by tdammers » Mon Mar 11, 2019 10:52 pm
@jomo, re the yellow glider problem with the CRJ700/900 at the last event. I've done some additional experimenting and research, and found a few interesting things. But before I go into details, I'd like to clarify that this isn't to blame anyone or anything, I just wanted to satisfy my curiosity, and thought I'd share my results, for future reference.
So first, some observations.
- In the video, there's another pilot flying a CRJ700, and their model also shows as a yellow glider.
- OpenRadar shows my aircraft type as "CRJ9". The other CRJ however is shown as "CRJ700" (full model name, not ICAO designator).
- At the end of the session, when we tried a couple combinations, there were three differences between the last failed attempt and the successful one: 1. I had been using mpserver01 for the failed experiments, but mpserver03 for the successful one, and so I initially suspected mpserver01 to be the culprit. 2. For the successful run, I used the CRJ700, while the unsuccessful ones were done with various other models from the CRJ700-family (900, 1000, 1000ER). 3. Between the unsuccessful runs and the successful run, you apparently deleted your CRJ700 install, and then installed the CRJ700-family from github.
Armed with this, I tried to replicate the problem, using mpmap03.flightgear.org as a convenient way of checking the aircraft ID as sent over FG MP. Since I suspected mpserver01 to be the culprit, somehow truncating the aircraft name from "CRJ900" to just "CRJ", I logged onto MP in various ways, exactly as I had done during the event - however, no matter which MP server and aircraft model I chose, they all showed correctly. So I had to dismiss this theory - neither the 700 vs. 900 difference, nor the mpserver01 vs. 03 / 19 / ... change made any difference.
And then I remembered that there are hard-coded paths inside the CRJ700 family, which means that installing it into a directory other than "CRJ700-family" could cause it to not load correctly. I also remembered that yours had been installed into "CRJ700", not "CRJ700-family". So that is the probable cause of the problem, as far as I can tell. And further, it's possible that those paths inside the CRJ aircraft that referenced the "CRJ700-family" directory weren't there in older versions of the aircraft, which would explain why it worked a few weeks ago, when I last visited: without those references, it doesn't matter how you name the directory, as long as it's directly under your aircraft directory.
So where does the "CRJ9" come from? Simple: that's what I entered on alwaysdata when filing the flight plan. IRL, you would use ICAO designators for the aircraft type, and most flight plans I see on alwaysdata do exactly that - you file "B772", not "777-200" or "Boeing 777-200", and likewise, you file "CRJ9", not "CRJ900" or "CRJ-900". OpenRadar, then, automatically fetches the flight plan from alwaysdata, and displays the filed type ("CRJ9"), ignoring what it sees on the network ("CRJ900"). The other CRJ that night hadn't filed a flight plan, so OpenRadar had nothing else to display than what the FG MP network said ("CRJ700"). So it's not that my client didn't send the full model name, and the mpserver didn't eat half the type name either; this is simply a feature of OpenRadar, and it worked as designed. ORCAM, which uses the exact same aircraft model resolution logic as the full-blown regular FlightGear codebase, doesn't care though: it always looks at the MP network, and completely ignores alwaysdata and OpenRadar (in fact it doesn't even know alwaysdata exists). Even if OpenRadar says "CRJ9", ORCAM still gets "CRJ900", and uses that to find the correct model. I could file "Foobar 12345" as the aircraft type, and ORCAM would still use whatever I was actually flying, while OpenRadar would happily display "Foobar 12345", at least once you have imported the flight plan. I think this is what made you believe that my client was cutting off the aircraft name, and what got me confused during the session.
Hope this clarifies things, if not, feel free to ask.