Viraj's Weblog

David Cronenberg is a director that can take something that we see in almost every TV show nowadays, a gunshot, and turn it into the most visceral and graphic scene imaginable. He pushes the limits on what the viewer will accept. Living in our society dulled by all the violence that pervades television, this director shows us just how disturbing death and violence really is, without actually pervading the movie with such scenes.

Cronenberg's movies are about the dark side of human nature, the animalistic, raw, uninhibited. He interwines such polar ideas like violence and love, sex and anger, and does it in such a way that you begin to question what it means to be human. You start wondering whether you are anything more than an animal. Ebert said this movie is about the survival of the fittest, and I have to agree with that.

I really thought the movie was amazing. It plays like a chess game, each move coming closer and closer to a catharsis. It's as intelligent as A Clockwork Orange, and I highly recommend it for anyone that wants to come out the theatre with more questions than answers.

The trailer for this movie plays down the violence in the film. There are some quite graphic scenes and I wouldn't recommend any children seeing it.

The above picture is from a BBC show about violent behavior in humans. I found it hilarious. Scientists did a study and found that murderers had lower brain activity (glucose metabolism) in the prefrontal cortex. They say this correlates physical differences in brain structure to different types of behavior, notably violence. I can see it now, brain scanners at airport security checks. A flashing red evil light would blink whenever such a person came through. Then they could be re-routed to the psychotic plane, or short-plane, like the short-schoolbus for mentally challenged kids. Can we really dilineate a murderer from a normal person with such scans? It seems rather simplistic to me.

There are all sorts of experiments on violent behavior. The show claimed that low serotonin levels and high dopamine levels in the brain directly contribute to violence. Serotonin acts as an inhibitor, to stop feelings and thoughts from getting out of control. Dopamine is just the opposite. So scientists are searching for genes regulating those chemicals, and one such gene is the one encoding MAO-A. Studies have found that not enough MAO-A could contribute to violent behavior, because it results in lower serotonin levels. But if the MAO-A gene is expressed, there is more serotonin, and there is less tendency for violence.

So what makes the gene expressed? Nurturing. It turns out if a child is nurtured by his mother and is taught at an early age about social bonding, love, bad behavior vs good behavior, the child has a less tendency to grow up with problems. Take the same child and rear him without motherly love and he or she is more likely to turn out problematic. Is this really surprising? No, but to me the environmental influence on genetics is. There is a chemical basis to the behavior. Someone predisposed genetically to a certain behavior, or even disease, can in effect cancel out, or amplify, a gene based on their experience.

This is one of the many examples where genetics by itself cannot determine anything without also considering the experiences of the person. The person's experience itself determines whether the gene is expressed or not, and it ultimately shows that the brain can control gene expression. I find that rather amazing, and it shows the true power of the brain. It's not too hard to fathom the brain having control of other disease-related genes. Let's say you have a gene that predisposes you to cancer. Is there a possibility the brain can control whether that gene is expressed or not? I don't know, but it's interesting to think about.

No I'm not talking about the $3 per gallon kind. I'm referring to flatulence, or farting.

I was reading my Chemistry book and it mentioned a product called Beano, which is an over-the-counter supplement to prevent gas when eating certain foods. Why in the world my college Chemistry book would be advertising fart-reducing pills is beyond me, but it did get me interested.

Now I'm not saying I have a flatulence problem, but there are some foods that I eat and well, umm, let's just say I don't smell like Gucci afterwards. It's actually strange, because its the healthy foods that give me gas. Things like lettuce, broccoli, other green things, sweet potatoes, certain breads, etc. So I pretty much hate eating such foods. I know they are good for me, but they always give me stomach problems. I can't stop eating such things because I'd like to have a healthy diet once in awhile. It sucks because I can't even eat a salad without getting some problems.

They have a pretty good explanation of what gas is on Beano's website:

What causes gas?

The body does not digest and absorb some carbohydrates in the small intestine because of a shortage or absence of certain enzymes. This undigested food then passes from the small intestine to the large intestine where bacteria break down the food, producing gas. The most common symptoms of gas are flatulence, abdominal bloating and abdominal pain. In essence, the body lacks the enzymes needed to breakdown the carbohydrates found in some gassy foods like vegetables, legumes, grains, cereals, nuts, seeds and whole-grain breads. This is where Beano® steps in. Beano® contains a food enzyme from a natural source that works with your body’s digestion to breakdown the complex sugars in gassy foods making them more digestible, thereby preventing gas before it even starts.

