Google Maps Now Does Interactive Re-Routing

An anonymous reader writes "Remember how cool it was the first time you used MapQuest or Google Maps or Google Earth? You'll feel like it's the first time again, when you use interactive dragging of routes on Google Maps. Some of the folks from the development team have even whipped up a handy video to explain the concept."

Re:Nice, Yes, But It's Not Amazing

[EvilGrin5000]

Just because they did their homework and programmed the original Google Maps with reusable code, doesn't mean that a new feature added that uses already implemented code isn't 'Amazing'.

Looking at the feature by itself is pretty nifty! Let's not judge it by saying "well, they didn't add any new code for this so it's nothing new..."

Re:Nice, Yes, But It's Not Amazing

[Yurian]

True, but that misses the point. Yes, it was possible before - in the same way that satellite imagery was available on TerraServer for years before Google Maps. The difference between possible and easy is all the difference in the world.

Re:Nice, Yes, But It's Not Amazing

[niteblade]

From a strict 'new technology' perspective: Agreed - nothing revolutionary. As an improvement to the way the average Joe can plans trips, done in a incredibly simple, intuitive fashion that non-techies will truly appreciate: Truly awesome.

-NB

Re:Bloat?

[geekoid]

I have wished they had this feature almost every time I use it.

Re:excellent feature

[wile_e_wonka]

That's interesting. I don't know about the rest of the world, but in the US Navteq provides the data for both Google Maps and Mapquest (and Yahoo! Maps as well, I believe). So you would think the directions would be no less accurate. On the other hand, I have definitely noticed the different services do often recommend different routes from the others, despite all being run by Navteq.

Anyone else know more about this?

Re:Not that big of a deal

[catbutt]

Given that ajax techniques were pretty much non-existant in 98, yep, it would have been considered innovative. Probably "nearly a decade ahead of its time" innovative.

I find it amusing (well, "annoying" is probably a better word) the way people, who have presumably never innovated anything of note in their lives, love to declare what other people have done to be "non-innovative". Why didn't you produce it for us in the time between 98 and now, if it was so easy and obvious?

Re:Bout time.

[Anonymous Coward]

Now they just need to figure out how to show rational text for things. For example, I just did a routing in California where you have no toll roads, but you have bridge tolls. They listed this as "partial toll road". I travel a lot, so I am familiar with toll roads in Texas, in Kansas, in Florida, etc. But to a Californian - partial toll road is just gibberish. Can't it just say "Bridge toll $4"? Hello? McFly?

Re:Bloat?

[thestreetmeat]

I'm driving from Ottawa to Moose Jaw this August. I would rather take this [google.ca] route rather than the default suggestion [google.ca], because I'd like to see my own country first, it's shorter in distance (although not in time), I have a place to crash in Winnipeg, and I won't have to worry about arriving in Chicago at the wrong time.

This feature is very useful. Before, to find out how long the trans-Canada route was, I had to make three separate trips and add the distances manually to get the total.

Advice for Google Maps, Mapquest Too

[Mana Mana]

Someone explain WHY THE FUCK Google Maps ruotes are STILL inferior (i.e., generally longer) to Mapquest's?!

In March 2006 I was raving about Google Maps to someone when she said: Their routes are always too long compared to Mapquests, no thanks. I thought, this chick is ignorant, turned out I was the ignoramus. This was in Loudoun County, VA. I am a born New Yorker, "no sleep till Brooklyn," boy and till 4 weeks ago I endured GM. Finally, I added a bookmark[1] to Mapquests driving directions page (their site is a blight, ads and obfuscation, hence I had never really tried them before compared to GM's Tufte-like simplicity/power-of-information-conveyance).

Go ogle stop with the add-ons -- Street View; Low-level-zoom edifice schematics; subway station locations; Drag & drop the blue line to customize your route; Avoid highways; My Maps; and most annoying to a (dial-up) user the new 'enhancement' of "Link to this page?" which I formally used to unstick/refresh-complete the graphic tiles when they become stuck and left me with missing route tiles OR to alleviate Google servers loading kajillion layers OR endlessly loading gad-knows-what!!! Dangit three bangs as this one really annoys. *sigh* -- and fix the basics!

