Bicycle Travel > GPS Discussion
GPS Data Wish List
Fred Hiltz:
Jennifer asked in the topic just below this for our thoughts on improving the GPS data that Adventure Cycling publish. That thread grew rather large and unwieldy, so I have started this wish list with several points from that thread.
Please take a look at Jennifer's introductory post to see what she is looking for, then comment on these points or add more.
In the order of their appearance:
1. Don't change too much.
2. Start and end the sample routes at the edges of maps where possible.
3. Include contact information in the waypoints marking points of interest.
4. Include routepoints for follow-the-road navigation by GPSRs that do not calculate this on their own.
5. Put the waypoints that mark points of interest and those that mark the riding route in separate files. Include routepoints in the latter file.
Fred
mdxix:
I vote for 5 & 6. In my view, it is essential to keep the data sets independent, modular, and compatible with many platform.
I thought the blog that Jennifer wrote was very promising. She made statements about "robust database", "mobile mapping solution is not if, but when", & "waypoints can be kept up to date easily and in an unobtrusive way". I could not agree more.
On the several trips that I have taken on four ACA routes, I never used the files as is. Nor have I read in this forum that someone simply loaded the files as is in their GPS unit, regardless of make and model.
That is, the more these files are easy to handle, manipulate, and use on various tools the better. Hence, I add:
6. Keep open & accessible across multiple platforms.
In other words, I would rather there not be a preference toward a few models versus others. People are using all kinds of technology that is advancing every year, including mobile.
Instead, I prefer to see the files use open standards that are modular working on multiple platforms. Create, for example, a quality check list that includes:
* Test conversion to multiple formats and various online mapping tools (edited 24aug2012)
* Test on multiple GPS units (recruit some test volunteers from the membership community)
* Test the file on a mobile device (test volunteers can help)This applies to both routes & waypoint—but keep them separate ;)
Best of luck to your team. This is a big effort. Thank you for your time.
tsteven4:
Here are some nits I would wish for assuming that the data continues to be distributed in gpx format:
7. Add elevations (gpx ele elements) for all waypoints (gpx wpt elements) and routepoints (gpx rtept elements, but not necessarily gpx extensions rpt elements). Today I toss all the ele elements and fetch them from http://gisdata.usgs.net or if that fails from http://ws.geonames.org and build a SQL database of elevations. This gives me elevations for all points, but the validity is problematic on bridges and tunnels as they fetched points will be on the surface, which in these cases is not where the route is. The elevation data is useful in displaying profiles, although because of the sparsity of the gpx rtepts they are rough. If your software cannot fetch the elevation data I could help with software but in my experience this is non-trivial and requires some maintenance as the web servers and protocols change with time.
8. Delete all times, i.e. (gpx time elements). I don't think these are meaningful for the ACA data and they can cause difficulties when viewing the routes in some tools, e.g. google earth. I could provide software to this if you don't want to take care of it upstream of the gpx file creation.
9. Delete all gpx extension Depth elements. There are only two in the entire set today, and they are used incorrectly. I could provide software to this if you don't want to take care of it upstream of the gpx file creation.
10. Validate all distributed gpx files. While I haven't found invalid gpx files from ACA, invalid files are a common cause of issues reported on the gpsbabel mailing list. I can supply software to do this. Note that this step doesn't change anything, it just guarantees that the gpx files that are about to be distributed obey the rules for gpx files as defined in the appropriate XML schema documents (xsd files).
11. Consistently organize and name the folders and files for each route, including case and blanks in directory and file names. These makes automated processing easier.
And one that I consider more important:
12. Have contiguous route segments, gpx rte elements, share a route point (gpx rtept). This make display of the overall route appear as one, instead of a bunch of separate and disconnected segments. Accidental violations of this are the most common reason I write Jennifer.
And one major request:
13. Continue to provide public access to all the gpx data, including the points of interest and route data. There was an article in one of last years magazines that stated that points of interest data would be restricted to members in the future. To me it seems restricting access to the points of interest is in conflict with the ACAs mission "Adventure Cycling Association inspires and empowers people to travel by bicycle." I am concerned that a revised policy on the waypoint data might prevent me from providing online viewing tools and google earth files to the public, and I certainly believe the maps at http://tsteven4.qwestoffice.net/ inspire and empower people to travel by bicycle.