My Biology teacher described this symbiotic bacterial environment in class. Basically, when you are born you don't really have much bacteria in your intestine. It takes some time to build up that "flora," as my professor put it. I don't know about you, but "flora" gives me the impression of a garden growing in my colon, and that's not very cool to say the least. I wonder if this bacteria can ever get out of control. Well if you know someone that farts alot, then I think the bacteria have the upper hand in that person.

Wikipedia also has some more delightful information on gas. I especially like this bit:

If sitting on a cushioned surface, the gases can be directed into the open-cell polyurethane foam and somewhat quarantined. Following the fart, standing will not release the odor, in fact, the gases will be further pushed to the center of the cushion. The gases will not leak out and be detectable, unless the cushion is compressed again under the weight of another person. The use of this phenomenon as a practical joke is obvious.

Wikipedia always has funny wording, which changes alot of course. Like on this page about sleep, the caption for the picture is "A human female sleeping." They've since changed that.

Anyway, Beano seemed like something worth trying, so I bought some at the local CVS. You are supposed to take it with your first bite when eating certain foods. All I can say is wow, it really works. I ate something that I knew for sure causes problems for me, and when I took Beano with it, there were no problems at all.

So buy yourself some if you have gas problems. Hm, I'm probably going to regret ever writing this.

I was reading a brochure from the Juvenon guys about aging experiments. They describe an experiment where scientists took away an enzyme that checks for mitochondrial DNA errors in mice, sort of like an error-checking enzyme. Not surprisingly, more and more cells in those mice started getting mutations in that DNA. What they observed however was that the mice appeared to have accelerated aging.

So they are claiming that mutations in mitochondrial DNA correlates to aging, and it would make sense to say that if you can limit those mutations, the cells might live longer. That's the selling point of Juvenon.

Disregarding whether the drug actually works, what I find interesting is this ability to accelerate aging in an animal by isolation of this enzyme. Cloning studies could benefit from this. When animals are cloned, the scientists do not really know whether it is truly healthy or not. Many problems arise later in the animal's life, so they have to watch it for a long time. It's very difficult to 'prove' that the cloning resulted in a successful organism. That's why we don't see such quick advances in cloning, it requires time for analysis of the organism. But what if we could accelerate the aging of the animal in this process? Basically doing a fast-forward through the animal's life in order to analyze it's health without actually waiting and observing the normal lifespan. Then by doing many studies like this, we could find out the problems in the cloning techniques quicker and try new ones.

As far as Juvenon goes, I've tried taking it on an off. In some instances it made me a bit jumpy, like sort of 'wired.' I seemed more energetic and could think a bit faster, however it was difficult to relax. I had trouble sleeping when I took it before bed, so I won't do that again. I have a pretty messed up mind as it is, so this may be less than beneficial in my case. I can see how it would help older people though. I may continue taking in lower than the recommended dosage. They suggest 2 tablets daily, but I might take 2 every 3 days or something like that. It's all about dosage and how your body responds to it.

I have also been taking fish oil supplements, which I feel is good idea. I don't eat any fish at all, and there are many studies showing that such nutrients are beneficial. Honestly the best regimen for me involves napping during the day. I can think alot sharper and feel alot better. Polyphasia should not be slept on .

A pretty funny ad on 'examining' yourself.

Here is your typical sysadmin day.

Update: There is a song as well.

This week I've been doing some consulting. A heavily loaded website (about 300 req/second) was having some issues such as Apache processes going haywire and eating up all CPU, and Apache processes segfaulting continously. However, the site was functioning, and only when there were alot of apaches eating up CPU did the system go to a crawl. The segfaults were of the apache child processes, not the main one.

