Open MPI and Debian

Branden sent me a link from Planet Debian where one of the Debian developers talks about improving the packaging for Open MPI in Debian. Nice to see people outside the core development community get excited about Open MPI.

Now if only we could get people to spell it Open MPI instead of OpenMPI 😉

Open MPI in Mac OS X!

http://www.apple.com/macosx/leopard/technology/multicore.html
Apple finally announced that Mac OS X 10.5 (Leopard) will include Open MPI as part of its developer tools. The discussion to make this happen started over 18 months ago, and it’s great to finally see it announced. I’ve been working on this for a long time and it looks like we’re finally there. Woo hoo!

Sun’s working on adding some DTrace hooks into Open MPI for tracking things like message arrival and the like. Won’t make it into 1.2, but should be nice once they’re in, especially considering the announced support for DTrace on Leopard.

Happiness on the Open MPI front!

Instrument Flight 3

This morning was the third instrument rating lesson. Originally, we were going to take Kyle out to dinner last night, so instead of scheduling for 7:00am like normal, I scheduled it for 11:00am. The plane was late getting back from a lesson, then needed some oil, so it was 11:40 before we could get the fan spinning. We did a 0/0 takeoff (meaning I was wearing the hood during the takeoff roll, able only to see the instrument panel. I had done two (maybe three) of these in the 152 back in B’ton and had no problem. I was really nervous this time, I think because I’m still getting used to the G1000 panel. I didn’t get well aligned with the runway and didn’t set the heading bug quite right, so it wasn’t my best work. I don’t particularly like sitting on the runway at a non-towered airport, but next time I’ll take my time getting lined up and should do considerably better.

Since it was mid-day by the time we started rolling, it was already pretty hot out there, and the heat of the day was causing some good thermals. It made it a little difficult to work on keeping level flight, especially in turns. We worked on constant rate descents and climbs, which went mildly ok — would have been considerably better if not for the thermals tossing us around. The bulk of the work was doing partial panel. First, we simulated a failure of the PFD (the screen in front of me), so I had to use the MFD (the screen on the right). No biggie — a bit of situational awareness lost because there’s no way to have the “6 pack” displayed with a moving map on the MFD (there is on the PFD, plus when the PFD has the 6 pack, the MFD is basically one huge moving map). Then we simulated losing the MFD as well, leaving me with compass, airspeed, backup attitude, and altimeter. None of which are in a reasonable locations for use in flight. We then did some turns based on the magnetic compass, both timed and compass reading. In the C-172, a 15 degree bank results in a 360 degree turn taking about 2 minutes. So a 90 degree heading turn with a 15 degree bank takes about 30 seconds. Only problem with that is that 1) the airplane timer is in the “failed” G1000 and 2) my watch has a polarized lens and my glasses are polarized and the position of my watch when holding the yoke makes it mostly unreadable.

The magnetic compass is a slightly entertaining because it isn’t smooth in its travel. When turning towards the South, it will precede the real heading. So if you want to roll out on a heading of 180, you have to wait until the compass reads 210 (for the continental US, anyway). The opposite is true for North — you have to roll out 30 degrees early. In both cases, once you roll wings level, the compass will swing to the proper heading and the world rights itself. Rolling out east or west is done without any extra math. All good entertaining fun that makes me happy that I almost always have a directional gyro that doesn’t have these problems.

Instrument Flight 2

[yes, the entry is backdated to the end of the lesson. cope.]

This morning was my second instrument lesson. Friday’s weather was beautiful in the morning and choppy and hot in the afternoon. I was hoping to get the same thing this morning, and was lucky to be right. We took off on 22 and headed to the northwest practice area (just north of KAEG). Takeoff was a bit harrowing — I had the engine leaned a little too much and wasn’t making full power. I had a 100 foot/minute climb which is way low, but at least survivable. Scanning instruments, the EGT (exhaust gas temperature) was high, so richened the mixture and we were able to establish a good climb rate (600-800 foot/minute) at Vy. Much happier.

Instrument flight was uneventful – I was able to keep the plane mostly where it should be. I still drift a bit more than I’d like when I’m trying to deal with cockpit resources (like tuning and verifying the right VOR or ILS). We then came back to KAEG to do the ILS 22 approach. Until about .5 miles before the middle marker (about .5 mile from the end of the runway), I had it nailed. I started losing the glidescope (the vertical guidance) a bit, but was probably +/- 100′ or less and I could have gotten it back together. however, someone decided to start using runway 4 and everyone else switched just as quickly. The runway numbers are magnetic heading rounded to the nearest 10 degrees, with the last digit dropped. So I’m coming in on 22, people are departing on 4. Those good at math have probably figured out that planes were coming right at us. So we broke off the approach early and offset for traffic.

