Over the last several weeks, I fine tuned the software, ironed out several bugs, and generally made the interface more user-friendly. Not surprisingly, the LAF found nothing except a few false returns from some communication satellites, muon showers, and stray cosmic rays. The results were logged to a text file on an SSD, and I inspect the log each night after work.  If it is empty for three or four days straight, I sometimes tweak LAF’s search strategy.  This may be a lot harder than I thought.

Exactly two days ago, I came home, ordered a pizza, and turned on the TV.  Zoe (my girlfriend) came over a bit later and we watched some old Star Trek episodes for a bit.  Eventually I got up to check my email in my office and noticed my computer was dead. Well, it wasn’t exactly dead. It just wouldn’t boot. I try everything. At first it looked like it was just a hosed OS. Luckily, I had a Linux box handy, and I mounted the drive with LAF as a secondary.  After about an hour of checking and fiddling with recovery software, I finally got some information off the disk.  I could see the filenames and their structure, but their contents were corrupt. At first, nothing seemed unusual.  Zoe suggested that I check the LAF log directory. All the correct filenames were present. The odd part was the log file sizes. All together, the logs were nearly 1.5 terabytes in size!

I think I can now fit the pieces together. Looks like LAF locked onto a signal and wouldn’t let it go.  Apparently, the LAF thought this was the real deal.  The data was captured to the log file which eventually filled up the disk.  The OS then collapsed without any temporary storage to write to and eventually corrupted itself.  What a mess.

At first I was devastated. But Zoe reassured me.  The LAF had found something! But, what had it found?  Was it significant? An alien intelligence, or just radiating sun spots?  If it was significant, maybe it could change the world. I thought about sending the corrupted hard drive out to get it properly salvaged, but decided against it when Zoe looked up the current recovery rates. They want two grand! I didn’t even spend one grand on the home-built computer, and I certainly wasn’t going to spend twice as much on fixing it.

But the data is definitely toast. I can see the log file times and sizes, but the records aren’t formatted the way they should – consist of all zeros.  They must have been corrupted when the computer ran out of space.  There is no way to figure out what LAF found with what I have.  Still, there is some good news. I have a recent printout of the code for LAF software, and I even have another computer.  The only thing I needed to know now is where in the sky LAF was looking, and the exact time of the failure.  After poking around the recovered directory listings, I can see that there are valid time stamps on the files. The main log has a modification time listed as 3:13pm and 44 seconds. Now all that was left is to figure out where it was pointing.

The answer to that question may be easier to figure out than I first thought. We just went outside and circled the dish. To me it looks like it is pointing straight up. Maybe the dish was trying to return to its “stowed” position after the crash.  But Zoe says it isn’t exactly pointing straight up.  This is really puzzling…

Zoe is right as usual.  The dish is definitely off-axis.  When I migrate the LAF code to my backup computer I’ll get the dish’s precise orientation first.  And before I start up the LAF again, I’m going to make some changes to the code and hardware so this scenario doesn’t play out again.  Tonight, I’m going to buy several large external RAID drives.  Over the weekend, I’m going to type in the LAF code base from the printout I have, and make a few changes to it as well.

First, I’m going to add new logic so the logs can grow to span the new drives I’m installing and formatting tonight.  The logs will not be written to my OS disk as they were this last time.  Fool me twice, shame on me!  I’m also going to put in additional code to limit the log sizes so there is always a buffer of %5 empty space left on each drive just in case.  Finally, at Zoe’s brilliant suggestion, I’m going to add triggers so that I will receive instant email and tweets when the logs experience significant change.  I’m not having this happen again!

Ok.  I bought six SSD drives, installed them in my backup computer, and configured the new disks as three RAIDS.  This replication of the data will ensure their integrity even if a meteor strikes.  I spent Saturday typing in the LAF code once again from the recent printout.  It didn’t take me that long, and I made some improvements along the way.  On Sunday I added the new logging logic discussed earlier as well as Zoe’s triggers.  I bought a real nice UPS (no birthday presents this year!) to keep the LAF running even throughout a long power outage.  My girlfriend points out that this is silly since the UPS can’t actually power the dish too, but hey – I’m not taking any chances!

