February 2007 Archives
« January 2007 | Main | March 2007 »
February 28, 2007
TV Sets Your Internal Clock
That is, when you watch normal television, you'll find yourself knowing what time it is more frequently than if you are watching other stuff. For instance, if you're just watching HD-DVDs, streaming media, downloaded media, and other such stuff you'll just finish watching stuff and then wonder what time it is.
If you watch TV, though, you're constantly reminded of what time it is. For instance, just about everyone who watches TV knows that if you put Leno or Letterman on it's around 11:30pm and the news has just ended.
You even get to know the day of the week better because of what's on and the constant reminder with the stations own ads.
Why is this?
Likely, it's some simple form of conditioning. Through repetition, you learn to know what time shows are on. You will also learn the day of the week shows are on. After long enough, you'll get to be able to just know what time it is based on what's on TV. Many people live in households where the TV is just always on. It's turned on when people get home. It's used as background to whatever is going on, including making dinner, eating dinner, doing homework and housework, and pretty much throughout the whole evening and, in some households, the whole morning.
The household I grew up with was somewhat like this. In fact, I was someone like this until just over 6 years ago when Laurie and I decided to drop cable and not have any sort of paid TV. We didn't have any TV reception, either, until about a month ago when we discovered we had digital TV reception in our area. Since we switched to using Netflix to watch shows, we weren't totally out of the loop. However, the times of the shows didn't matter.
For Heroes, though, we used the net to catch up (sort of) and now we've watched it a couple of times on the air. Of course, if you watch it, you're already well aware of the fact that we've watched it on a Monday evening (9pm, we're not central time... heh).
A strange thing that's happened to me with the TV, though. I've gotten to staying up later. Sure, I'm blogging more in front of it, but it also tends to keep me up. We haven't been following Studio 60, but it's on and it has a couple of great actors from the West Wing and, of course, Friends. After that, I'll probably watch most of the news because occasionally there is something interesting. After that, how can I possibly turn off Leno?
"It's a sickness!" exclaims Laurie. It's addicting, too. I agree, though.
February 27, 2007
How Fast is 802.11n?
Just how fast is it supposed to be? If you look 802.11n up on Wikipedia, you'll see that it's listed as a 540 megabit standard, a flat 10 times the speed of 802.11g. That's great and all, but that also requires a gigabit Ethernet connection on the other side. So why, then, do many of the current routers have 10/100 switches?
This Linksys 802.11n router claims 12x the speed of 802.11g, but only has a 100Mbps switch yielding less than 2x the 54Mbps of 802.11g. This Netgear 802.11n router has a 1000Mbps switch, but it claims only 300Mbps for the 802.11n port, making the wireless part slower yet the actual throughput to wired network devices much faster. This D-Link 802.11n router claims 12x the rate of 802.11g, like the Linksys, yet also only has a 100Mbps switch on it. They have a newer 802.11n router, though, that does have a Gigabit switch on it.
The story is not all that different on the adapter side. Netgear and Belkin both have Cardbus adapters that only do 300Mbps, although I don't know the maximum transfer rate of Cardbus. This Linksys adapter claims both 12x speed and 54Mbps rates, but that's probably just a type. This other Linksys adapter claims that when 802.11n is used, speeds can go "beyond 100Mbps." This Netgear PCI adapter claims a data rate of 270 Mbps, while the Belkin PCI adapter remains at 300 Mbps.
So is it over 100Mbps, 270Mbps, 300Mbps, 540Mbps, or 648Mbps (12x 54Mbps)? That would depend on how well your Draft 802.11n gear can interoperate with each other. It should all settle out at 540 Mbps in the end. However, this will only help when either communicating with all 802.11n devices or with wired devices on a Gigabit network.
It's nice to know, though, that when it does work properly, it could technically be faster than your USB2 connection. That is, a wireless 802.11n connection to a Gigabit network connected to a fast Gigabit network storage device with good cabling and good signal strength could have a chance of outperforming a 480Mbps USB2 connection to a hard drive.
