LiveJournal Project - December 2004

Hooray, year 3 down. Looks like 2005 has a lot of updates and then it really starts to decline. Are the answers yet to be found or have we already found them? Hmm...

Date: 2004-12-03 15:53
Subject: finals on the horizon
so it looks like ASU has found yet another way to piss me off with this new plus/minus grading system. in my POS 301 class, due to my rather poor attendance record, it's going to take a 99 on the final to get an A, as opposed to an A-. this is certainly doable, but it won't be easy - because if i miss one question, i'm fucked. this wouldn't bother me so much since i was expecting to pull an A+ in my POS 160 class and the two would balance each other out. however, as it turns out, the professor in that class isn't doing plus/minus, so even if i end up with an A+, it's still only an A. recockulous. i really wonder if the people who approved this whole plus/minus bullshit (cough michael crow cough) really bothered to take into account all the ways in which students can now be screwed by having some professors that do plus/minus and some that don't. i know i won't be getting an A+ in cryptography; i think even if i ace the final i'll be a percentage point or two too low. so that leaves japanese - and who the hell knows what's going to happen there. A? A+? she'll probably give me an A out of spite, regardless of the fact that i probably have higher test/quiz grades than anyone else in the class. bleh. in semesters past, i wouldn't give two shits if i had a 91 or a 99 - because an A was an A. i don't really see how it is that the powers that be at ASU really believe that students' GPAs are going to increase under this new scheme. i think it's also kinda bizarre that if i do finish the semester with all A's and one A-, my semester GPA will be 3.93 - and i'll be pissed about it. most people would probably be happy with that. =/

in other news... the pigskin project is underway - phase 1, gather data. once the code is finished and i've had a chance to run some simulations, i'll probably have more to say about it. but for the moment, just know that i've got some interesting ideas in the works.

and now we return you to your regularly scheduled world domination in progress...

Date: 2004-12-04 21:23
Subject: not much to see here, just a self-reminder.
so, regular readers of this journal would probably note that i often vacillate between being bored off my ass and not really knowing how to keep myself occupied and between being completely overloaded with shit to do that i really just don't feel like doing. i think, for a moment, at least, i'm happy to say that i've acheived a state which is neither of those. overloaded with shit to do - yes - but it's all shaping up to be stuff that i want to do. i'm not sure if it's because the semester is coming to a close and i'm freeing up neural cycles to do other things, or if maybe it's just the alignment of the planets, but there's a ton of cool shit in the queue these days. a short list, as a reminder to myself...

...the pigskin project
...searching for chicken little
...r/t video contest
...poketron (no, not pokemon)
Music: INXS - Never Tear Us Apart
Mood: hungry

Date: 2004-12-06 16:00
Subject: die, evil dog-like creature!
Quiz name: Which absurd cat are you?
Result: You scored as Pissed at the World Cat
What's it mean: And here we have the next serial killer. Try having some cotton candy, it'll make you feel all warm and fuzzy inside, Psycho.

Date: 2004-12-09 15:58
Subject: 1 down, 3 to go, so pissed off!!!
i walked out of the crypto final thinking that i'd easily scored in the upper 90% range. then i thought about it a little more, and i realized that once again, dumbfuck mistakes had bitten me in the ass. i knew how to do every problem on the exam, but for some reason which will forever remain a mystery, i wrote down 199-1 = 190 - and of course, this fuckup blew the computations for the entire rest of the problem. then i realized another mistake of a similar nature on a different problem. in both cases, i knew the theory behind the problem, but stupid fucking first-grade arithmetic kicked me in the balls. if this were an exam done in the math testing center with unlimited time, i'd have had the time to check my work and catch these errors. but noooo.... time crunch on math exam make ravyn become stupid!

so now i don't really know what i got on the crypto final. i need a 52/70 to get an A- in the course and a 61/70 to get an A. based on my estimations of partial credit, i'm predicting 62-64, which is fine; however, this still doesn't stop me from being severely annoyed. tomorrow is japanese and research methods. not too concerned about japanese, but i need to ace the research methods final to get an A. (87 for an A-) joy.