Oh, and before I did any of this, I used the backup computer to read the current position of God’s Ear.  Using this position and the time I got from the log file’s latest modification time, I now have an approximate location of where it was pointing last.  Indeed, it wasn’t pointing straight up.  Looks like it was oriented with a right-ascension and declination that put it in the neighborhood of sun HDS-288, which, by the way, is only about 60 light years away!

But, the recovered location is just an approximate value, and I’ve decided that the LAF’s first task will be to scan the narrow band of sky surrounding this star.  Curiously, most of the area surrounding HDS-288 is peculiarly dark.  Hmmmm.  Anyway, I’ll kick this first task off tonight, and maybe something will turn up during the week.  Crossing my fingers!

On Monday night I kicked off the LAF for a test run, but I had hit the panic button about 10 minutes later when I realized it wasn’t working.  I had set the log space limits to a ridiculously small value, but the LAF just kept recording anyway.  After a bit of debugging, I realized that I had set the limits in Kilobytes.   But, the LAF sees these limits as Megabytes. Doh!  Oh yeah, did I mention I’m 41,000 years old today?

Anyway, I’ll kick off the LAF tonight.  It’s late, and I have a big day at work tomorrow.  Also, Zoe wants to take me to a late showing of “The Fifth Element” tomorrow night.  I know, it’s an old flick, but I can’t resist a good Sci-Fi, even if it’s a bit dated.  Besides, we like long movies… 😉

This is crazy.  I kicked off the LAF Tuesday night before I went to bed, and by morning, it had scanned about an eighth of its tasked area by the time I headed out to work.  When I came home, it had surveyed nearly half of star’s surrounding area.  I received not a single email or tweet throughout the day.  I even checked on the computer again before we left for the movie and all was well.  The log was empty except for the occasional update on the antenna’s position.  As I said before, there doesn’t appear to be much of anything in the narrow band surrounding star HDS-288.  At least not in the visible spectrum.  I was mulling over the idea of relaxing the frequency scan range but I eventually decided to do this only on completion of the this pass.  After all, the job was now half done.  If this first run failed, I could always tweak the parameters and try again later.

So, we headed off to see “The Fifth Element”.  I really love this movie.  It is an adventure, comedy, love story, and sci-fi all rolled into one!  Anyway, on the way home I decided to turn on my iPhone.  I had turned it off when the movie began around 8:00pm.  It was now more than two and a half hours later.  The phone immediately began to shake like a kid in a candy shop.  I was driving on the highway at this point, so I threw the phone to Zoe.  She gave me a dirty look, but her eyes opened wide as she read the display.  She read it out loud to me: “You have 117 emails from the LAF… Oh, and one from a very important online vendor.”  I pulled over to the shoulder of the road, and slammed on my breaks.

The triggers were firing all right.  Zoe read the first message: “Are you having problems satisfying your loved one at night?  Your troubles are now over!  We deliver affordable Viagra discretely to your door…”  I smiled at Zoe, and shook my head.  I grabbed the phone back and read the first LAF message.  It had the subject: “Signal detected, locking in frequency”.  In it contained some low level diagnostic info that I could examine later.  The second message read: “Signal analyzed.  Frequency Modulation detected.”  I quickly jumped to the third message which read: “LAF now recording to disk 0x0”.

What followed were nearly 100 emails containing updates to the ongoing data recording.  The last email put the data recording at around 93%.  I panicked.  “Zoe, we gotta get home – now!”, I said.  I was about to slam on the gas when I glanced up and noticed the red and blue flashing lights in the rear window.  Crap, just what I needed.  It seemed like forever, but a policewoman eventually got out and made her way over to our car.  She asked me why I was parked on the side as I was.  She then looked at Zoe and said: “Are you ok?”  Zoe smiled and said she was.  Then I told her we needed to get home to my computer which had been sending us urgent email.  I didn’t realize how that sounded until the policewoman asked me to step slowly out of the car.

