Active Topics
-
Is there a section to talk about Java ME? Apps, etc. (0)
to General by Kalatti - 3 days, 6 hrs ago -
Extra softwares in Sailfish using CLI, repositories, etc (117)
to SailfishOS by nieldk - 5 days, 3 hrs ago -
Firefox with Leste (6)
to Maemo 7 / Leste by teroyk - 6 days ago - more...
 |
|
2006-05-29
, 23:04
|
|
Posts: 1,245 |
Thanked: 421 times |
Joined on Dec 2005
|
#62
|
Originally Posted by Ceklund
That limitation probably applies only to connecting the Nokia 770 to bluetooth-enabled phones (for things like internet access and file transfer). It definitely is not a limitation in the bluetooth specification, and as far as I know it's not a hardware or software limitation in the bluetooth implementation used in the Nokia 770.
It says in the user's manual that you can only have 1 Bluetooth device active at any one time. So how would you have the GPS receiver and the phone connecting you to the Internet both be active at the same time? This has me hopeful that the manual was wrong... if you guys are doing both simultaneously, then it can be done!
woohoo!
 |
|
2006-05-29
, 23:14
|
|
Posts: 78 |
Thanked: 1 time |
Joined on Sep 2005
@ San Francisco, CA
|
#63
|
Originally Posted by gnuite
Thanks for looking at this. I was able to dump data off my Socket GPS using this command from my Mac:Could anyone with a Socket GPS receiver please verify (with GPSD or rfcomm/cat) that their receiver does in fact emit the RMC sentence of the NMEA protocol? I can't find any information on the web about the messages it sends. The closest thing I can find is a datasheet for some "Trimble Recon GPS Card edition" that claims that it emits RMC and that it includes the "Socket Communications Bluetooth kit", which is a pretty weak connection.
You can check by looking at the NMEA output (with GPSD or rfcomm/cat) and checking if any of the output lines starts with "$GPRMC". If not, it would be helpful if you could tell me all of the message types that the Socket GPS receiver emits (such as the $GPGGA or $GPGSV messages).
Code:
sudo cat /dev/cu.SocketBTGPS-1 > testfix.txt
I did two tests, one where I know I had a fix, and one where I did not have a fix. I AM seeing lines beginning with $GPRMC in both. I have PM'd you both text files (the one where I had a fix ends in fix).
__________________
Nokia N800, 4GB ADATA SD (Class 6)
Nokia 770, 1GB Sandisk RS-MMC
LG CU500 for 3G data service (uses my iPhone SIM)
http://www.markandjo.com/markblog
Nokia N800, 4GB ADATA SD (Class 6)
Nokia 770, 1GB Sandisk RS-MMC
LG CU500 for 3G data service (uses my iPhone SIM)
http://www.markandjo.com/markblog
|
|
2006-05-29
, 23:18
|
|
Posts: 3 |
Thanked: 0 times |
Joined on May 2006
@ Cold
|
#64
|
Originally Posted by Ceklund
Works great. Have a BT Gps giving location and a 3G phone getting maps, no prob.
It says in the user's manual that you can only have 1 Bluetooth device active at any one time. So how would you have the GPS receiver and the phone connecting you to the Internet both be active at the same time? This has me hopeful that the manual was wrong... if you guys are doing both simultaneously, then it can be done!
woohoo!
|
|
2006-05-30
, 05:38
|
|
Posts: 114 |
Thanked: 11 times |
Joined on Oct 2005
|
#65
|
Originally Posted by hula
I also notice that in some google maps there are a few glitches......
After playing with google maps for a while, I noticed that the gps position often has a systematic error in some direction. It would be useful to be able to "nudge" the map to some known reference point (a suggestion would be by dragging the waypoint or gps dot to the reference point or something - similar to Atlas for the Palmpilot).
...
On the information pane you're considering, would it be possible to include the lat/lon coordinates for the currently shown map center? It would make it possible to verify some rough validity of any homegrown map without having to actually go there.
sometimes the road, and driving route consistently don't match.
I assume that this is an error in the maps coordinates.
SUGESTION:
It would be nice to mark "anchor" points. that way, whe just go somewhere, and manually select the exact position on the map, "where we really are". This would enable the calibration of the maps.
I guess 1 point could provide a simple fix, but 3 calibration points would create a much more accurate representation.
Also, for the "Information Panel", that has been mentioned, I sugest a simple and clean look, similar to the one found on the Audi Navigation systems.
http://shows.autospies.com/gallery/A....jpg?undefined
They use a simple black (traslucid) vertical bar on one of the sides.
I would sugest the following data: (from top to bottom)
Clock | Compass | Speed | Sat Fix (number) | Lat, Long | Distance to Next Change of direction + pictogram | Map Scale
|
|
2006-05-30
, 11:14
|
|
Posts: 3 |
Thanked: 0 times |
Joined on May 2006
@ Cold
|
#66
|
Originally Posted by lmf
Yes, multiple reference points is always better, but...SUGESTION:
It would be nice to mark "anchor" points. that way, whe just go somewhere, and manually select the exact position on the map, "where we really are". This would enable the calibration of the maps.
I guess 1 point could provide a simple fix, but 3 calibration points would create a much more accurate representation.