For now, I'd definitely stick with all of the same brand and make sure you get one with a gigabit switch to connect to current and future gigabit devices. Just don't expect it to make your Internet connection any faster. ;)
February 26, 2007
Thoughts on Amazon EC2
So, I finally got a chance to play with Amazon EC2, or Amazon's Elastic Computing Cloud. The experience largely went how I thought it would. However, there were a couple of interesting things that I noticed. These things revolved around security, the virtual machine environment, and the billing of usage.
First off, EC2 has a firewall around it. This is controlled via an API that allows you to authorize or revoke privileges based on many different rules all combined in to named rule sets. The rules are applied immediately, too, which is great. Compared to many normal data centers where the firewall might not even be accessible to you for your dedicated servers, this is particularly useful. Not only can you defined traffic from IP and IP ranges to specific ports but you can also defined traffic by security group and user. So, for instance, you could have a backend machine in the cloud that an only be accessed by a web server group you defined. Or, instead of that, you can just specific your username and all instances running as your username will have access to that machine. This is fast, flexible, powerful, and very easy to maintain. For higher security, other software firewalls can be put in place and you could even dedicate an entire instance to being a firewall, should you choose to.
Second, the environment that the OS is running in is basically as described. However, it's listed as being equivalent to a 1.7Ghz processor, but the procinfo shows it being an AMD processor running at 2.4Ghz. I haven't done any tests on the performance of the machine, but the response times are nice and snappy. Even if the compute power isn't fully that of the processor listed, the virtual configuration is running on top of high quality components so all parts are sufficiently fast. I'd be curious how hard to would be to create FlashMob networks with a handful of instances, although it doesn't look like much has been happening within that community of years. There really doesn't seem to be anything preventing a group from signing up to use more than 20 instances and going forward within bringing up any form of high performance computing cluster for science or profit. The ability to do so, even if it's just to test the software and proof of concept, without having to dedicate hardware is wonderful.
Finally, a couple of items about billing surprised me a little. Now, this has to all be taken into context. My little experiment ended up costing me maybe a dozen cents more than I thought it would. The billing for EC2 is listed as such: 10 cents per hour per instance being on, plus 20 cents per GB transferred in our out of Amazon (but not between S3), and 15 cents per GB per month for the storage of instances on S3 (billed by S3).
The first surprise was that the 10 cents per hour used isn't pro-rated. That is, 1 minute costs 10 cents, as does 59 minutes. So, just to boot an instance will cost you 10 cents. The impact of this is that booting 60 machines for one minute would actually cost 6 bucks not 10 cents like I may have implied before. It's still cheap, but it has implications for the above mentioned performance cluster tests; that is, assume anything you want to experiment with will cost up to an hour and thus you might as well use up to an hour. This is also true for machines that are on 61 minutes; you will pay for two hours or 20 cents. Don't get me wrong, though. This is still cheap for experimentation. It just might not be the pennies you thought to bring up a machine for 10 minutes a day.
The second surprise turns out to not be a surprise. At first, I thought that S3 was billing for the transfer in and out of S3 and that EC2 wasn't, so it wasn't double-billed. However, the original assumption that no traffic between S3 and EC2 is billed is correct. See, after creating the AMI that you'll be offered to create during the "getting started" document, I saw a charge appear on my S3 activity in the bandwidth. However, I didn't do the math on it. The charge was for a penny and I thought maybe it was for the transfer. As it turns out, it was likely just a round-up from other usage (still shows 0.000GB of bandwidth). The AMI is about 227MB, which does show up in the usage detail of the transfer (one time, though), but clearly wasn't charged for. See, if it had been for transfer like I was thinking, the fee would have been 5 cents.
Now, again, keep in mind that although the billing of the instances is rounded up to the nearest hour, we're still talking pennies. Even for non-business use, which is how I'm currently experimenting with it, this is cheap. So far, I have 20 cents in instance usage, a penny for S3 storage usage, and another penny for bandwidth usage. I've learned a lot about it without even spending a quarter.