“breaking out of the clouds” (removing the view limiting device) at 300-400′ was an interesting experience. The ground looks *really* close and coming up faster. I had practiced a VOR approach at Bloomington and an ASR (airport surveillance radar) approach at Terre Haute, but both of those go missed at 1000′, so you’re still a long way from the ground when you “break out”. I’m sure I’ll get used to it, but kind of weird.

We then spent some time doing touch and go landings so that I could get the feel of the 172 again. I finally got to the point where I was comfortable in the flare, but still was coming in high and fast. Each one was slightly better than the last, so I think there’s progress being made. It’s amazing how much different the sight picture out of the 172 is compared to the RV. And just as remarkable how much slower it comes out of the sky compared to the RV.

Next week is navigation, which should be fun.

Exhausting Flight Home

Today was a tiring flight home. Winds were directly across the runway at 14-18 knots, occasionally swinging to a headwind. It took a little while for the winds to calm down enough that we could roll (18 is pretty high), but we caught a break after sitting at the hold-short line for 5 or 10 minutes. The plane was squirly once speed came up, so I did a soft field technique and got off the ground early, then nosed over (in some good chop) to pick up a bit of speed. There was a choppy 20-25 knot tailwind on the way home, so we had to stay under Va (about 120 knots in the RV) on the way home.

Winds in ABQ were from the Northwest / North, so runways 26 and 30 were in use. I’ve landed on 3, 8, and 26 but never 30 (17/35 is almost never used). To land from the North on 30, you fly over the numbers of 17 at 7000′ (about 1800′ above the ground), make a left turn and enter right traffic for the runway, staying above 7000′ until south of runway 8/26. So now you’re heading southeast away from the airport, then make the 180 turn to head back towards 30. The 180 part isn’t unusual, but it’s odd that there’s nothing to the southeast of the airport. It’s just desolate wasteland, so it’s hard to judge. So I ended up high and fast on approach. Got it figured out and pointed at the touchdown zone with slightly high speed with a massive crab from the crosswind. Tower called winds at 360@10, so there would be about a 8-9 knot cross-wind. Coming down the crosswind bled off a bit as expected, and I was all setup for a nice crosswind single wheel landing. A little fast, but when you’re slipping to keep aligned with the runway, you bleed of speed really fast in the flare. So get down, start flaring, and the cross-wind drops out on me. No problem, throw out the rudder, put it down on both mains, and roll out (but fast). Not a great landing, definitely a tiring approach, but the plane is not any worse off than before so it’s all good.

Now that I’ve seen the approach to 30, I think the next one will be much better. I think I have a better idea on where the waypoints are and how high I should be before starting the turn to base. But I suppose I won’t know until the next flight requiring me to land on 30. Yippie.

How to get the FAA mad at you…

Mom – nothing in this entry involves me or the plane. Sorry, but they still let me fly…

ATC (Air Traffic Control) is responsible for controlling traffic around busier airports. This includes everything from a tower at an airport like Santa Fe or Bloomington making sure only one plane at a time uses the runway to controllers using radar to positively control both IFR and VFR traffic at busy airports like Denver or LAX. And you talk to ATC while you’re flying around, either for routing (IFR) or traffic advisories (VFR), although VFR doesn’t really have to talk to ATC in route. But it’s a good idea.

Anyway, for what not to do, there’s a guy in the news for being escorted down by F-16s after controllers thought he had been hijacked. Ooops. I’ve been in close proximity of F-16s and F/A-18s (they fly in and out of the Sunport), but never when one was escorting me. That would be double-plus not fun. There isn’t much information yet so it’s not clear whether he accidently scrolled through the magic code on his transponder or if he used one of the code words. Should be interesting.

There was an e-mail from the FAA Safety Team in my inbox when I got home from work today, a report on runway incursions for the last twelve months in the local region. The summary is:

12 - Entered runway or crossed runway hold bars without ATC clearance* 
11 - Issued and read back hold short instructions, then entered runway 
4 - Landed without ATC clearance 
10 - Departed without ATC clearance* 
2 - Departed from wrong runway 
* A single aircraft is responsible for multiple (2) RIs

So the 12 entered runway or cross runway hold bars is a difficult one. Clearly one guy took off that way, which is double-plus bad. But on complicated airports like KABQ, it’s not too hard to get lost and end up crossing an active runway. Not good, but you can almost understand how it happened. But 11 times, someone got instructions to NOT enter a runway, read back the instructions to NOT enter a runway, then entered a runway. I’m not sure how you do that exactly, but they did. *sigh*.

