| So - the older daughter is going to MD renfest this weekend with friends. It's kids' weekend (sprouts under 12 enter free). My car is going into the shop.
In the logic that is our world, this implies that the Rothman family (all of usn's) are piling into the van, dropping my car off at the shop, and proceeding to MD for the renfest, including the 9 and 11 yr-old, and delivering the 13-yr-old into the tender clutches of her peers.
we'd be glad of any company in our renfestery - given the understanding that this weekend is rug-rat infestery renfestery. | comments: Leave a comment   |
| Lori and the kids and I, my sister and neice, all went up to Maine camping for a few days. Very very cool. We went up during the Pleides, and saw volumes of shooting stars.
Also, while we were up star gazing, I had a sky map, and reference for constellations. I finally figured out how to find several constellations in the summer sky that I hadn't previously - including particularly Cygnus.
Good fun was had by all. No injuries were incurred, mild exercise and fresh air was had. None of the children incurred injurious degrees of wrath - despite the antagonizing singing of the Phantom of the Opera. | | comments: 4 comments or Leave a comment |
| Alright all - my daughter is now 12 (and a HALF). She's a mature young lady, and it's about time to start hitting her with the heavy books and stories that can starkly influence your life in your early teen years. As an unrepentant and non-reconstructed geek, I have a small list of materials that really hit me at around that age (though probably a couple years later). I'd also like to prod the audience for books/ideas/memes that had a significant impact during their formative years.
My list:
1) Stranger in a Strange Land (Heinlein) [though now I might replace/supplement that with 'To Sail Beyond the Sunset']
2) Atlas Shrugged (Ayn Rand) [also potentially 'Fountainhead' - too bad I didn't get to that one earlier - I have more identification with the architect's imperative than the rail magnate's...]
3) The Naked Ape (Desmond Morris)
More lately I think I'd add
4) Guns Germs and Steel (Jared Diamond)
I'd probably also add a movie to this - though almost more as a 'topping off'
5) Tucker: The Man and His Dream (1988)
and from the music genre
6) The Wall (Pink Floyd)
7) The Who Collection [OK, I've lost track of particular albums... Baba O'Reilly, Who Are You, Won't Get Fooled Again]
8) Rush [again, I'd have to review albums, but definitely The Trees, Subdivisions, Tom Sawyer..]
9) Thick as a Brick, Songs from the Wood (Jethro Tull)
All thoughts and suggestions welcome! | | comments: 4 comments or Leave a comment |
| So.. I'm coding like hell on my Advanced AI term project. Going like the dickens.
Sunday night, my code blows up. Badly. Running out of memory.
OK, says I, must be a memory leak. Hook up the memory profiler (open source freeby - how cool is that?) and find the offending widgets. I found the offending widgets, then started through my code with a tooth comb, looking for circular references and poor exits from routines - I figure I've got some sort of leak that's giving the garbage collector fits. And I do this for HOURS. I've got really cool disassemblers on about 1/3 of my object classes now. Doesn't help. I wire up timing-based invocations of the garbage collector - maybe it missed my objects because it wasn't looking at the right time? no joy.
Then I step-step-step through my code, looking for the place where things are getting built and disposed (on-the-fly call tree analysis - more freeby open source - how cool is that?). I really start looking at the things I'm building and destroying. Turns out the 'dynamic and adaptive' nature of my code was killing me.. some long-term recursive-like aspects of my code were causing allocation of larger and larger and larger objects. So I went back into the algo, ensured my 'grow' and 'shrink' were balanced statistically, then installed some hard gates (if it's too big, or too small, dispose of it and try again..). Problem fixed.
So why is it ![[info]](http://l-stat.livejournal.com/img/userinfo.gif) zeekar's fault? Because a while ago (few weeks? few months?) on LJ, Mark started a great thread about some of the issues with garbage collection in Javascript. OK, I'm not Javascript'ing, I'm straight-up Java'ing, but the discussion stuck in my brain. So my immediate intuition was to jump straight to try to resolve issues in my code that might cause Javascript to blow a gasket with its garbage collection. Wrong leap of intuition. But I blame Mark for the hours of debugging in the wrong direction. :) | | comments: 2 comments or Leave a comment |
| Our household is just recovering from a quick summer cold that came through. I think I actually brought it home from work . I believe we've cleared it though, and despite the little buggies' best efforts, we managed to keep them from lodging in either sinuses or lungs (sinusitis and bronchitis being recurring issues in the Rothman household).