My next tasks with Amazon EC2 are to run some performance tests to try to determine what kind of compute performance is really meant by 1.7Ghz. That is, is it like a 1.7Ghz Core 2 (solo, of course) or like a 1.7Ghz Pentium 4? There's a huge difference between the two and it might actually be more like a 2.4Ghz modern AMD processor, which would also be great. Another task I want to do is see if I can turn my current server into an instance. I've been a little confused by some of the AMI creation documents without actually doing them, so I've just got to roll forward with it. That's a topic for another post, though.
Until next time, compute on!
February 23, 2007
Canon PowerShot SD900: Thoughts and Review
In this post, I'll attempt a travel review of the Canon PowerShot SD900 Digital Elph Titanium 10.0 MP pocket-sized point-and-shoot digital camera (that's a mouthful). This will read like a list of various things since it's my thoughts while using the camera during a short trip to Maui. This is not a comprehensive review, but rather a practical review from trying to use it on what might be a typical vacation.
Things I Like About It
It powers on fairly fast for a point-and-shoot (P&S) camera. It's certainly not as fast as the 30D, although people are start to think that's because it doesn't actually power off but rather goes in to a deep sleep mode so it doesn't have to boot up when the power is turned back on.
The resolution is nice at 10MP, but some early shots show that it may be a bit dirty when zoomed all the way in.
It takes video at VGA (640x480) resolution at 30fps. We've been using DSLRs exclusively for a while, so video is a bit of a novelty on it's own. The rating makes it almost as good as our regular DV video camera, but without the advantage of the DV format. However, it can also take XGA (1024x768) video, although only at 15fps.
The UI is surprisingly fast and smooth. Previewing images is very smooth when zooming in and out and going between the images. Although, based on the few images we've looked at on a laptop, I'm not convinced that it's looking at the full-sized original.
The camera has a manual mode that has a lot of settings that can be changed. It's odd, though, that to enable 3200 ISO you actually have to choose a non-manual shooting mode. That said, it's great that it has 3200 available although it's not so great that it's only 1600x1200.
It has a custom timer setting that can take between 1 and 10 images separated by 0 to 30 seconds each. This is a pretty neat feature.
I like that it comes with a normal battery charger rather than a cradle. The charger doesn't have any cables and is nice and small (smaller than the Canon 30D battery charger).
The flash worked surprisingly well both inside and outside. It seemed to have a pretty good range and didn't completely wash out people that were a few feet away to maybe 10-15 feed away. That said, it did make completely white images when taking pictures of something in our hands.
Things I Don't Like About It
It doesn't shoot RAW images, or any other lossless format. This leaves little to be done with post-processing.
It's battery life is rated for only 230 shots with the screen on, which comes out to about half of a 2GB SD card. Even with the screen off it only is rated to take 600 shots. Since it has a proprietary battery and charger, we'll have to get more of those batteries to help out. Some of our older and now gone P&S cameras took regular AA or AAA batteries, which was nice since they can be purchased pretty much anywhere in the world.
The continuous shooting mode is rather slow with the 2GB SD cards we have. I think it's just under a frame a second. It can also only be put in to continuous mode while in manual mode. If you switch out of manual mode and then back, it'll lose the setting for continuous mode. I understand why, though, since it shares the continuous shooting setting with the timer settings.
When in stitch assist mode, if the screen turns off while you're trying to line up the next shot, it'll lose where you were. Also, the manual claims these images can only be used with the software provided and that they can't be previewed. I'm not really sure what that means yet.
Summary
The Canon PowerShot SD900 Digital Elph Titanium is a great performer that targets consumer photography while providing some fun manual settings to override the automatic settings. It's lacking in normal manual settings, such as Av and Tv, will keep it out of the amateur photography market, though. Shots done in fully automatic mode always come out pretty good, though. At 10MP, the sensor helps make up for the lack of zoom range when framing images, so you can crop out a good image and still have enough pixels for it to be useful. It's small size is unobtrusive in restaurants and at parties, while being light enough to keep in your pocket all the time.
