I just trashed the Google Maps team on their blog

[Stanovoy]

I finally lost it. I hate G1's Google Maps so much that I went out of my way to seek out the blog the Maps team uses, and attacked the members on their poor design decisions. :(

I think what pushed me over the edge was their audacity to call "re-centering the map when My Location reaches the edge" a feature! This is the "feature" I have been having trouble with. I simply couldn't imagine how a human being can become so stupid as to think that is a good way of doing things. :mad:

I would've had no problems with them if they said, "the G1 couldn't handle constant re-centering. So we settled for this option." I also would've had no problems with them if they said, "we didn't get to constant re-centering yet because we needed the time to work on something else." But to call that a feature and brag about it?? Just how stupid are these developers?? Are they basically a bunch of socially inapt outcasts who can code very well? Can't Google hire someone to oversee the usability portion of the software?? :mad:

At the end of the comment, I described Google Maps for the G1 as garbage, much like a car with a loud muffler but no engine. In view of their other versions on devices such as BlackBerries, I don't think that analogy was over the top.

I hate Google Maps for the G1. I used to love Google Maps dearly... :(

Here's the blog: http://googlemobile.blogspot.com/2008/10/google-on-android-maps.html

[Cow_King]

lol your crazzzzzzzzy

[PixelHenge]

no, really see my signiture?? Im right there with you about those two things and more! Where did you leave your comment? i would love to leave one too. PM me the link if you dont want to post it here.

heres my thread about issues with google maps http://androidcommunity.com/forums/f12/devs-lost-map-pins-and-locations-reward-if-found-4692/

and heres another with even more complaints: http://androidcommunity.com/forums/f41/some-google-maps-questions-for-those-with-android-4512/

and yet another: http://androidcommunity.com/forums/f3/still-no-exact-adress-in-street-view-5549/#post58567

i truely think all of these things should have been addressed pre-production as google is indeed the maker of google maps- this shouldnt be left up to 3rd party devs.

[Stanovoy]

Here's the blog I talked about: http://googlemobile.blogspot.com/2008/10/google-on-android-maps.html

[PixelHenge]

Read your comment- very good, i would have to agree- i wouldnt even need turn by turn if it would recenter automatically.

[Stanovoy]

The pins are important too. Every weekend, I have a lot of events to go to. When I had my BlackBerry, I would type all the locations in so I will always have them on my Google Maps. In addition, that will also give me an idea as to the best way to attend all of them. With the G1's version, I'm back to pen and paper. If I were the boss of Google Maps, I would fire each and every single one of those morons. Honestly, while getting Google Maps to work through a browser is a challenge, getting it to work through an application is a joke. So I'm guessing the guys on the app team are mostly brain dead idiots.

[PixelHenge]

well ya, i know.. so ok, its difficult to implemet.. ok, i get that, google... but why dose every other platform have these features then? I know i kind of imagine it went like this

"ok guys, we'er about to launch android, are you done?"
"yup"
"oh, wait, crap, we forgot to add maps to the new streetview bling"
"oh ya"
"well, never mind, just throw something on there, it doesnt have to be usefull cause you know, as long as you can find 'pizza' who cares? this is america, who uses maps?"
"oh, well i might..."
"too bad, time to launch it"
"oh, ok then.. heres a version of maps from 10 years ago"
"ya, use that"

I think people like us who ere expecting google maps to be usefull and functional on a google OS were maybe giving them to much credit for inter-department co-operation.

I mean, i know the maps has a history and such, but i have to sift through all the other things i looked at between then ast time i looked for "this thing" and now, plus ya, there are alot of places it cant find because its not listed in the directory or whatever, i ought to be able to mark it with a pin then. AND, it should tell me how to get there, and update and recenter the map as i'm going.

Anyway, i posted a comment there too.

[Stanovoy]

Quite honestly, I actually "conditioned" myself to lower my expectations and only expect Google Maps for the G1 to work like the one for the BlackBerry 8800 so I won't be disappointed. I was wrong.

I believe the Google Maps team actually went out of their way to make the re-centering on the G1 to work like that. Here's why:

Logic for "constant" re-centering:

if (location.refreshed == true) && (location.previous != location.current)
{
map.centeron(location);
}

Logic for "garbage" re-centering:

if (location.refreshed == true) && (location.previous != location.current)
{
coordinates = convert_to_screen_coordinates(location, map);
if (coordinates.x > screen.width) || (coordinates.x < 0) || (coordinates.y > screen.height) || (coordinates.y < 0)
{
map.centeron(location);
}
}

See how much more complicated it is to implement their garbage?