I also just took a quick trip to San Diego for work - out Monday PM, back red-eye Wed PM/Thur AM. Good prodeutive work, but not much by way of sight-seeing. I ran into some of my old crew from the Collaboration Management Office - apparently this week was the annual chat conference Lorraine holds out there. Who'd have know'd it?
So a quick story from San Diego... I got into San Diego Monday evening, and pulled into my hotel there about 9:00. I was a bit tired and wrung-out, so I decided to grab some food at the hotel, and then turn in. The dining room was closing, but the kitchen was still open - I convinced the waiter to seat me out on the deck. While I was eating, I stopped the waiter to ask what the fireworks were for, and he casually said, "oh, that's just Seaworld. How is everything?"
"How is everything?" I responded. "I'm on a deck by a swimming pool - 70 degrees, low humidity. Overlooking a marina. The hotel next door has an outdoor jam band with a really good sax player. The food is quite good. Now Seaworld is making fireworks just for me. You know - I think I'm doing all right." | | comments: Leave a comment |
| After Doctor Who, PBS runs a couple of music/concert shows, with a fascinating variety of different styles. Tonight the cards turned up Bonepony that I'm really enjoying. Aside from the fact that I really like their particular blend of traditional folk, bluegrass, rock, and a bunch of other eclectic influences, and the fact that they're amazingly talented (guitar player Nick Nguyen - wow...), there's a component that really goes past that. These guys play with joy. Pure, unadulterated, lovin' every minute of what they're up to.
This put me back in mind of Bruce Sterling's novel 'Holy Fire', and what he expressed in there as the glaringly dangerous, consuming pursuit of the thing that fires you up and rewards you with deep joy. It's dangerous because it can draw you in, make you neglect everything else. It may be maddening, and frustrating. It may be teasing and demand blood and sweat. But if you get it right, if it's the thing for you, swallowing the flaming thread, filling yourself with holy fire, can cause you to do things beyond what you imagined you could do.
Right now I don't have it. What I'm doing at work doesn't even touch on it. Work is going to give me ulcers in pretty short order. In my last gig I was trying to force it - I knew the collaboration area had a core of inspiration, but it wasn't mine to tap into. I'm still pokin' around the doctoral dissertation thing, and I'm still lacking inspiration. Unfortunately, passion and professionalism often masquerade as true inspiration, but they're not the same. There are topics in security and command and control that I'm passionate about, particularly some that are blatantly broken, particularly ones that chap me regularly. There are also areas that I recognize that are important and in need of fixing/improving/work, which I recognize I have unique insight into (DoD does not tend to generate or reward abstract thinkers - there are about 4, maybe 5 of us I know of in DISA...).
I'm going to make myself crazy if I can't find a real source of holy fire before long. Marking time at work, and working my way the slow, standard, and risk-free way through my academic program is starting to get taxing, and pushing me quickly to that 'haven't found my font of greatness' mid-life crisis pretty inexorably.
But for the moment, participating vicariously of Bonepony's holy fire is very enjoyable. | | comments: Leave a comment |
| Following the popular meme, here is a list of places I spent the night in 2007:
Sterling, VA (home!)
Atlanta, GA [1-night for work - FORCECOM]
Orlando, FL [grandad's 90th b-day!]
Oak Ridge, TN [work, Oak Ridge Labs]
Suffolk, VA [work, JFCOM]
Hurley, NY [Lori's folks - Xmas]
Poughkeepsee, NY [my folks - several visits]
St. Louis, MO (well... IL side of the river) [1-night for work, TRANSCOM]
Raleigh, NC [skellington's wedding!] | | comments: 1 comment or Leave a comment |
| Tonight was Operating Systems. Boy am I glad I just had that class last semester and knew what was on tap... nasty test... it doesn't likesess usss..
Tomorow's networking. It'll be mystery meet - I haven't done any networky stuff at GMU yet.
Sample question from tonight's exam:
(2 hr exam. question was 1 of 4, worth 25%...)
3 classes of accessors to a shared resource: searchers, inserters, and deleters.