- the tiles are pretty small so reference points will most likely be on different tiles
- deviation need not be the same for all tiles at a given zoom
- keeping track of reference points requires that you have a map scope larger than the view port requiring much more calculations
- calculating correction based on multiple point's relation to eachother also requires quite a few calculations.
So I think a simple thing, like:
- Take a reference point (x, y, dx, dy), where x,y is location and dx,dy is correction.
- Reference points should adjust the map only locally, so they have an effective radius within which they add to the correction, say r (the equivalence of screen width perhaps?). This also means that only points within r distance from the viewport center will be included in the calculations.
- Within the radius they also have a diminishing effect, linearly decreasing. If the distance to the center point is d, that would mean corr_x = dx * (r - d) / r.
- Multiple points add to the equation by calculating the average corr_x for all included points.
The maths should work equally well in pixelspace so in theory you could probably do this with integer math.
Originally Posted by lmf
Also, for the "Information Panel", that has been mentioned, I sugest a simple and clean look, similar to the one found on the Audi Navigation systems.
http://shows.autospies.com/gallery/A....jpg?undefined
They use a simple black (traslucid) vertical bar on one of the sides.
I would sugest the following data: (from top to bottom)
Clock | Compass | Speed | Sat Fix (number) | Lat, Long | Distance to Next Change of direction + pictogram | Map Scale
Since you rarely need all those at once, I would prefer to be able to choose between a few sets of (much smaller) panes with different kinds of information. When driving, very few care about lat/lon, for instance. The totally uncluttered map view is one of the things I really like about this program.
|
|
2006-05-30
, 13:21
|
|
Posts: 182 |
Thanked: 3 times |
Joined on Mar 2006
|
#67
|
I have a question about the advanced notice for flite (thanks for that feature by the way), how much does each bar represent? One second each?
 |
|
2006-05-30
, 15:30
|
|
Posts: 3,096 |
Thanked: 1,525 times |
Joined on Jan 2006
@ Michigan, USA
|
#68
|
I recieved no response from the last time, so I will try again. Any chance maemo-mapper will work with TerraServer, they have better resolution than google maps and appear to be free to use and distribute, see below.
http://terraserver.microsoft.com/about.aspx?n=AboutFaq
Are there any restrictions on what I can do with the images that I download?
The images from the U.S. Geological Survey, and are freely available for you to download, use and re-distribute. The TerraServer team and the USGS appreciate credit for their work on this project by displaying the message "Image courtesy of the USGS".
http://terraserver.microsoft.com/about.aspx?n=AboutFaq
Are there any restrictions on what I can do with the images that I download?
The images from the U.S. Geological Survey, and are freely available for you to download, use and re-distribute. The TerraServer team and the USGS appreciate credit for their work on this project by displaying the message "Image courtesy of the USGS".
|
|
2006-05-30
, 15:37
|
|
Posts: 373 |
Thanked: 56 times |
Joined on Dec 2005
@ Ottawa, ON
|
#69
|
Originally Posted by tigert
If it is a choice between a high CPU fancy rubber-band selection or a simple low CPU remember the last two taps, I would be very happy sticking with this existing two taps solution. It is much better than those fields being left blank.I also think the "select rectangle" thing is the best way to select downloaded area.
The reason I wish it worked in the device, instead of google earth etc, is because I am carrying the device around more often than a laptop, and it would be nice to be able to select the downloaded area when under wifi connection like in a coffee shop, before going out for a walk etc.. I do have a phone with 3G connection, but it eats a lot of battery when connected, so it is useful to be able to fetch the maps beforehand.
Would the drag-panning-with-stylus be too much work for the CPU to handle by the way?
//Tuomas
In my experience, the current usability problem with the current approach is that the map scrolls when you are trying to make your selection ... often times scrolling your intended second point off the screen. The solution is to zoom out to a bigger area and then manually change the zoom selection back to your intended scale to download.
It would be great if the selection box idea could be done efficiently but I am pretty happy with the esisting simple approach.
One feature that I could like is a reverse of this "Download Area" though ... call it a "Purge Cache of Area". I am finding that there is the odd map that gets downloaded from Google Maps that is in the wrong scale (I believe this is Google's fault as it seems to cough up random things when the server is busy ... a "Refresh Area" option or "Ignore Cache" checkbox would solve this too) or I want to clear the cache of the detail of a city that I have finished travelling in there is no easy way of doing so other than manually clearing files with the file manager. I know this is fairly simple to do since the file locations have some corelation but it is still a bit labour-intensive.
|
|
2006-05-30
, 17:25
|
|
Posts: 1,245 |
Thanked: 421 times |
Joined on Dec 2005
|
#70
|
Originally Posted by kutibah
The bars don't really represent any concrete unit, but the advance notice _does_ depend on your speed, so it is definitely temporally based. 0.3 to 0.7 seconds per bar is probably not a bad estimate. The exact formula to define how much advance notice (in distance) that you get responds linearly to both the customizable "advance notice" variable and your speed.
I have a question about the advanced notice for flite (thanks for that feature by the way), how much does each bar represent? One second each?
Last edited by gnuite; 2006-05-30 at 17:28.
woohoo!