Time: 21:42
Subject: a letter i'm writing to michael crow...
Dear President Crow:

It is my understanding that the decision to use plus/minus grades is in all cases left to the discretion of individual faculty members. While I certainly respect the right and obligation of faculty members to evaulate their students' work as they see fit, I believe that the lack of uniformity of implementation created by this position engenders a situation in which students are ultimately treated unfairly. For instance, the grade of A+ was added to the system as a means for students at the higher end of the achievement spectrum to offset any grades of A- that they might receive. The success of this plan, however, is predicated on the idea that it will be possible for students to get the A+ in the first place. This is not a question of achievement, but a question of availability. This semester I have seen courses in which the professor is not doing plus/minus grading at all as well as courses in which the professor is offering an A- but not an A+; so no matter how well a student does in those courses, he simply cannot get an A+. Should this student receive an A-, his GPA will go down, and there is nothing he can do about it except hope for the best come the following semester, in which it is entirely possible that this exact situation may repeat itself.

Moreover, suppose that this same student has done A+-quality work in a class in which plus/minus grading is not used or in which the A+ is not available. When grades are posted, he will have only an A. Should that student later decide to apply to graduate school, the admissions committee at his school of choice will see no differentiation between the A received in that class (which should have been an A+) and an A that he might receive in another class which was genuinely just an A. If part of the rationale behind plus/minus is to enable a more accurate and precise evaulation of student work, then a situation as I have described here is actually presenting an INCORRECT evaluation of the quality of that student's work as well as artificially lowering his GPA.

The most obvious way to solve problems such as these is to eliminate plus/minus grading; although this would admittedly be my preferred solution, I recognize that it is an unrealistic option, and so instead I would like to offer a set of recommendations that I believe would improve the system so that it is truly fair to all students.

1. Require all faculty members to use it. This is essential, both to providing the most accurate evaluation of student work as well as to avoid punishing students who do exceedingly well in courses where plus/minus grades are not currently being used.

2. The A+ must be an attainable grade. It was added to the system to help students offset potential A- grades, but its existence in the system does not help them if there is absolutely no way that they can achieve it. Moreover, the grading scale in the A range should ideally be the same as the grading scale for any other letter grade in which plus/minus grades are possible. If, for instance, a B+ is 87-90, then an A+ should be 97-100, not 99-100.

3. Faculty should be required to include information about the grading scheme on their course syllabi. Students should know from the outset exactly what is expected in order for them to achieve the grades they desire. This should include information both about whether or not plus/minus grades are being used as well as what the cutoffs are for each fractional letter grade.

I realize that this is the first semester that plus/minus grading will be in effect at ASU, and as is inevitable with the rollout of any new policy, there are likely to be bumps along the way which will gradually be smoothed out. I also know that this is the best time to offer solutions to these issues - before existing practices become firmly established and more resistant to change. In the interest of fairness to all students and the realization of a system which is closer to what was intended than the existing one, I urge you to consider these modifications to plus/minus grading.

i'll probably fire this off within the next few days. comments and feedback much appreciated.

Date: 2004-12-17 01:20
Subject: captain, i'm giving her all she's got!
(said in my best mr. scott voice)

you know you've got a computationally-intensive simulation when even a parallel implementation of your code running on a 64-bit dual-processor machine is still making you scream "fuck, this is too damn slow!" and there's not really too much i can do about it. i've already optimized the code 10-fold, and even though there are other machines in the house that i could consider adding to the cluster, because they're so much slower than the machine i'm currently using, given the way the process scheduler works, i think i'd actually end up making things worse. i should go to bed. it might be done in the morning.

96,000 generations remaining. current best-fit estimate: 756.7 (smaller is better - if i get down to anything under 100, i'll be really stoked. under 50 and i'll probably piss myself.)

and yes, eventually, for anyone that gives a rat's ass, i will explain what all of this is for if it turns out to be something worthwhile.

more notes for myself... it appears that i get better as well as faster convergence when i precompute differences, rather than expecting the GA to figure it all out. it's not looking like this particular run is going to sufficiently converge (changed error / fitness function to mean absolute error instead of total error magntude, but i'm still looking at an average error around 12, when i need it to be under 5), so pass 2 will need to be run with a different set of preprocessed data. at least this may end up speeding things even further, since it'll cut the number of variables in half.

edit 2... incarnation 1 has proven to be a failure. there's not much point in allowing it to run to completion. 47,300 generations, best fitness value: 12.31 (mean absolute error) - and it looks like we've reached a point where it's not going to improve. so, moving on to pass 2 / pass 3...