- each deleter must have exclusive access
- only one inserter at a time may access, but that access may be shared with many searchers
- many searchers may concurrently access the resource, and may share access with inserters (well, the one allowed inserter).
write pseudo code for concurrent searchers, deleters, and inserters. You have no knowledge of how many searchers, deleters, or inserters may present themselves (no maximums). Use only semaphors for concurrency control. | | comments: 2 comments or Leave a comment |
| well, Mon evening was the first 2 of 4 doctoral qualifier exams. I just finished AI and Algorithms. I figure it's about 60% I passed 'em both, less than 10% I failed 'em both. Personally, I like Algos a lot - but I haven't had the local course ("Describe the Church-Turing Thesis, and current technical support for it" ?!?! someone's pet rock...).
I did have AI locally, about a year ago. But yick. AI (or at least the prof I had here) has gotten a lot more realistic, and a lot less hand-wavey. But I'd still argue that it hasn't graduated to an academic discipline yet - if physicists did their work the same way, we'd have 53 kajillion versions of slinkies, but no semiconductors...
Needless to say, I'm a bit more confident in my performance on the algos test than the AI test. But we'll see. Local report is that the pass ratio on both tests is about 50%.
Tomorow's OS. Wednesday's Networks. Once more into the breach! | | comments: 1 comment or Leave a comment |
| I'm at a stage where I have to start thinking about an appropriate doctoral thesis topic. I'm going to start blogging some thoughts, ideas, hopes, and concerns out here. I know a bunch of folks out there that see this have been through the nut-roll, and/or been close to those who have. So... I'm open to all advice, critique, and stray hairs that anyone's interested in contributing.
A couple of initial data points - my background is a mix of math, operations research, and CS. I'm pursuing a doctorate in CS. My main areas in CS are algorithms and analysis, networking, with lesser depth/interest in compilers, OS and AI (though the AI/optimization cross-over is interesting). A thesis in graphics, hardware, systems, natural language, or other areas is probably not happy for me.
I have a couple of initial candidates:
- information theoretic analysis of bidding in bridge as a noisy communications channel. (OK, I'm not going to school in california, so this probably won't fly).
- analysis of competitive advantage in ant populations, with applications to autonomous and distributed systems. I'm actually fairly serious about this one - I think ants have some serious goodness going on, and much of it probably has application in autonomous computing. Part of the problem I have here is that this is almost necessarily a cross-discipline topic, and will entail some serious mod/sim on ant behaviors/populations that could easily be an animal behavior thesis as well...
- characteristics of non-hierarchical distributed data storage. not well defined - but the general idea would be to look at grid and/or peer-to-peer distributed algorithms for data storage and retrieval (i.e. distributed hashing - some of the TOR stuff, etc.). I like distributed, and peer-to-peer is a particular favorite of mine. I've also done some work on web caching, and have been perenially dissatisfied with the methods and means employed. My biggest concern in this area would be a negative result - if the big brains at Akamai think hierarchical and dns-naming-based is really good, they've probably got a point. The bet here would be that their criteria was implementability in 1997...
- accomodation of perterbation in vehicle routing optimization. This is really an operations research topic - how to make vehicle routing (airplanes, trucks, trains) less 'fragile'. You know, so that weather in Houston won't make your flight from PA to MO 4 hours late (as happened to me this week). A way to allow intelligent slack to accomodate snow storms, canceled routes, and congestion events. Some of this applies to data networking as well, although the constraint sets are different. Danger here is that this is well-tilled ground - the airlines have been paying people a lot of $$ for a long time to improve their routing capabilities.
- routing and admissions calculations in MPLS based multi level security networks. This is one that could draw on a lot of background and capabilities from where I work - and it's fairly fresh ground. I think MLS in networking from a DoD employee might get a fairly easy ride as well. But I know the guys in our backbone groups have been wrestling with automated admissions and routing for a couple years now, with some pretty high-powered network gear vendors, and all without much success. This one may be thankless...
- application of optimization algorithms in quantum computing environments - analysis of complexity, and enhanced capabilities. What happens if you run standard linear programming algorithms in quantum environment? what about some advanced implementations like elliptical solution? How does violate-and-recover work? Initial value crash? Does integer programming (branch A LOT and bound) cease to be scary? This has a lot of promise, but not an awful lot of existing literature or practice. I also don't speak Hamiltonians, or any of the other quantum-physics specialized maths - and I'm not sure the hardware guys have abstracted enough to get away without it...
So as I said - fire away. I'll be cooking up some more of these. I'll expand and enhance ones that seem to be promising. But I need to start brainstorming so I've got a few decent proposals to start shopping in the department. | | comments: 2 comments or Leave a comment |
| 1. Go to http://www.careercruising.com/.