The systems (load balanced) were hosted at Rackspace, and they pretty much set everything up for the customer, with things such as PHP, Apache, MySQL, etc. So the first thing I looked at are the haywire httpd processing eating up CPU. What could I do? Well an strace on it to start:

strace -p PID

But it didn't show anything, so it's hung somewhere before any other system call is made. Next thing, attach a gdb to that bitch:

gdb /usr/sbin/httpd PID

Then I did a 'bt' to show a backtrace of function calls and end up with:

0x00265383 in malloc_consolidate () from /lib/tls/libc.so.6
(gdb) bt
#0 0x00265383 in malloc_consolidate () from /lib/tls/libc.so.6
#1 0x00264b29 in _int_malloc () from /lib/tls/libc.so.6
#2 0x00263ecd in malloc () from /lib/tls/libc.so.6
#3 0x0040f500 in OpenSSLDie () from /lib/libcrypto.so.4
#4 0x0040fb5c in CRYPTO_malloc () from /lib/libcrypto.so.4
#5 0x006bd86b in ssl3_new () from /lib/libssl.so.4
#6 0x006c41e2 in tls1_new () from /lib/libssl.so.4
#7 0x006c5c66 in SSL_new () from /lib/libssl.so.4
#8 0x00639545 in ssl_engine_disable () from /etc/httpd/modules/mod_ssl.so
#9 0x080722bc in ap_run_pre_connection ()
#10 0x08072476 in ap_process_connection ()
#11 0x08066ac1 in ap_graceful_stop_signalled ()
#12 0x08066c14 in ap_graceful_stop_signalled ()
#13 0x08066eb9 in ap_graceful_stop_signalled ()
#14 0x08067550 in ap_mpm_run ()
#15 0x0806da2f in main ()

Hmm, ok so it's in some malloc call having to do with SSL. Well this site is not even using SSL, so the next thing I did was disable SSL. Not surprisingly, this reduced the frequency of the process going haywire, but did not rid of them entirely. Something else is going on. From the malloc calls, it appears having to do with maybe a memory leak or double-free happening somewhere. The next time it happened, it was hung in some pthread call, but not really related to any apache module. The problem is somewhat similar to this bug.

At this point I really had nowhere to go on this CPU issue. Apache was also segfaulting, and I guessed that the segfault problem is related to the CPU problem, so I decided to look into that. Now it gets really interesting. Trying to debug an Apache server getting 200-300 req/sec is not very fun. First of all you can have virtually no downtime before everyone starts realizing it, so whatever I needed to do, I had to do it with the least impact.

Apache reported to the logfiles something like:

[Thu Sep 15 04:14:26 2005] [notice] child pid 24195 exit signal Segmentation fault (11)

I also saw lines such as:

Allowed memory size of 16777216 bytes exhausted (tried to allocate 10 bytes)
free(): invalid pointer 0x9910020!

But this is not very helpful to me. What was PID 24195 doing? Which file was it serving? I first added PID logging to the logfiles with the '%P' custom log statement, similar to:

LogFormat "%h %l %u %t \"%r\" %>s %b \"%{Referer}i\" \"%{User-Agent}i\" PID: %P" debuglog

Now I could correlate the segfaulted PIDs with the requests. But get this, it told me nothing! All of the crashes seemed to be random, some on .php files, some on even .jpg files! WTF? This told me more and more there is a memory leak somewhere, and the symptom is not showing up until much later. Next step, I want Apache to dump cores. This was done by adding a:

CoreDumpDirectory /tmp/cores

Great, now I have all these core files. I look at a backtrace:

gdb /usr/sbin/httpd core.12345
bt
#0 0x06ca51b0 in _zval_ptr_dtor (zval_ptr=0x53de8f3)
at /usr/src/debug/php-5.0.5/Zend/zend_execute_API.c:395
395 zv->refcount--;
(gdb) bt
#0 0x06ca51b0 in _zval_ptr_dtor (zval_ptr=0x53de8f3)
at /usr/src/debug/php-5.0.5/Zend/zend_execute_API.c:395
#1 0x06cb71f8 in zend_hash_destroy (ht=0x8821c3c)
at /usr/src/debug/php-5.0.5/Zend/zend_hash.c:519
#2 0x06cae8de in _zval_dtor (zvalue=0x881f47c)
at /usr/src/debug/php-5.0.5/Zend/zend_variables.c:52
#3 0x06ca51d8 in _zval_ptr_dtor (zval_ptr=0x53de8f3)
at /usr/src/debug/php-5.0.5/Zend/zend_execute_API.c:397
...