Flying “blind” (Instrument Flight 1)

I need to find cheaper hobbies. The next logical step after getting a private pilot license is either to get a bunch of endorsements (taildragger, high performance, complex, etc.) or an instrument rating. High performance around here means a T-41 or a 182RG. The T41 is a 172 with a huge engine and not that interesting. It goes slower and uses more gas than the RV-7. The 182RG is high performance and complex, but really expensive and anyone who rents one would want lots of hours before they’d let me go alone. The only taildragger for rent in all of central New Mexico is a Pitts at Bode, but it’s ungodly expensive and not a fun plane to fly straight and level. It is, however, an awesome plane to fly acro. It’s also $300+ an hour, so not in my price range. That left instrument training. Today was my first lesson. I’m training in a Cessna 172 with a Garmin G1000 glass panel.

We went up for about 1.2 hours, about 1.0 of which was under the hood (wearing one of these things, so all you can see is the panel, not the outside world). We worked on level flight, turns, climbing, descending, and turning to a heading. While I can do all this from muscle memory when I can see out the window, even in something as “quick” as an RV. But when the outside world goes bye-bye, all the normal clues to what is going on are gone. The scan for the G1000 is definitely easier than the scan in with steam gauges (we have to do 3 hours of instrument training for the private pilot), as everything is much closer. On the other hand, the G1000 requires a lot more focus than steam gauges for setup. It’s not just dial in a VOR and hope for the best, it’s add a flight plan, blah, blah, blah. Thankfully, most of the interaction is very similar to the GNS430 in 813T, so it wasn’t too bad. But I definitely need a couple more hours before I feel comfortable.

Next weekend, I have my second lesson. More getting the flow down and possibly a couple of ILS approaches at KAEG. So sometime this week I need to read up on ILS approaches. Hopefully, I can get Galen to safety pilot for me for a couple of flights this week to further practice. Even 0.1 or 0.2 hours of instrument flying will really help keeping me in shape. I hope :).

IKEA’s clearly not American enough…

So I have a couple of pieces of furniture from IKEA. Last summer, all my furniture was stored in a storage unit in B’ton while I was in Los Alamos. During that time, the unit was broken into and my tool box (along with some other things) was stolen. In the toolbox was all the screws / bolts / etc. necessary to reassemble the IKEA furniture. I finally got around to calling about it (the furniture didn’t get out here until January and I’ve been busy).

Anyway, they’re willing to send me replacement hardware for free. No charge at all, not even shipping. Amazing. Mad props to them, although passing up the ability to charge someone ten different fees seems un-American too me :).

Quickies

  • Bought some fun stuff recently:
    • An ICOM IC-A24 handheld nav/com aviation radio. Nice backup to have on long cross-countries, and even on local flights, since Albuquerque International (Class C), Santa Fe (Class D) and Los Alamos (special restrictions) all require a radio to enter their airspace. And with Double Eagle II about to go Class D, a radio out would mean either waving at the tower in one of the Deltas to get lights or going to Belen. Neither are fun when you just want to land. Most of the radio was paid for from gift certificates Mom and Dad gave me for Christmas and my B-Day. Thanks Mom and Dad!
    • A fryer. Somehow it’s Pete’s fault, but it seemed like it would be fun to have and was only $20.00 at Target.
    • One Six Right, in principle a movie about the history of Van Nuys Airport (KVNY) but is just as much about the history of General Aviation in the United States. There’s a lot of beautiful birds in the movie, and it is extraordinarily well done.
  • My car went in for its 60,000 mile service today, as well as a new set of tires. She’s apparently aging well and there’s nothing big that needs to be done in the near future. Woo! We’ll ignore the 60,000 mile service waited until 65,000 miles on the car. Wow, almost 5 years and over 65,000 miles. Doesn’t seem like it’s been that long!
  • Weather has been sucking lately. We didn’t get to fly up once this week and it looks like next week is going to be just as bad. Hopefully we’ll catch up on some squawks while we can’t fly.
  • FInally getting around to some much needed web server maintenance on marvin, the server that hosts bbarrett.org. Trying to setup some of the services like bug tracking that worked at one time or another…