that was intended to live in the network as a bit of digital graffiti — in effect, the cyber-equivalent of writing “Kilroy was here.” However, because of an error in coding that caused the software to rapidly replicate, a vast number of copies spread through the network, overwhelming it.

The researchers said they were concerned about such issues of network resilience, but they also saw the possibility that MegaDroid might be used to build new kinds of distributed applications. For example, disaster workers in an earthquake zone could harness the accelerometers in their phones together to provide a digital seismograph, tracking aftershocks. Or small and inexpensive chemical and biological sensors could be added to smartphones carried by government workers, creating an early warning system in the event of terrorist attacks.

The researchers said that they were sensitive to privacy issues and that they believed the simulator could be used to build systems that provided useful data in national security situations while protecting privacy.

Right now the system models only the open-source Android operating system designed by Google, but the Sandia researchers said they hoped to add other operating sytems like Apple’s IOS. This would give a more realistic model of the wireless data world.