[PixelHenge]

Wow.. now you got the code up there....

I had thought from watching it function there was a lack of code for constant recentering... but no you got the logic up there...

just wow, thats... um, sad... like going out of your way to make something stunted. I woulda expected that from any other company except google.

Ya, i got to agree with you then, idiots. I cant even THINK of a way that this could be better than just recentering with movement. I mean, i really tried to imagine it from a marketing standpoint, and i simply cant. No better word here: dumb.

[PixelHenge]

you ever seen that movie "idiocracy"?

"extra big-a$$ tacos! now with more molicules!"... comes to mind here. hehe..

[ariesboy]

Think of re-centering in terms of battery management. 20-30(not exactly sure of the number here) refresh of the screen vs. 1. The latter being the "garbage" re-centering. Which do you think will eat up your battery?

The better option is if the allowed both options through settings. I dont think your garbage re-centering is garbage. It's useful if i'm plugged in and charging. I wouldn't need to worry about battery drain. But if i'm unplugged, battery conservation is king.

[Stanovoy]

Think of re-centering in terms of battery management. 20-30(not exactly sure of the number here) refresh of the screen vs. 1. The latter being the "garbage" re-centering. Which do you think will eat up your battery?

The better option is if the allowed both options through settings. I dont think your garbage re-centering is garbage. It's useful if i'm plugged in and charging. I wouldn't need to worry about battery drain. But if i'm unplugged, battery conservation is king.

Actually it doesn't quite work like that. The screen is refreshing at a constant rate. When you see an static image displayed on it, what is happening in the background is the same image being displayed over and over again. So, the only extra energy drain would come from the background calculations, not the refresh.

As for the refresh, the current algorithm has to go through the same number of steps as mine when not re-centering. It is true that my re-centering happens more often. However, Google Maps are based on tiles. Once a tile is downloaded, it is simply a shift in pixels. The pixel extraction and rendering are both constant. The only thing extra is adding a constant to the starting and ending positions. Thus, that is an extremely cheap operation.

Now, let's assume that the operation is actually a very costly one. Even if that were the case, an application that conserves battery power but does not serve one of its major purposes is fairly useless. Would you care for a flashlight application that doesn't light up the screen? How about an email client that only syncs once every week? I think the options should be constant re-centering and manual panning. I can see absolutely no purpose for the edge re-centering they have right now. In fact, because they must have code for it, that is probably a drain on system memory, which we have found out are not plentiful.

All in all, I believe the current implementation failed to serve the purpose it is designed to fulfill, and that the better alternative will only have negligible impact on battery life, if at all. Thus, the cost and benefit scale leans heavily toward the latter than the former in all scenarios.

[Stanovoy]

you ever seen that movie "idiocracy"?

"extra big-a$$ tacos! now with more molicules!"... comes to mind here. hehe..

I love that move! Why Brando? Because it has electrolytes! Do you even know what electrolytes are? They are what plants crave. How do you know that's what plants crave? Because Brando has what plants crave. :eek::eek::eek:

[thepaddedroom]

It'd be nice if they updated the maps. Tonight, the app couldn't find an address in a three year old neighborhood.

[PixelHenge]

awe man! thats like my favoritest movie ever! (next to huckabees and forbidden zone, oh and dead alive)

your kids are starving. Carls junior believes no kid should go hungry. You are an unfit mother, your children will be placed in the custody of carls junior... welcome to coscto i love you.. haha

Man, i cant sit in a carls junior without cracking up, than and fudd rutckers. :D :)

[PixelHenge]

Oh, i was thinking the same thing about the screen refrehes- technically more code would equal more resources being used, but it would be so totaly small a difference.. Agreed though, you should beable to turn my location with centering on or off. And ya, if i'm in a car, id have it plugged in anyway. I could only see it being a problem when i was biking, but i doubt that function would make much impact on battery use.

@thepaddedroom - yes, having the same problem, google maps online could even find some of the newerthings whereas the andoird apps couldnt.

[Stanovoy]

Yo judge, check this out. This guy, he is guilty as ****! Just look at him! First, he talks all like a fag. Second, we have got all... all this evidence. He didn't even pay his hospital bills! So here's what I say, your honor <slams fist on the desk> GUILTY! Peace!