I see pretty much random location crashes, but they are all within PHP shutdown sequences. This is still not telling me much. The next thing I wanted to do was an strace on a working apache process and wait for it to crash. Luckily, the crashes were so frequent, I could strace on literally any apache PID and after a few minutes would be guaranteed a segfault.

Now what I saw was interesting. Almost always before a segfault, I would see some Mysql traffic that involved trying to access a nonexistent table. It turns out some the website's PHP code had some legacy stuff doing this. So I thought, ok a missing table is not supposed to segfault apache, but lets get this code fixed either way. After a note to the developers, it was fixed. And guess what? There was a significant decrease in the number of segfaults per minute. Before it was about 1 every second, now it was 1 every minute. Clearly something was related to mysql, but at this point in my debugging process I just did not realize it.

All of this has been compiled and setup by Rackspace. During all of this, I'm opening tickets on their system and having them look into the problem. They are doing their own debugging, and even late one evening ran apache within a gdb and watched it. They pretty much thought it was bug in PHP, and asked us to report it to the PHP developers. Rackspace did do alot of debugging and I was impressed with their work, but I felt this was just going in the wrong direction. We spent many days trying different debugging techniques, PHP versions, and what not. But still no luck. At one point we found one .php script that ALWAYS segfaulted, and it had to do with XML stuff. We installed stable XML/XSLT libraries and it went away for that script, but still segfaulted elsewhere .

Rackspace showed me some straces of their crashes. They were all meaningless except for one thing, I noticed that shortly before each crash, there was a mysql insert statement executed. The fact that a mysql 'table not found' had increased the segfaults got me thinking, could this be something related to mysql or libraries?

I did some searching and came across this bug, which explained that PHP might crash if it was compiled with one mysql library API version, but there is a different version on the server. The next thing I look at is phpinfo(). I see it was compiled with mysql librares 3.23.58, however the mysqld running on the system was 4.1.11. I had a eureka moment. Could a change in data structures between the versions in the API cause some issues? I thought that could likely be the case.

It just so turns out, that when Rackspace compiled PHP, they did so on a 'build host', which had mysql 3.23.58. Then they took this PHP and installed it on our server. I asked them to recompile it with the same mysql version that we have. Well they did so, and it has been about 5 hours and not one segfault. I think we got it. I'll have to watch it longer, but sometimes you just get a feeling that you just know you figured it out.

This all took place over a week. When I should've been studying, instead I was in the FAU library debugging apache crashes . It's all good though, because I am still learning something in the process. It is fun to do these kinds of things. Now I just need to make my invoice .

Lesson learned? Build systems are not very good if they differ from the target system.

Update: This did in fact fix the problem.

2 days ago I got in a little accident on I95. It was luckily very minor. It was around Gateway Blvd. northbound, and that place is always jammed around 6-7pm as many lanes merge into two. Anyway, first the guy in front of me slams on his brakes. Then I do and successfully stop before hitting him. Then I watch in my rear-view and see the guy behind me successfully stop ok.

At that point I'm like 'whew' and go on listening to my music. Then BAM, the guy behind me runs into me. It turns out a lady behind him hit his car which pushed it into mine. We all 3 pulled over and I didn't get any damage to my rear bumper. Neither did the guy behind me to his front bumper. We decided it was not enough to report it. I'm not sure what happened to his rear bumper, but the lady was apologizing profusely saying she had a baby in her car distracting her, etc. I said "is everyone ok?" and then took off. No point in increasing my insurance over this if everyone agrees to let it slide.

Anyhow, I95 sucks.

Here is a very heartwarming story about someone dedicating his life to help kids in Haiti.

I read this in The Homeless Voice, which is a great newspaper sold to support the homeless. The articles are all very inspiring and touching. It makes you realize how important someone, even a homeless person, can be. So it really saddens me to see on the website a story about teenagers beating a homeless man to death "for fun."

There is also a great little 'advertisement' in the newspaper. It shows a picture of someone resembling Jesus and holding a 'Will work for loaves and fishes' cardboard sign. On the top it reads:

How can you worship a homeless man on Sunday, and ignore one on Monday?

Indeed.