For the next ten minutes I was asked to walk lines and perform endless tricks.  Eventually she asked me what I do for a living.  When it became clear that I was a “geek” by trade, my behavior no longer seemed as strange.  Eventually she let me go, but the whole ordeal took way too long.  When she left, I tried not to rush home because I didn’t want to stop again, and this time have to explain why I was speeding.  When we arrived home the house was dark.  Zoe turned on the lights while I ran to my office to check on the LAF.  The screen had timed-out and was black, so I quickly jiggled the mouse and brought up the LAF console.

The last entry in the LAF log read: “Disk 0x2 exhausted, LAF shutting down collection”.  Crap.  Crap.  This is exactly what I didn’t want.  I looked at the current status indicators which displayed the current position of “God’s Ear”, but “No Signal” was currently being detected.  I was exhausted at this point, but I was even more curious, so I  scanned the beginning of the recording on the first disk.  For frequency modulation I simply record the demodulated wave.  On the screen I saw a single pattern continuously repeating. There were equal sized waves followed by silence for an equal amount of time.  After closer inspection, I counted and found each set contained 42 waves.  Not very exciting.

This was most likely a terrestrial signal, and not one from space.  Probably just a test pattern from some military vehicle.  Ah well.  What a day.  Tomorrow night, after work, I’ll cross reference the last sky position, time, and frequency with likely terrestrial emitters to see what the LAF found.  At some point I need to add logic to the program so signals like this are automatically rejected.  Right now, though, we’re going to sleep!  What a day…

After work, I had Zoe cross reference the LAF’s locked-on position with know terrestrial sources.  She said it was a waste of time – though I couldn’t see why.  Meanwhile, I continued to scan through the logs.  So far, the pattern simply continues – 42 waves, 42 silences.  This effort was going too slow, so tonight I just wrote a small application to systematically search the remaining logs and detect if this pattern ever varies.  It is probably not a good use of my time, but hey.

The real interesting news is that Zoe confirmed that the signal was unlikely from a published terrestrial emitter.  I asked here how she knew this.  “Easy !”, she explained, “The log continuously shows each position adjustment “God’s Ear” made after lock-on last night.  Based on these adjustments, it’s easy to tell the movement of what it was tracking with respect to our position.  If it were an airborne platform, it could be either moving or stationary.  If it was a stationary platform, the dish wouldn’t have moved – but it did.  If it was tracking a non-stationary airborne platform, the LAF position would change, but not in the same way as it would if it were tracking a distant object in space.  In fact, the way the LAF was tracking, it was basically compensating for the Earth’s rotation.  This means that whatever it was tracking was either very far away, or local… but moving in a path with ridiculously significant coincidence with the Earth’s rotation.  I’m betting on the former.”

Well, I’m done with the simple search application, and I’m kicking it off as I type this.  We’ll see if it turns up anything tomorrow.  Either way, I can’t think of anything terrestrial or celestial that would emit such a meaningless signal, so I’m curious about what the LAF found.  At this point, though, I can’t help but think that I’m getting sidetracked.  Maybe I should just ignore this signal and keep searching for something more interesting?  Or, maybe I should just go to sleep and worry about this tomorrow!

What can I say.  I woke up this morning to the sound of a steady beep.  At first I thought it was a dream.  Then I wondered if someone broke in.  finally I remembered that my log scanner app was set to beep once if it finished, continuously if it found something.  I don’t know how long it was beeping, but I stumbled out of bed and dashed into my office.  The app had found something something in the log.  The boring cyclic pattern ceased just prior to the termination of the log on the second drive.  In it’s place were what appeared to be short and tall pulses at the same frequency as the 42 pulses seen earlier in the log.  The first grouping simply had 42 tall pulses.  From that point on, if there is a pattern I can’t see one.  Looks pretty random to me.

But, if the pulses were something though, I’d guess the short and long pulses were standard binary bits.  Each complete number would then be composed of 42 pulses, or bits.  I don’t have much time before work, so I just wrote a quick script to collect groups of 42 bits and record their decimal number equivalent.  Maybe I’ll see a pattern in that.  I don’t know where this is going.  It looks like another long weekend ahead of me.  The fact that it’s Friday the 13th isn’t helping…