About the F-22 software bug
in [ADA]
|
Prev: Ada in Boeing 787
Next: Private or public task ?
From: Martin on 6 Feb 2010 04:28 On Feb 5, 4:50 pm, Hibou57 (Yannick Duchêne) <yannick_duch...(a)yahoo.fr> wrote: > On 5 fév, 11:31, "Dmitry A. Kazakov" <mail...(a)dmitry-kazakov.de> > wrote: > > > And shift does not solve the problem anyway, if that existed. The potential > > problem is that the angle is not a real number. It could be represented by > > one, but then the operations like +, -, *, /, =, /= must be replaced and > > ones like <, >, <=, >= disallowed. > > I would have thought for reals, operations like +, -, *, /, <, >, /= > changed and =, <=, >= disallowed > Is there something I don't understand ? Longitude is cyclical - < and > only 'work' if there is a hard-stop at each end. You can make up some rules that might work for a project, e.g. 2 points are only ever compared using the shortest route between them (and not the route that goes [almost] round the earth!) but it isn't really an honest abstraction. Cheers -- Martin
From: sjw on 6 Feb 2010 07:12 On Feb 5, 11:39 pm, Phil Clayton <phil.clay...(a)lineone.net> wrote: > Interesting to see that 'cyclic fixed point types' are under > discussion: > > http://www.ada-auth.org/cgi-bin/cvsweb.cgi/ai05s/ai05-0175-1.txt?rev=1.2 > > http://www.ada-auth.org/ai-files/minutes/min-0911.html#AI175 Interesting that those discussions are in terms of cyclic *fixed point* types. I agree that at the boundary you need to handle cyclic fixed point. Back in the day, the Ferranti name for these was 'standard angle format'; the F1600 series was a 24-bit machine, and the most significant bit typically represented -180 degrees. Overflow was represented by an optional check, so wrap-round was as easy as pie. But why (on a machine with any sort of floating-point capability) would you want to preserve this anywhere other than at the boundary? On my present system, angles are stored in radians. Handling wrap- round (and sometimes we need to deal with a range -2pi .. 2pi) was not properly dignified with library support from the start, but we've achieved a generic which lets you normalise an angle to -pi <= a < pi or to 0 <= b < 2pi or to -2pi <= c < 2pi with minimal pain. We don't need to worry about latitude!
From: jonathan on 6 Feb 2010 10:30 On Feb 4, 6:46 pm, Pascal Obry <pas...(a)obry.net> wrote: > Just curious, does someone have information to share about this problem: > > << > While attempting its first overseas deployment to the Kadena Air Base in > Okinawa, Japan, on 11 February 2007, a group of six F-22 Raptors flying > from Hickam AFB, Hawaii, experienced multiple computer crashes > coincident with their crossing of the 180th meridian of longitude (the > International Date Line). I did another quick web search ... this was the best I could find: http://www.murdoconline.net/archives/4595.html/comment-page-1#comment-324995 quote: ... it should be noted that they did test the code in a simulator, and it worked fine. But the simulator did not use the actual GINS [GPS Inertial Navigation System] hardware (because the simulator does not actually move, so a GPS receiver would be useless to simulate movement). The GINS hardware was tested separately and also did not show an anomoly... By the way, if you want an awesome display of some real F-22 software in action, there are several youtube videos around. I googled: F-22 youtube Supermanouverability. All I can think of watching these is what the programmers have go through to make these things happen. To me its seriously scary ... complexity quite possibly on par with that of in-flight entertainment systems;-) Jonathan
From: Pascal Obry on 6 Feb 2010 11:35 Le 06/02/2010 16:30, jonathan a �crit : > On Feb 4, 6:46 pm, Pascal Obry <pas...(a)obry.net> wrote: >> Just curious, does someone have information to share about this problem: >> >> << >> While attempting its first overseas deployment to the Kadena Air Base in >> Okinawa, Japan, on 11 February 2007, a group of six F-22 Raptors flying >> from Hickam AFB, Hawaii, experienced multiple computer crashes >> coincident with their crossing of the 180th meridian of longitude (the >> International Date Line). > > I did another quick web search ... this was the best > I could find: > > http://www.murdoconline.net/archives/4595.html/comment-page-1#comment-324995 Thanks to all for your feedbacks. Pascal. -- --|------------------------------------------------------ --| Pascal Obry Team-Ada Member --| 45, rue Gabriel Peri - 78114 Magny Les Hameaux FRANCE --|------------------------------------------------------ --| http://www.obry.net - http://v2p.fr.eu.org --| "The best way to travel is by means of imagination" --| --| gpg --keyserver keys.gnupg.net --recv-key F949BD3B
From: Martin on 7 Feb 2010 05:11
On Feb 6, 12:12 pm, sjw <simon.j.wri...(a)mac.com> wrote: > I agree that at the boundary you need to handle cyclic fixed point. > Back in the day, the Ferranti name for these was 'standard angle > format'; the F1600 series was a 24-bit machine, and the most > significant bit typically represented -180 degrees. Overflow was > represented by an optional check, so wrap-round was as easy as pie. SAF16 and SAF32 still in use today!! ;-) Cheers -- Martin |