Our friend stylishly spells his name as U-P-G-R-A-Y-D-D with two D's at the end, which stands for a double dose of pimpin'

[RegGuheert]

Think of re-centering in terms of battery management. 20-30(not exactly sure of the number here) refresh of the screen vs. 1. The latter being the "garbage" re-centering. Which do you think will eat up your battery?I agree that recentering likely uses significantly more battery power. Why? The cost for the battery has nothing to do with updating the screen, since, as Sanovoy has stated, it is the same in both cases. While you will download the same number of tiles in both cases, in the case of "garbage" recentering, you only need to transmit data (to request new tiles) once each time you reach the edge of the screen, not each time you come across any new tile. The biggest difference is in how much GPS activity is needed. If you are not following the user, then perhaps you simply estimate time to the edge of the map based on speed and then turn off the GPS until you approach that time. I don't know if this can really be done, due to the time it takes GPS to acquire the signal, but it seems like the only possible justification for the strange implementation that they have made.
The better option is if the allowed both options through settings. I dont think your garbage re-centering is garbage. It's useful if i'm plugged in and charging. I wouldn't need to worry about battery drain. But if i'm unplugged, battery conservation is king.Agreed.

One final thing I will mention is that in May I had a Blackberry 8830 with Sprint Navigation. While it was a GREAT feature, I found that the battery draw was SO high in some instances (like when it was not able to access the 3G network) that it would sometimes completely drain the battery EVEN WHILE PLUGGED INTO THE CAR CHARGER! I'm wondering if this was a consideration since T-Mobile's 3G network is somewhat non-existent many places. While the amount of data moved using a 2G radio is the same, the power consumption will be much higher since the radio would need to be on constantly. (Still, as Stanovoy has mentioned, the amount of data is not decreased by using the current, braindead, approach to navigation.)

[Stanovoy]

While you will download the same number of tiles in both cases, in the case of "garbage" recentering, you only need to transmit data (to request new tiles) once each time you reach the edge of the screen, not each time you come across any new tile.

I actually thought that was the reason for their implementation, but that turned out not to be the case. Let's pretend our screen can only accommodate 4 tiles (A1, A2, B1, B2, with A and B running across the top, and 1 and 2 running down the side). Constant re-centering of a location moving towards the right edge would cause A1 and A2 to be downloaded, and then at some later point, B1 and B2. Edge re-centering would have all 4 of them downloaded at once, as a huge stream. While the total amount of data downloaded will be the same, the difference appears to be the amount of data start requests and data end confirmations sent by the phone, with constant re-centering requesting at least twice as much. In a realistic Maps implementation, this could amount to 10s of times more. This would cause a more pronounced drain on batter.

With that said, I have come to the conclusion that this is not the case. Google Maps for the G1 is implemented to download tiles asynchronously and separately, similar to torrent files. One way you can see this for yourself is to scroll quickly in any direction so the entire map will be blank. Watch the load order. Each tile is loaded as soon as its download is complete. If a sequential data stream is used, then you will see the tiles loading in a particular order, much like JPG files that always appear from the top to the bottom. However, what you will see is an out of sequence loading pattern. In addition, the pattern varies from screen to screen. This means there has to be a separate data start request and data end confirmation for each and every tile.

In view of the above, the only conceivable difference is that the current implementation uses a batch data request. If so, the save on power would be marginal at best. The actual request is simply a short text sequence of tile IDs, e.g. "C1, C2, D1, D2". The data start and data end packets will still be needed.

To put everything into perspective, each cell phone broadcasts its unique identification number (not just the IMEI) every few seconds or so, depending on the previous signal strength. This number is likely to be far longer than any request Google Maps can possibly require, and sent at a frequency far higher than Maps will ever need. There is no way anyone will likely go through a set of tiles at that speed, unless you are fully zoomed in while driving on a highway.

So all in all, this current implementation not only makes no sense from a user standpoint, it also lacks merit from a technical standpoint.


The biggest difference is in how much GPS activity is needed. If you are not following the user, then perhaps you simply estimate time to the edge of the map based on speed and then turn off the GPS until you approach that time. I don't know if this can really be done, due to the time it takes GPS to acquire the signal, but it seems like the only possible justification for the strange implementation that they have made.

I think what you are referring to is GPS interpolation (GPS-I). Off the bat, Google Maps does not use GPS-I. Why? I don't know, it makes little sense. Let me explain what GPS-I is first.