February 22, 2007
What Ever Happened to FlashMob Computing?
I remember when the event was about to happen and almost drove up to it, but recall having to work that weekend or being away or something like that. This was nearly 3 years ago, in April of 2004. The last update to the site was in May of 2004 and Google only (currently) has 17 pages for "flashmob computing." That's pretty sad, actually.
Their goal was to break in to the Top 500 supercomputer list. Three years ago, this would have been possible with a supercomputer with less than a teraflop of performance (and to think Intel demonstrated a single CPU recently that could do that). Last year, the mark was already up to over 2.7 teraflops. That's just to make it to position 500, which wouldn't last very long at all. Although more than 700 computer arrived, they never used more than 256 of them that were rolling at around 180 gigaflops. The final number was with only 150 computers and 77 gigaflops.
That was three years ago.
I downloaded the ISO that they used. Instead of burning it to disk and booting off of it, I wanted to try it within VMware as I have this notion that I might be able to use it on EC2 for a fun test. It booted fine. It only has flags for up to a P4 system and there are some autodetection issues with the Intel Core Duo I have. Aside from that, it ran fine with a single CPU and gave me a 1.35 Gigaflop result inside a VM with only 256MB of RAM. I brought up a second one, with the two on their own network. This resulted in full use of both Cores and a results of about 2.35 Gigaflops. This is on a single CPU, with 2 1.66Ghz Cores (first gen) running on a 12" laptop inside of two VMware VM sessions! Three years ago, 2x 2GHz celerons and 2x 1.8Ghz P4 systems only produced a result of 2.35 Gigaflops. (I'm actually itching a bit to try this on our QX6700 system, but that'll have to wait a bit.)
It surprises me how few people sent in results since this is so easy to run and presumably even easier if you burn the CD. Maybe 200MB was too much 3 years ago for most?
Without spending any time on it, I'm trying to make sense of what the boot CD is doing and how I can maybe convert it over to a more modern version of Linux. It's running on a pretty stock Knoppix boot disk with kernel 2.4. I don't think there is much to it other than a couple of directories.
It seems like technology like this would still be in active development. Reading the difference between this and Grid computing makes it seem like this sort of solution is much more general purpose.
The closest thing, on the Grid side, might be MacResearch's OpenMacGrid, which is a distributed computing environment with a difference. It allows any research to submit projects for acceptance on to the grid. It's unclear, though, how something less popular than SETI@Home or Folding@Home could gather enough people to have a really fast and reliably fast distributed computing environment. You also can't really have more than one such client running as they all typically run in idle cycles.
Google actually used to have something similar. Although it wasn't openly accessible, it wasn't as narrowly focused as most of the distributed computing clients. According to the Wikipedia entry about Google Toolbar, Google Compute ended near the end of 2005 and was never used for more than assisting with Folding@Home.
The Wikipedia entry on "flash mob computing" has only duplicate information from the USF site. It's as if the failure to break in to the Top 500 list killed the entire concept. I've tried contact them. If I hear anything, I'll add it to this article (and so far I haven't).
A SourceForge search pulled up Cbench which led to information on the Thunderbird Linux cluster at Sandia. This led to Open MPI, an open source project to implement a standard API for the Message Passing Interface as well as OpenFabrics, which gets even more beyond what I can understand in just a few minutes. The Thunderbird Linux cluster currently runs at 58 Teraflops with 8694 processors across half as many machines. The pictures on their site show that it is clearly an exercise in high performance networking more than anything, as the FlashMob Computing folks learned at their trial.
Sure, given enough time and experience, one could probably develop some sort of boot disk that used these tools, but I still think it might be easier to modernize the FlashMob one. ;) Sandia, after all, has the funds to do as they wish.