edit3... this is one of those aha! sort of moments... as it turns out, increasing the population size is a better way to improve fitness (at least for this particular project) than running more and more generations. after 1400 generations, running with a population size of 20,000 (instead of 1000, as i was doing before), my best MAE fitness value is 8.4 - almost a 50% improvement, and it's still going down- but i still need a factor of 2 improvement from this current level. so, if a 10-fold increase in population size produces a 50-100% increase (decrease) in fitness, where do i hit the limit? would a population of 100,000 work even better, or would it be waste of resources? and of course, the larger the population, the longer (MUCH longer) it takes to run to run. which brings me to my real problem - i need more bloody CPU power, and in a big way. i've got a lot of different ideas for tweaks and different directions i'd like to take with this, but if it's going to take me a day or two for each sim to run to completion, who knows how long it will be before i even sort through a lot of my first-order ideas. i could, i suppose, go drop several thousand dollars on more hardware, but that's not really an expense i can justify unless i know for certain that i'm going to get usable results - not to mention the fact that there's no room in here for more boxen and no network drops on the other side of the house. (and do i really want to go to sleep every night listening to the incessant whirring of case fans from 4-5 computers???)
Mood: hopeful

Date: 2004-12-19 03:38
Subject: back to our normally-scheduled programming (ranting)
so, by now anyone who's had any contact with news media outlets in the past few days knows the story of lisa montgomery, the woman who went over to another woman's house (whom she had met on the internet, no less!), strangled her, and sliced the fetus out of her womb and then went home to her own family and said that she'd given birth. now, normally not too many fucked-up things that go on in this world really elicit much display of surprise from the jaded ol' blackbird, but this one truly baffles me. i can understand what might motivate someone to do a lot of really twisted things (even the woman who sliced the arms off her infant doesn't strike me as all that befuddling given various other parents that have committed similarly-twisted acts of infanticide over the years) - but what could possibly possess someone (satan made her do it!) to go kill some random stranger (premeditated murder, even, not just some random act of violence!) and slice her open just to steal her unborn baby?!? the media reports say that montgomery had a miscarriage awhile back - as if they're trying to suggest that maybe she lost her marbles because her own baby died and had to go get another one.

HELLO?! lisa montgomery, what the fuck is wrong with you?! babies aren't that hard to make!! you've got a husband, you murderous twat, and it's obvious that you're able to have children since you have two already, so what the hell were you thinking? rather than go fuck up some poor bastard's life (i.e., the widower of the newly-deceased) by taking his wife away from him (and almost getting his kid, too), all you had to do was get with your husband, fill him full of beer so he'd forget how ugly you are, and start screwin'! but nooo... you had to make life difficult for people (including your own family, dumbass) by unleashing your little failed-to-take-her-lithium miscreance upon the world. good fuckin' job. i bet you won't even make it to your trial. i'm sure that somewhere inside leavenworth there is at least one angry woman who's just waiting for a choice opportunity to demonstrate the effectiveness of her homemade knife collection. you could use a little plastic surgery.
Mood: shocked

Date: 2004-12-23 08:10
Subject: more thoughts on politics, dubya, and iraq (but really nothing new)
the news today says that they now believe that the cause of the latest attack on US forces in iraq was due to a suicide bomber, supposedly someone who'd been working on the base for the last four months. what really boggles my mind is that they seem to be SUPRISED by this. come on, you idiots, do you really think that every iraqi who's working with you is doing so with love and hope in his heart? i don't know about the rest of you folks, but if i were an iraqi, pissed off at the fact that my country had been overrun by a bunch of christian zealots, lacking in any sort of real military training and without a stash of RPGs with which to go on a rampage, you can bet your ass i'd try to become a sleeper agent. hit the bloody infidels in their soft underbelly!

and in more news... a few days ago, dubya held a news conference, in which he came right out and said that he believes that "free nations don't attack each other" - in other words, he believes in the democratic peace theory - and it seems that he holds this belief without reservation. i'm sure i've said this before, but i'll say it again anyway - that's some scary shit. it's one thing to sit around in the hallowed halls of academia and debate theories, but when you take something which is merely a hypothesis that's far from universally validated and base the foreign policy of the world's most powerful nation around it, you're really asking for disaster. i wonder if mr. bush has ever actually read any of the criticisms of DPT - the most salient one at this point being that forced democratization is a very unstable process. it's one thing to look at the countries of eastern europe after the end of the cold war, but assuming that models which seem to work pretty well in the case of poland will apply to places such as iraq and afghanistan is completely ludicrous. the eastern europeans wanted democracy. they wanted to be a part of the EU, part of NATO, and to enjoy the various economic benefits that would come with all of that. moreover, there was never really much of a cultural or religious divide between eastern europe, western europe, and the US. so it's no real surprise that democratization seems to be working for them - especially since it wasn't foisted upon them by an external power.