If you know the direction of travel, the path taken, the acceleration of the object, then you can check on the exact location of the object and guess the rest based off that location. Of course, this not a perfect world, so there will be errors. That's why the location is refreshed every few seconds or so to make sure any possible errors are not compounded. If GPS-I is used, then we should be seeing a smooth moving dot, rather than the current jumpy one.

Generally, GPS-I is not used except in automobiles and military applications because it requires an accurate accelerometer, but more importantly, a very accurate compass. Accelerometer is prone to error compounding. So when there's an angular shift, i.e. change in direction, its precision will be negatively impacted. With an accurate compass, this error could be eliminated and the only the vector of the shift is taken into account, which is simply just acceleration. This allows accurate tracking of cars and missiles over uncharted areas.

On the other hand, the G1 has neither. While the accelerometer is not bad, the compass has been demonstrated to be slow acting and inaccurate. Thus, interpolation based on these mechanisms will not be very accurate. However, that does not end our inquiry.

First, any inaccuracy will be eliminated once the GPS is refreshed. So for 99% of time when the G1 is used as a guidance system, i.e. walking or driving, the inaccuracy will not affect the quality of the results significantly.

Second, Google Maps uses pseudo-GPS positioning. It takes the GPS reading, and snaps your location to the nearest road within a certain range. Test this out if you want. Go to a street corner, and stand on the sidewalk of one of the intercepting streets. You will see, that over time, you location will jump between the two intercepting street, even though you are clearly closer to one of them. Any inaccuracy produced by the interpolation will go through the same snapping algorithm. Thus, its effects will be significantly mitigated, i.e. you will not likely to find yourself off road all of a sudden.

Now that you understand what GPS-I is, I hope you can see why Google Maps is clearly not using it. Therefore, it is completely dependent on the actual GPS or triangulation of cell towers, and constant re-centering will not demand any more power than edge re-centering.

[Lukster]

There is a lot of room for improvement i agree.

[PixelHenge]

.....Second, Google Maps uses pseudo-GPS positioning. It takes the GPS reading, and snaps your location to the nearest road within a certain range. Test this out if you want. Go to a street corner, and stand on the sidewalk of one of the intercepting streets. You will see, that over time, you location will jump between the two intercepting street, even though you are clearly closer to one of them. Any inaccuracy produced by the interpolation will go through the same snapping algorithm. Thus, its effects will be significantly mitigated, i.e. you will not likely to find yourself off road all of a sudden......



I have a question then. Now i understand everything that you were saying, and agree- i know the g1 uses a pseudo-gps (or a-gps? {assisted}). But im currious about the 'snapping' you mention.

I have not found that to be true, if i understand correctly. This is just for curiositys sake and you seem like the one to ask.

ok- heres where the g1 gsp said i was:

http://androidcommunity.com/forums/members/pixelhenge/albums/randm-stuff/881-me.jpg

Or atleast thats about where my dot was in the android maps UI. Which is really rather spot on even down to the location in my 4 unit building (alittle off the the side of that street on the top). I must say, i was impressed being im in the middle of a military base and most of the t-mo cell towers are a ways away, and that from INSIDE.

But im just curious as i havnt seen it snap, but your explaination makes sence.

[Stanovoy]

I have a question then. Now i understand everything that you were saying, and agree- i know the g1 uses a pseudo-gps (or a-gps? {assisted}). But im currious about the 'snapping' you mention.

I have not found that to be true, if i understand correctly. This is just for curiositys sake and you seem like the one to ask...

The snapping behavior is built in to compensate for the inaccuracies in either the GPS or the digital map itself. You can detect GPS inaccuracies by zooming all the way in and stand still. You will see your dot shifts a tiny bit once in a while. This is due to the movements of the satellites. For map inaccuracies, if you switch to satellite view with "Show labels" checked and look around, you should be able to easily find locations where the digital map seems to be off.

What we can conclude from this is that the targeted errors are very small. That is why you have to be very close to the road, i.e. on the sidewalk or literally on the curb. Instead of displaying you as a few feet off the road, Maps will snap you on to the road. Also, the faster you travel, the more range will be given to the snapping algorithm. So let's say if you are traveling each bound on a highway at 60mph. The GPS reading might put you just north of the west bound lane, but Maps will snap you dead on to the east bound lane. The reason for this is the time difference between satellite readings. It takes a minimum of 3 satellites for triangulation. When you are going at a fast speed, the time gap between reading of the first satellite and that of the last can be significant enough to generate a location-translation error. That is compensated in various ways, but the simplest is to allow a wider range for snapping.