It was just pointed out to me that OS X is case insensitive:

Viraj-Alankars-Computer:~/a virajalankar$ ls -l
total 0
-rw-r--r-- 1 virajala virajala 0 11 Sep 09:31 a.txt
-rw-r--r-- 1 virajala virajala 0 11 Sep 09:31 b.txt
Viraj-Alankars-Computer:~/a virajalankar$ mv b.txt A.txt
Viraj-Alankars-Computer:~/a virajalankar$ ls -l
total 0
-rw-r--r-- 1 virajala virajala 0 11 Sep 09:31 A.txt

Hmm, that seems kinda dangerous doesn't it?

I saw a BBC documentary about the brain and they were talking with one person who had incredible memory skills. They would shuffle 20 decks of cards, have him look through the decks, and then ask him to recite every card in each deck in order. Of course he was able to do it, but how?

It turns out his method involved first some preparation. He took a walking trip around the city and got into his mind certain locations and monuments. Then he associated each card with some object, say a queen of spades being a dog. He then created a story that described him walking around the city and seeing strange happenings with his cast of characters. The wierder the scenario is, the easier it is to remember. The location aspect makes a big difference, because his storyline proceeds as he walks through the city.

Another person had to memorize a list of unrelated words, and imagined themselves walking through their house seeing all of these strange object that were or were related to the words he needed to remember.

There are all sorts of memorizing tricks, but the storyline and location method was new to me. I decided to try it for my first Biodiversity exam. It works, and I'm pretty sure I got a 100% on the exam. This class is pretty much rote memorization of phylums, species, and what not. For some of the sections I just created a very wierd scene in my head. Others I created a ridiculous story. Either method would have images that would 'hint' at what needs to be remembered. Let's go through an example.

First here is a small section of notes that I needed to remember:

Phylum Pyrrophyta - Also called dinoflagellates

- Chlorophyll A+C
- 2 flagella
- cellulose cell wall
- Pfiesteria is an example genus
- These organisms cause red tides
- zooxanthellae are the organisms in corals
- some exhibit bioluminescence

Here is the 'story' that I used to remember this:

A pyromaniac with 2 whips was very fiesty and was walled up in a zoo. He drew blood from zoo animals to supply energy to his glowing coral necklace. He was cadaverized by a dinosaur.

This is absolute rubbish. The main point is to not so much memorizing the wording, but the imagery. Once you get the crazy images in your head, the words will come. Then these words are hints towards what is really needed to be remembered. The word pyromaniac is linked to pyrrophyta. The fact that he was killed by a dinosaur gives rise to 'dinoflagellates.' His 2 whips mean 2 flagella. The term 'cadaverized' is used to remember chlorophyll C (A is common to all of the plant-like phylums on this test, so there was no need to incorporate this extra info in the story). The mention of 'zoo' relates to zooxanthellae, and the 'glowing' to bioluminescence. 'Blood' refers to red tides. 'Fiesty' to pfiesteria, and so on.

Obviously there is a language that needs to be applied here, and it is nonsensical. The crazier it is, the easier it is to remember. I of course need to know something about the topic, and be able to induce 'zooxanthellae' from 'zoo', but that's just vocabulary you will already have from reading. There is no silver bullet to memorizing, but different techniques can be applied at different places, depending on what makes sense (or nonsense) to you.

This is just one example, I also used much more complicated storylines that incorporated many pages of notes to remember. In the end the stories were much shorter than my actual notes, so there is a condensing effect. The stories were so wild and were easy to remember the imagery. Things like Smurfette from The Smurfs swimming in slime (representing a fungus-like slime mold organism). On the test, when I saw keywords in the question, they reminded me of the different parts in my story, and from there I was able to get the answer. It takes alot of work though, but I find coming up with the wierd stories kinda fun. I will need to work at it more to sharpen my storymaking skills.

So I'm browsing through Discover magazine and notice an advertisement for Juvenon, supposedly an anti-aging drug. I decided to do some more reading into it. I've been reading alot about cognitive enhancement drugs, and also trying some, and a common theme about these drugs is they offer some sort of anti-aging, anti-cancer, anti-epileptic, or anti-Alzheimer's disease effects. I would not be surprised if anti-aging drugs are actually cognitive enhancers as well. They are ultimately affecting cells, including neurons.