2. Put in Username: nycareers, Password: landmark.
3. Take their "Career Matchmaker" questions.
4. Post the top ten results.
Mine:
1.Computer Engineer
2.Electrical Engineer
3.Electronics Engineering Tech
4.Electrical Engineering Tech
5.Physicist
6.Optical / Ophthalmic Lab Technician
7.Aerospace Engineer
8.Mechanical Engineer
9.Boilermaker
10.Mathematician | | comments: Leave a comment |
| I'm feeling inordinately pleased with myself. In my rollover IRA (tax sheltered) I just executed a 24-hr, +80% options trade. VTR is a nice little REIT that specializes in health care properties. They got beat up in the whole subprime thing. They've got earnings coming in on 8/9, and their current crop of options expires 8/20. So... in the unsettlement of the last few days I put in a limit order on a straight call option (strike @30, v. yesterday's trading @32.15-ish) for 2.25. I got that, and figured I'd sit on it a few days - hop out middle of next week. During the day today, however, VTR jumped, and was reaching mid-33s. So I put out a limit sale on the options for 4.40. It executed just at close.
My thought on this was that with subprime cold feet running rampant and the markets broadly tumbling, all the bad news for the sector was getting plastered across all bystanders - including VTR. With a few days to cool down, and a potential upside in earnings report, there should be a substantial but very temporary "bump" ahead of earnings. If earnings actually were good, the bump might "stick", but I know VTR has been doing some restructuring of debt and properties this quarter (taking advantage of the subprime discount...) and I wasn't sure how well the street would take a report with a lot of one-time components. So I planned to sell calls ahead of the report. I also hold some long equity, so good stuff would still be good, just not as leveraged.
So... I'm feeling clever this afternoon. | | comments: Leave a comment |
| Your Score: Clark Gable You scored 28% Tough, 19% Roguish, 9% Friendly, and 42% Charming!
You're a helluva guy, a real split personality and a bit of an enigma. On the one hand, you're a man's man, tough talking and ready for anything. But on the other hand, you soften your rough and tumble core with a disarmingly smooth exterior, and you make the ladies swoon. You're equally admired by both men and women alike, drinking other men under the table all the while charming the socks off half a dozen lovelies. You're a commanding presence, and you know how to get what - and who - you want when you want it. You're drawn to women who, like you, are savvy enough to deal with the world on their own terms. You work well with spitfires. Leading ladies include Joan Crawford, Myrna Loy, and Jean Harlow. No damsels in distress for you.
Find out what kind of classic dame you'd make by taking the
Classic Dames Test.
| | comments: 1 comment or Leave a comment |
| Charlotte is 10 now, and has joined a baseball league. Not a softball league, but a baseball league (unintentional, but... OK). She's the least experienced member of her team - this is fanatic baseball territory, and they start 'em (ALL of 'em) as wee tots. So... Charlotte's got the girl thing to get passed, and also some inexperience.
So yesterday, Charlotte and I went out to some batting cages for a while. I've never been to a batting cage before (it occurred to me). So Charlotte and I took some turns swingin' the bat. I got some rust out of my system after the first half dozen or so _slow_ pitches. They were still a bit quick for Chalotte, but she got some good practice getting the bat around quickly - and understanding why that's important.
It occurred to me later - I haven't swung a bat in a good, oh, 15 years? I used to be pretty good (up through HS level), and my cut's still OK, though my eyes and my swing need tuning.
It occurred to me this morning - I haven't swung a bat in a good, oh, 15 years? My fingers have been very claw-like and painful all day, and several core muscles that I haven't exercised regularly in the last, oh, 15 years are complaining a bit.
Getting old sucks. But aches and pains in exchange for a nice evening at the batting cages with my daughter is.... priceless :) | | comments: 1 comment or Leave a comment |
| | On our way back from Atlanta, we stopped in with my sister in Charlotte, NC. Although the date wasn't quite Easter, we did the traditional Easter things with the kids. A notable addition to the pack of rugrats is my niece Anna. Pics behind the cut.
( Read more... ) | | comments: Leave a comment |
| | Also as part of the great Spring Break trek, we stopped through at Stone Mountain, and took in the sights. We were transported on a variety of interesting media - amphibious "Duck", train, and sky lift. Pics behind the cut.
( Read more... ) | | comments: 1 comment or Leave a comment |
| |
![[mood icon]](http://l-stat.livejournal.com/img/mood/classic/smile.gif){kind=small}
chipper