As for your image, your location is too far to be in the "snapping range." You would have to get much closer to a road to see your location being snapped onto the road. The easiest way is to drive around on a highway that shows the two sides separately, e.g. Route 115 right next to you. You will constantly be placed perfectly on the side of the highway you are actually on, throughout the trip. Now, pull over to the shoulder. Wait for a while, or even start walking against traffic. Then, depending on the accuracy of the map at that location, you might see yourself suddenly "jumping" to the opposite side of the road.

Finally, A-GPS is not pseudo-GPS. This is a commonly made mistake because of the similarity between A-GPS and a-GPS. Pseudo-GPS, or a-GPS (asynchronized-GPS), is based purely on cellular tower triangulation. This is the type you see in phones without GPS chips but running Google Maps. On the other hand, A-GPS (assisted-GPS) is based on satellite positioning assisted by cellular tower triangulation for faster initial position acquisition and cellular servers for faster satellite lock. There are two types of A-GPSs. The first type is what we have in the G1 and most cell phones. It works as I have described. The second type is used for high precision guidance needs such as air traffic control and military avionics. It operates like our phones with higher precision hardware. In addition, it can accept signals from ground based positioning beacons to pinpoint your location down to inches. We obviously don't and won't have access to that. Otherwise, I'm sure those damned terrorists will rig a car with it and you can imagine the rest...

[PixelHenge]

The snapping behavior is built in to compensate for the inaccuracies in either the GPS or the digital map itself. You can detect GPS inaccuracies by zooming all the way in and stand still. You will see your dot shifts a tiny bit once in a while. This is due to the movements of the satellites. For map inaccuracies, if you switch to satellite view with "Show labels" checked and look around, you should be able to easily find locations where the digital map seems to be off.

What we can conclude from this is that the targeted errors are very small. That is why you have to be very close to the road, i.e. on the sidewalk or literally on the curb. Instead of displaying you as a few feet off the road, Maps will snap you on to the road. Also, the faster you travel, the more range will be given to the snapping algorithm. So let's say if you are traveling each bound on a highway at 60mph. The GPS reading might put you just north of the west bound lane, but Maps will snap you dead on to the east bound lane. The reason for this is the time difference between satellite readings. It takes a minimum of 3 satellites for triangulation. When you are going at a fast speed, the time gap between reading of the first satellite and that of the last can be significant enough to generate a location-translation error. That is compensated in various ways, but the simplest is to allow a wider range for snapping.

As for your image, your location is too far to be in the "snapping range." You would have to get much closer to a road to see your location being snapped onto the road. The easiest way is to drive around on a highway that shows the two sides separately, e.g. Route 115 right next to you. You will constantly be placed perfectly on the side of the highway you are actually on, throughout the trip. Now, pull over to the shoulder. Wait for a while, or even start walking against traffic. Then, depending on the accuracy of the map at that location, you might see yourself suddenly "jumping" to the opposite side of the road.

Finally, A-GPS is not pseudo-GPS. This is a commonly made mistake because of the similarity between A-GPS and a-GPS. Pseudo-GPS, or a-GPS (asynchronized-GPS), is based purely on cellular tower triangulation. This is the type you see in phones without GPS chips but running Google Maps. On the other hand, A-GPS (assisted-GPS) is based on satellite positioning assisted by cellular tower triangulation for faster initial position acquisition and cellular servers for faster satellite lock. There are two types of A-GPSs. The first type is what we have in the G1 and most cell phones. It works as I have described. The second type is used for high precision guidance needs such as air traffic control and military avionics. It operates like our phones with higher precision hardware. In addition, it can accept signals from ground based positioning beacons to pinpoint your location down to inches. We obviously don't and won't have access to that. Otherwise, I'm sure those damned terrorists will rig a car with it and you can imagine the rest...

Oh, wow... See, i had no idea A-GPS and a-GPS were two differnt things. I had been readin about the g1 since it's anouncement and the iphone before and both were discribed as having a-gps.. i googled a-gps and have since been confused. That would explain why :rolleyes:

Yes, i did notice my dot moving around slightly. I figured maybe it had something to do with the weather or what have you, but now i know, thank you!

Also, when you had talked about snapping to streets, hadnt realized you had meant only with in range- that makes alot more since now too.