now let's look at afghanistan and iraq. here we have islamic (or at least arab, since saddam hussein's baathists were secular) nations being overrun by, as i said earlier, christian zealots. maybe bush doesn't see it that way, but then again, bush probably never bothers to look at things from the other side's point of view. we've gone in, ousted their leaders (who were rat bastards, don't get me wrong), and now we're attempting to turn these two bass-ackwards countries into shining pillars of democracy in the middle east. who the fuck are we kidding?! bush makes a big deal about how the afghans have held elections and that elections are coming up in iraq next month. but you know what? elections don't mean dick. they had elections under saddam hussein. and just because a candidate receives a majority of the votes cast certainly doesn't mean a: that the elections were fair (look at what's happening in the ukraine), b: that a majority of the PEOPLE support the winner, or c: that anything is really going to change. i mean, come on, what does bush really think is going to happen after the elections in iraq are held - that the insurgents are just going to say "oh, well, we have a president now, let's put down our bombs and go home" ?!? hell, we've still got allegations of voter fraud from our own last presidential election that haven't been (and probably won't be) resolved - and we're going to tell someone else how to do things??

i don't know why it's so bloody obvious to so many people that we've really screwed the pooch over there yet bush and his inner circle still can't (won't) see it. if they're so damn concerned with building a free iraq, then why don't they just buck up, admit that they're bumbling it, and do what they need to do to FIX THE PROBLEM. if that means we have to send more troops to get things stable - fine, do it. if it means that we have to go to the UN with our tails between our legs and ask for help - fine, do it. we broke it, now we gotta buy it - and the longer we dick around, the more it's going to cost.

Time: 13:49
Subject: going to someplace cold.
off to NY/CT to do that overcommercialized december holiday with _purpleglitter_ and her family. back on the 27th. having never been to NYC before, i wonder what i'll think of it. i wonder how much hassle i'll get from the security droids. i wonder if i'll be able to find a place to park at the airport... =/

Date: 2004-12-28 13:53
Subject: i'm probably the only one who cares, but...

After a long absence, the reclusive Kate Bush is suddenly popping up everywhere — in an interview with OutKast's Andre 3000 last month, in which the rapper says he'd love to record with her; in the title of John Mendelssohn's recent novel Waiting for Kate Bush, which opens with the protagonist threatening to leap from a tall building unless the "Wuthering Heights" singer promises to release a new album within six months; and now, in her own announcement to her fans that seems to answer the Mendelssohn character's plea. In a letter to her fan club, the 46-year-old singer writes that her new album, her first since 1993's The Red Shoes, is finished and will be out in 2005. London's Sunday Times reports that the album may be out as soon as March.

from http://www.ew.com/ew/report/0,6115,1012101_10_0_,00.html

Time: 16:48
Subject: balance, grasshoppa
Quiz name: What is your alignment?
Result: You scored as True Neutral
Meaning: A True Neutral person has two faces- either these people are merely apathetic, preferring to focus their minds on more important things, or these people truly believe in a balance of all things. To these people, there can be no light without some darkness. These people also have no dedication to, or intrinsic distrust of, laws.
Editor's note: I don't know if I agree with this one. Maybe it seemed more accurate at the time.

Time: 17:52
Subject: ok, some substance this time.
weekend recap...

my weekend in CT with 's family went about as well as it could have - which, rather than being a statement of resignation, is actually a pretty favorable assessment. the flight up there was hell - i was stuck in a row with a couple of PDA-happy lovebirds that couldn't seem to keep their hands off each other. blech. go in the damn lavatory and join the mile-high club and leave me the hell alone. i took countless shots to the shoulder from people walking down the aisle of the plane - which only reinforced my belief that anyone over 200lbs should not fly coach on long-distance flights. next time i'll know better. the woman in front of me put her seat back right at the outset of the flight, denying me the ability to use my laptop - but it didn't matter anyway, because i didn't have the deCSS libraries installed - thus making playback of the DVDs i'd brought with me impossible anyway. ugh.

get to the airport, cory picks me up, and we drive out to her parents' house which truly is in the middle of nowhere. i meet the dad, and the brother, and we seem to get along, so that's a bonus. people start rolling in for christmas eve dinner around 4:00, and over the course of the night they just keep coming. didn't really have any complaints with any of them, although i did find one person i met just a little bit, well, um, strange. also, i got to hear the quote of the weekend - when a couple of the people were getting ready to leave, one of them said "you're much better than the last one." ha! too funny. overall, it seemed to go pretty well.