+ Fast routes.
+ Accurate routes! In NYC I once had GM direct me from one part of Williamsburg to another via the Williamsburg Bridge![2]
+ No toll routes!!!
+ Total tollS calculation! Come on! I want to know if a route from 10013 to 20147 is worth ~$17.50 in tolls or an extra 90 minutes of my travel time?! And if I had a scenic choice, WTF, no brainer. That Jersy Turnpike is a boring mother! And why the fuck is the Battery Tunnel always a vital route to Manhattan from Brooklyn[4]. What, you California boys are like Bloomberg, out of touch? FYI, 99.99% of New Yorkers take the effing Hicks street approach to the Brooklyn or Manhattan Bridges. Granted, MQ does the same fuggin thing in this respect, but, at least they offer "Avoid This" in their options. Although it's buried under "Add a Stop" and they drop the Web 2 Ou sliding map for a fixed old timey one. Good grief, get back to work you Manhattan Portege carriers. My, this was a long item. ^.^ <-- New school. Fuck that: =)

+ Integrate traffic congestion into ETA. Further, allow entering departure time, incorporate it to realistically PROGNASTICATE ETA thusly? Using my GPS/navi unit I know to add 15 minutes when leaving my environs for the Holland Tunnel; Belt Pkwy East AFTER 3 p.m. on Mondays through Thursdays, and, 2 p.m. on Fridays. It's not magic, I have observed and can now predict the pattern. In New York City. But other routes' insights elude me, and Garmin tarffic covereage is limited to Metro areas and is not gratis. Geddit. Entre into meo corazao, de nuevo.

+ When are you Google guys going to ply Bloomberg, a la Sen. Patty Murray, for (D, WA, ah, ah 'nough clues? 8) for your 76 9th Avenue Google Lair influence? Come on, get with the MTA train (oh, pardon, Subway), Bus line mapping scheize already! I want to know how to get to Brooklyn Brewery's faux, nominal[3], beer facilities from Red Hook, hipster nouveau. BTW Red Hook beer for sale in Fairways is from New Hampshire, and WA I think. I know I know, almost got me too. Oh, it got you? Soright, try Sam Adams Cream Stout, it's what I imagined Guiness ought to taste.

+ Shortest distance routes.
+ Scenic routes!! How about beating Mapquest MEANINGFULLY and the AAA for once.
+ Remove the 100 entry barrier to my favorites routes -- this one made me buy a GPS/navi unit.
+ May be I missed it but: Use/make a standard way of dumping route data to XML so I can harrass Garmin into a way to import it. Garmin sucks in this regard. Pointers welcome, /.'ers?

Now I use MQ for routing printouts, GM for easy VISUALIZATION of destinations, and GPS mapping. <-- Here MQ sucks dick. WTF, low haging fruit.

But, seriously why the fig has not GM improved the route algorithms?

[1] http://www.mapqu

Re:Nice, Yes, But It's Not Amazing

[shellbeach]

People keep posting on other sites about how brilliant this is and how amazing Google's programmers are. Although I do really like it myself, all it does is make add another trip location where you click, and it's just a matter of calculating the route from the start to your point and from your point to the end, using the same stuff they've allready programmed. There's almost nothing new here.

The difference between a user-hostile application and a user-friendly application is most often not the skill of the programmer, but the intuitive ui design of the programmer. The fact that this functionality was simple to implement merely makes the solution more elegant.

Maps != Routes

[brunes69]

The map data has nothing to do with the routes. All Navteq sells to Google and Mapquest is a massive amount of vector-data that maps streets in geo-spatial coordinates. It is up to Google and Mapquest to determine the shortest path between any two points using this data.

It is more difficult than it sounds. Discovering the shortest path in a weighted map is a simple, well known algorithm that any third year computer science major would have studied. The problem is in the weighting. Things such as speed limits, number of traffic lights, road conditions, speed limits of intersecting roads, ourly traffic patterns - all of these affect the amount of time one route takes over another.

Aside from the fact that it is impossible to be up-to-date with this data on a constant basis, some of it changes based on the time of day of your planned trip. For example your morning "shortcut" to work may not be any faster on the weekend when the main route is not as congested.

I think in general, all the mapping sites to a remarkable job given the data they have access to. It is highly unlikely ny one site is "more accurate" than the other picking routes all of the time. What is probably happening is the place where you are going has some factors that have changed recently, or have not been acounted for, in one site vs. the other. You would for certain be able to find counter-examples that make the other site look better at other places in the country.