February 20, 2007
Sometimes the Sun Sets
This was the sunset from last night. I thought that the sunset would have been nicer tonight because there is a fire currently burning to the east of Lahaina on the hillsides. The wind blows down from the hills and you could see all of the extra haze in the air. It just wasn't quite the same as this one, though. (More information on the fire can be found at Honolulu Advertiser, but not currently at Maui News.)
Of course, that's not to say that the one tonight was bad or anything. Laurie caught some great shots of it, too. We were using our Canon SD900 point and shoot for both of these. Clearly, it's not a terrible camera. There are some things missing, of course. I'll get in to that on another post.
Well, that's the end of our short Maui trip. Lots of wonderful food, great snorkeling, a wonderful wedding, a fun hike, and some fun sunsets!
February 16, 2007
An Amusing Hotel Experience
So, we get to our hotel, right? Well, one of the first things we do at a hotel, for some unknown reason, we turn on the TV. This time around it was this dinky little TV of maybe 20" and was just a regular, old tube TV.
One time walking by another room, the door was open. In it we see a nice, decent sized wide-screen LCD TV. "Hey, how come we don't have that TV?" Then a little while later, we see a cart with many nice, widescreen LCD TVs going to upper floors. *Grumble!*
Well, while eating lunch, we got a pounding on the door. Guess what we got?
Travel Pictures: Mama's Fish House
We got to go out to Mama's Fish House for dinner. It's one of the best places we know to eat. Anywhere. We had great fish, pork, and duck. This was followed by a wonderful dessert coupled with sauternes and Kona coffee liquor.
While eating, we saw a Blue and Gold Macaw flying around right outside. When we left, it was still hanging out in the palm trees, so we got to see it up close and grab some pictures of it.
The view from our table was brilliant. You are so low down that the water feels like it's rising above you out in the distance. You can only see this through the palms on the beach, too. As dusk sets in, they light some Tiki torches (like just about everywhere here).
These are just some of the few pictures from our new Canon SD900 camera. I'm sure Laurie will be posting more and better pictures. ;)
February 15, 2007
On Our Way To Hawaii
So, I'm typing this while on the plane to Maui. We are being "forced" to go to Hawaii. The horror! Heh. ;) The first time I went to Maui was with Laurie last April (2006). Right after returning from that trip we learned the Laurie's friend Kate was going to be getting married on Maui in February of 2007. Well, it's now February of 2007.
15 Hours Ago, Valentine's Day Evening
I left work at the usual time or a little bit early. Laurie had picked me up, leaving our Jeep behind to drive up to Concord and go out for Valentine's Day. The drive was fairly light on the traffic and we got to take advantage of the carpool lanes (although there are far fewer than there should be).
We decided on the way up to try to go to Claim Jumper with my grandmother since it was near a couple of other restaurants for backup as well as a CompUSA. Well, we got up there in good time and only had to wait about an hour for a wonderfully large dinner. Claim Jumper has super-American-sized portions. I mean, the portions are large even for a typical American restaurant. The food was good, though.
Why a CompUSA? Well, we currently have two Canon 30D DSLR cameras. However, when fully kitted they can be a bit too much in many situations. So, we've been wanting a regular, small, pocket-sized point-and-shoot digital camera. What better than trying one out while going to Maui for a week?
So, we picked up the Canon PowerShot SD900. It's in a titanium case, has a 10MP sensor, a reasonably sized screen (although other cameras like the SD630 had larger screens). It still has an optical viewfinder for saving battery (from 230 pictures to 600 pictures per charge). Of course, I'll be writing more about this later. We're not yet totally convinced this is the point-and-shoot for us, but we'll see. This size and weight are great, so far.
12 Hours Ago, The Middle of the Night
We tried to get to sleep early since we had to leave the house at 5:30am. So, we got to bed at about 11:30pm and about half hour later I got in to a terrible coughing fit that juts wouldn't stop. I must have been coughing violently for an hour or so before getting back to sleep. I kept poor Laurie awake who also didn't get any sleep the night before.