Can you tell i've never had a GPS unit of any kind before the g1? :o Not to mention, i dont think i ever even got it out of the house save maybe once and looked at the maps before it died on me. I had the 1st gen iphone, and it would find me down to a radius, but i was well aware that that location was based off of my range to the cell towers. I did know that the g1 uses both that and satellite data- but the actual workings and such werent very clear.

Ok, ya, wow.. It's 4am, i am deffinently going to have to reread this when i'm more awake.

But may i ask, what do you do for a living? Code, satellites... Dont tell me you have to kill me if you answer that ;)

[Stanovoy]

But may i ask, what do you do for a living? Code, satellites... Dont tell me you have to kill me if you answer that ;)

Lol, well let's just say I was with one of the armed branches and had a position where my life would end if my GPS unit went dead or out of calibration. After that, I spent a lot of time at the Pentagon working with a top secret project that made extensive uses of both aerial and ground GPSs (I had TS/SCI). Now, I work on the legal side for a company whose entire product is based on the ability to determine a user's location :)

I'll let you know which company it is once our product is released for Android in Q1 of 2009 :)

That said, this is why Google Maps is so frustrating for me. Give me the damn source code and I can make it so much better in a week or so! I guess my company can write up an overlay, but that would take too much time away from our own software :p

[PixelHenge]

Lol, well let's just say I was with one of the armed branches and had a position where my life would end if my GPS unit went dead or out of calibration. After that, I spent a lot of time at the Pentagon working with a top secret project that made extensive uses of both aerial and ground GPSs (I had TS/SCI). Now, I work on the legal side for a company whose entire product is based on the ability to determine a user's location :)

I'll let you know which company it is once our product is released for Android in Q1 of 2009 :) <----I'll hold you to it...

That said, this is why Google Maps is so frustrating for me. Give me the damn source code and I can make it so much better in a week or so! I guess my company can write up an overlay, but that would take too much time away from our own software :p

Well, that first bit would describe a whole lotta military ;) BUT, ive had way too much army for one lifetime, so i'll leave that there. But the legal side, pentagon? Didja get to play with aliens?!!? :D kidding sorry.. 4am.. ill stop..

That would certainly explain all the above posts though, thank you.

So anyway, they released the sorce code for android and not the maps app? Thats just silly. Well, if you ever feel like having a lil side project that might envolve map pins, um, i do laundry and wash dishes!!

[Stanovoy]

Well, that first bit would describe a whole lotta military ;) BUT, ive had way too much army for one lifetime, so i'll leave that there. But the legal side, pentagon? Didja get to play with aliens?!!? :D kidding sorry.. 4am.. ill stop..

That would certainly explain all the above posts though, thank you.

So anyway, they released the sorce code for android and not the maps app? Thats just silly. Well, if you ever feel like having a lil side project that might envolve map pins, um, i do laundry and wash dishes!!

Google only released the documentation for a portion of Google Maps' API. The application itself is closed to the public. The publicized API is not complete. However, with enough time, dedication and talent, one can recreate and have access to the full API. Google is simply not helping you with it :) While it is possible to recreate the app itself (which is actually very simple), the drawback would be another layer of complication, since it's impossible to capture the full functionalities of Google Maps without either 1) modify the Android OS to replicate the close integration, or 2) overlay new functionalities on top of the existing Maps app. Needless to say, neither is ideal.

[Frito]

Hey guys i agree with you on google maps needing auto centering and pins n such but you do realize that the program andnav does this centering and in general is allot easier to use than google maps while driving, zoom in/out as well as other buttons on screen all the time, on screen keyboard for putting in addresses
it needs some work like most/all apps out so far but its definitely allot better than google maps atm on the G1,

oh and did i mention it has a compass mode for orienting the map? very cool, though it does show how the compass in our phone is very inaccurate it seems to work well heading north but heading south it very inaccurate might be the phones screen blocking earths magnetic field im not sure but its interesting to say the least

[remister]

I can't wait for Google Earth to get on our G1's!

[tinareeahelp]

Hi Everyone,Just found this site and thought it might be useful to chat to people who are trying to accomplish the same as me! I spend too much of my life on the internet and look forward to chatting to you all and picking up and sharing ideas along the way!Tina

[Iceberg]

Hi Everyone,Just found this site and thought it might be useful to chat to people who are trying to accomplish the same as me! I spend too much of my life on the internet and look forward to chatting to you all and picking up and sharing ideas along the way!Tina

You DTF Tina?
har har har
Spam