christmas day was much quieter - only two people came over for dinner - and i proceeded to get into a discussion with one of them about iraq, the war, and such, much to the dismay of cory's mom and the dude's wife, who basically ended up coming into the living room and telling us both to shut up. the conversation went a little like this... me: saddam had nothing to do with 9/11. him: i disagree. me: what evidence do you have? him: saddam paid money to suicide bombers in israel. me: what does that have to do with 9/11? him: israel's a democracy. uh, what?! ok, dude. whatever. later in the evening we ended up talking about less inflammatory topics, and i got a few more project ideas, so overall i can still call it a success.

the day after, i went to the gym with cory's brother, and then she and i went into NYC to poke around. and, well, all i really have to say about it is that it doesn't deserve all the hype. it's a city, it has stuff in it. big deal. i didn't hate the place, as ryanjamieson suspected, but i wasn't in awe of it. we had some pizza in little italy which wasn't too bad, but it was too thin and crispy for my tastes - cory says i can't say that i've really tried NY-style pizza yet, though, so i guess i'll have to reserve judgement for now. chinatown was interesting - lots of people trying to sell fake purses and rolexes. it's obvious the stuff is fake, but people must buy it or there wouldn't be so many people trying to sell it. anyway, we took the subway over to rockefeller center, i saw the tree and all that good stuff, and we wandered around some more. cold like chicago, crowded like london. again, no real opinion - it's a city. it has stuff in it. i'd never live there.

monday - back on the plane for a return to the land of cactus. woo.

Date: 2004-12-29 22:06
Subject: hoo-ray
new photo, new style. can you tell i'm bored and don't feel like working on feature creep?

Date: 2004-12-30 18:05
Subject: on disconnectedness
during the last 50 years, communications technology has become smaller, cheaper, faster, and more ubiquitous. we have our 24-hour cable news channels, we have our internet, we have our newspapers, our embedded reporters, our satellite radios, our weblogs - all coming together to form one unimaginably large pool of information that's there at our fingertips, just waiting for the push of a button, the flick of a switch, or the keying in of the right query. global economies have become more interconnected as free trade has expanded, nations have specialized in one way or another, and as intercontinental travel and transport has become more advanced. as a result of all of this, people say that the world is shrinking.

they're wrong. the world is just as large as it's always been. just because i can turn on my television and see what's happening on the other side of the globe doesn't actually put me there. it doesn't give me any feel for the atmosphere. it doesn't give me any connection to the people and the events that are unfolding in these places. if i want to see 120,000 people dead as the result of a natural disaster, i can turn on CNN and watch coverage of the tsunami's aftermath, or i can go to blockbuster and rent a movie. what's the difference? one of them is real, one of them is fictional - sure - but at the same time, in a sense there really isn't all that much difference. these people that were washed away in the real disaster - if you didn't actually KNOW any of them - are just as nameless and faceless as the characters in a film. i watched 9/11 on television as it happened - first the towers were smoking, and then they collapsed - live, in real-time, in my living room. and you know, the first thing i thought - damn, that looks like something straight out of a movie. no sorrow, no shock, no anger. i didn't know anyone who died in 9/11, so really, aside from the bullshit that's unfolded in this country and around the world as a result of it, the event itself simply had no meaning for me. i suspect that i'm not the only one that's felt this way.

so what does it all mean? is it simply my (our) own disconnectedness from that ideological construct that we call the "whole of humanity", or is there something larger going on here? how many people can genuinely say that they care about a bunch of random dead folks on the other side of the world, and how many are simply saying that they care because it seems the right thing to do, or because they're afraid to say that they really don't give a shit because they're afraid it'll make them look bad? how can people be upset about the loss of something that they never knew in the first place?

Time: 20:20
Subject: rant #2
y'know, nobody's going to deny that this tsunami was a big disaster and that it's going to cost a shitload of money to rebuild everything over there - but i am getting really fucking tired of getting hit up for money everywhere i turn. message boards, livejournal, television, news websites, radio - you name it, somebody's out there screaming for donations. but if all of that weren't enough to piss me off, now i'm getting spam about it! jesus h. christ, people, i heard you the first twenty times, there was no need to violate the sanctity of my inbox. any possible inclination i may have had at any point to contribute to the relief effort has been summarily erased.
Mood: annoyed