|
From: Paul G. <gk...@co...> - 2007-02-01 17:32:46
|
Hello All, I wrote about 4 weeks ago about a "problem" I'd encountered in fix.dat, specifically that there are many duplicate waypoints with vastly different coordinates. I cited as an example CAT05, located both in El Salvador and in Sicily... I received a response from a friendly fellow who has a copy of the most recent DAFIF data. These are the data which are 'cooked' down into, among others, fix.dat. He sent me a clip showing how the original data 'names' several fixes with different names, but 'identifies' them identically. Unfortunately, the process of 'cooking' the data eliminates the waypoint names (which identify each waypoint without duplication) in favor of the waypoint identities (which results in many duplicates) in the resulting fix.dat file which is used by FGFS's autopilot waypoint selector. He also said that he uses a perl script to work around the duplicates by selecting the one that is closest to wherever he is. However, it seems that the waypoint code is not so clever, as it consistently seems to select the last of a series of identical waypoints (as listed in fix.dat), regardless of which is closest to where you are (or the last waypoint in a list that you are compiling, as applicable). I worked around this by 're-identifying' the western-hemisphere duplicates by inserting a '1' after the letters (e.g., "CAT05" becomes "CAT105"). However, I view this as a strictly temporary workaround (although FGFS cheerfully accepts 6-character waypoint IDs, it violates the official 5-character convention). I would prefer to see a better solution (a la Mr. Anderson) built into the autopilot waypoint selection code, or an option to allow the user to select one from a list, or put up a list while auto-selecting the closest (particularly if two identically ID'd waypoints are equidistant +/- some configurable number of miles)... (on a side note, I'd like to see all text inputs converted to all-caps (as appropriate) so that a keyboard entry of "cat05" will match fix.dat's "CAT05", but that's another issue...) So with all that said, I've looked at the list of maintainers, and I am unsure which I should contact as there are many who've had their hand in the autopilot code. Who should I be talking to about my concerns, ideas, etc.? Paul Glanville |