Juvenon was developed by a biochemist named Bruce Ames. Here is a highly respected person, who also created a widely used test to identify cancer-causing substances, now developing an anti-aging drug. After doing some more reading, this does not look like quack technology, and is based on sound studies. Here is an interesting quote from one of the articles linked above:

In his lab, he and his colleagues have been systematically depriving human cells of one essential nutrient after another and then looking to see what happens. What they've observed isn't pretty. When cells fall short on any one of a handful of nutrients that have been tested so far, Ames has discovered symptoms of severe genetic damage - the kinds of mutations in DNA that are believed to lead to cancer. Nutrient deficiencies also appear to disrupt the function of mitochondria - the tiny organs that provide cells with fuel. The result: premature aging of cells.

What they are essentially saying is a lack of nutrients can result in cancer. Ames attacks current diets of most Americans, stating that we do not get the proper nutrients. You are doing yourself a disservice if you do not take a multivitamin daily. I agree with that completely. You are literally what you eat. If you are not taking vitamins, go buy some. Most are water-soluble, so any 'extra' nutrients would be removed by the kidneys. It's insurance for your body.

I'm fascinated with mitochondria. These are small organelles within each cell that have an evolutionary origin of bacterial cells. It is supposed that at some point, these mitochondrial bacteria came into the animal cells and an endosymbiosis took place. They even have their own DNA and replicate on their own. These machinery supply 98% of the energy for our cells. So the familiar ancestral evolutionary tree sort of breaks down, and rather you have separate organisms fusing into one. I find that pretty amazing.

So what is this Juvenon? First of all the proceeds for selling the drug claim to support research on the drug, and there is no profit for Ames in this. That is the claim anyhow. Juvenon is a combination of Acetyl-L-Carnitine and Alpha Lipoic Acid. There has been alot of research on both of these, and there are positive effects. Essentially, carnitine is required for mitochondria function, and ALA is like a catalytic converter. Mitochondria release free radicals when functioning properly, and even more so when given carnitine. ALA acts to remove those. The goal of Juvenon is to keep these mitochondria in top condition, avoiding deterioration.

My opinion? It can't hurt to try. The fact that this was created by a well-respected biochemist gives it some credibility. There are some good interviews with Ames and he gives very good advice. I've ordered a bottle of this and will see if I notice any effects.

For a funny anti-Juvenon critique, check here.

Here's an interesting critique on how New Orleans is being handled. Warning, article tilts a bit towards the left .

If there was ever a director that made me feel like I was at the location where the film took place, it is Fernando Meirelles. I was blown away by City of God, and when I saw the previews for The Constant Gardener mentioning it was directed by the same person, I had to see it as soon as it came out.

Well I saw it tonight and am amazed. The amount of energy the film conveys is just mesmerizing. You feel like you are in the heart of Africa. This director makes alot of use of contrast, and there are some great scenes of beautiful dinner parties or golf resorts, and then the slum and chaos just a few inches away. The use of color in this film is really well done and enforces this contrast.

It's a very dark, romantic, and suspenseful film about a not-so-nice pharmaceutical industry. It's rare a film could make me cry, but this one definitely gave me some tears. It was just beautifully done and is everything a movie should be. The story is not difficult to believe, and makes you really wonder. Go see it!

One thing that you invariably see in university life is the latest fashions. It's sort of an experimental petrie dish for the wierdest and sometimes sexiest stuff. As far as the girls, I don't think I've seen one girl without a tatoo on the small of her back, and bare midriffs are the norm. Also t-shirts with quotes such as "All guys want a blonde" written on them in big letters seems to be 'in'. I must say I am having trouble keeping my concentration.

I remember seeing one guy in the library wearing about 10 watches on his arm and some wierd red and black clothing with Slipknot written all over. I thought it would be kinda funny to ask him for the time, but held back. Anyway I wondered what this band Slipknot was about. Then I heard Duality on the radio, and man this group kicks ass. How can I not love a song with a chorus of "All I've got...all I've got is insane..."?. I had to buy some tunes on Itunes. It's very hard metal, and listening to that song in the morning at 6:30am is like a shot of adrenaline for me. I blast it in my car and most people are like WTF is that noise.

I have pretty strange tastes in music. I go from really lame teenage love songs to hiphop to mind numbing metal depending on my mood. Actually, it's usually the music that determines my mood so I have to be careful what I listen to!