I question 5, the separation of the point of interest (wpt) and route (rtept and rpt) data. I see this as a violation of 6, "keep open and accessible across multiple platforms". I think the separation is convenient for some applications, e.g. mdxix seems to want this, but I am not convinced this is generally desirable. I would prefer they remain together.
I am still unsure of what Jennifer meant by "Do you like/use the sample routes provided?". Can you cite an example of a sample route? In 3 Fred seems to imply a sample route just corresponds to the route as shown on a paper map panel.
Thanks for all your work and including us in the conversation.
Fred Hiltz:
--- Quote from: tsteven4 on August 23, 2012, 06:19:11 am ---... assuming that the data continues to be distributed in gpx format
--- End quote ---
This is a given. We have no intention to abandon GPX. When the project began, we distributed two other formats as well, attempting to cover the most popular of the many proprietary GPS data programs. Now we have standards, and the wonderful thing about standards is that there are so many of them. For practical production reasons, GPX and its Garmin extensions are likely to remain our mainstay. KML? Maybe, if enough members want it for bicycle navigation, this would be worth spending staff time (=money) on.
--- Quote from: tsteven4 on August 23, 2012, 06:19:11 am ---11. Consistently organize and name the folders and files for each route, including case and blanks in directory and file names.
--- End quote ---
The File names section of the GPS Data User Guide describes our convention for naming the GPX files. If you find some that do not follow it, please let us know of the oversight. If we implement WL #5, we would add some indicator of routes vs. services.
The distributed data contain no folders. A single Zip file contains all the data for one ACA route.
--- Quote from: tsteven4 on August 23, 2012, 06:19:11 am ---12. Have contiguous route segments, gpx rte elements, share a route point (gpx rtept). This make display of the overall route appear as one, instead of a bunch of separate and disconnected segments. Accidental violations of this are the most common reason I write Jennifer.
--- End quote ---
You are right, those are accidental. We intend to make the last waypoint defining a GPS route and the first waypoint defining the next be the same.
--- Quote from: tsteven4 on August 23, 2012, 06:19:11 am ---I am still unsure of what Jennifer meant by "Do you like/use the sample routes provided?". Can you cite an example of a sample route? In 3 Fred seems to imply a sample route just corresponds to the route as shown on a paper map panel.
--- End quote ---
From the GPS Data User Guide: "We include GPS routes in the published file to make the ACA route stand out on the map. They might also serve as sample routes to help you create the actual routes for a day’s riding." We have built these from 25 - 30 waypoints in order to fit as many as possible into limited GPSRs. WL #2 would improve on that.
Fred
tsteven4:
--- Quote --- For practical production reasons, GPX and its Garmin extensions are likely to remain our mainstay.
--- End quote ---
GPX is great.
--- Quote ---The distributed data contain no folders. A single Zip file contains all the data for one ACA route.
--- End quote ---
I was referring to the files and folders within the zipfile for each route.
Some examples of small inconsistencies are:
1) the sierra cascades gpx files are in the root directory of the zipfile, while those for all others are in a sub-directory,
2) the sub directories are usually named something like "UC01v004 Folder", but there are two cases where the directory name does not include " Folder", these are AM01v002/AM01v002.gpx and GM01v003/GM01v003.gpx
3) Sometimes Folder uses an upper case F and sometimes a lower case f.
4) usually the gpx files use a lower case v, but the following use an upper case V instead:
GD01v009 Folder/GD01V009.gpx
GD02v009 Folder/GD02V009.gpx
GD03v009 Folder/GD03V009.gpx
GD04v009 Folder/GD04V009.gpx
GD05v009 Folder/GD05V009.gpx
GD06v009 Folder/GD06V009.gpx
WE01v008 Folder/WE01V008.gpx
4) the pdf files mix the case of the v as well.
5) the great divide pdf file has an extra digit, GDAboutV0011.pdf
I will grant you that all of these issues are nearly insignificant, but consistency would make automated processing of the data easier.
Navigation
[0] Message Index
[#] Next page
Go to full version