7 Hours Ago, Rude Awakening
We got rudely woken up by a very loud mechanical alarm clock that wasn't anywhere near us. However, it went off at the right time. My phone went off a few minutes later, but I was already up by then. Some Starbucks, a trip to Longs for some cough stuff, and a drive to Oakland airport, we were ready to go through security.
It's always fun going through security with multiple laptops. Luckily, they only ask you to take laptops out and not cameras, phones, as other gadgets. Although the line was quite long, it went relatively quick and we had maybe a 15 minute wait at the gate before boarding.
Back to the Present
So now I'm still sitting here on my Dell XPS M1210 laptop writing this entry. I'm in a middle seat, but with Laurie next to me it's not so important. Laurie is also on her Dell Latitude D820 doing some photo editing (some exciting news recently got her even more excited about such things and her blog is now far more popular than mine).
So what other kit do we have with us? We're actually traveling fairly light. Of course, you know we have both of our laptops. These are to help with dumping pictures from the one Canon 30D DSLR that we brought and now the Canon SD900.
For entertainment, we have my Nintendo DS and Laurie's Nintendo DS Lite (yeah, she gets the better one -- we actually had that one first and then I got a regular DS for me as it was a good price). Of course, we also have our Apple iPods which is her 8GB Gen2 Nano and my 80GB Gen5.5 Video.
We also have a Bluetooth mouse with us, some extra batteries for the 30D and the mouse, in addition to some books in hardback format. Yes, those are real paper books. That's about it for the electronic gear. We have some Netflix DVDs and are hoping that the hotel connection will be good enough for some streaming Netflix if that isn't enough.
All of this stuff with built-in wireless raises an interesting question: is it really against the rules to use Bluetooth or WiFi on the plane? There are planes with WiFi for Internet access (we even got use that on our around-the-world trip). First, we were playing Club House Games with our DS's. Of course, they were connected with WiFi. Secondly, we're using the Bluetooth mouse to help out with the photo editing.
Naturally, when we were connection the Bluetooth mouse we saw a phone or two that were still on. One of them had the persons name. I thought about calling out the name, but didn't really want to advertise that we were using wireless devices, too. ;)
February 14, 2007
Netflix Streaming Video: Is It Any Good?
Finally got to watch something via Netflix's "Watch Now" on the web. We watched Congo as our first movie. Neither of us had seen it in quite a while and so we didn't remember the store in full detail.
One annoyance is that the streaming client will only run in Microsoft Internet Explorer. I have IE7 installed, so I used that. This isn't completely annoying, though, because that leaves my Firefox window available on the laptop monitor for browsing and such.
I plugged my laptop (the Dell XPS M1210) into Ethernet so that the connection would be as stable as possible. Netflix detected it for the "High" quality, which is their best quality of three levels automatically detected by the actual bandwidth available to the server. This category requires over 2.4Mbps of bandwidth, IIRC, which is nearly half of our available 6Mbps.
I also hooked up the 40" TV to my laptop using 1920x1080 resolution for a nice, native wide screen display. I did a "F11" maximize within IE7 so the top bar would float away. After clicking play on the movie, the Netflix plugin (or whatever it was) had to be installed. This installed in seconds and seamlessly. The movie started playing in a tiny little window right away (although, on checking for the above screen shot, it was 720px wide). That was impressive, so I hit the "full screen" button and got a dialog saying that Windows Media Player had to install something for a requested action to work. Clicking yes had it install something and success dialog popup mere seconds later. After this, the full screen button worked fine. A small control bar was at the bottom of the screen and also hid itself after a few seconds.
The viewing experience was great. The screen had no borders on it, unlike some of the playing experiences with the networks. It fit almost perfectly within the full 40" display, too, and played without any annoying tearing that many of the players experience. Throughout the entire movie, we never experiences any skipping of frames or other sort of troubles that might be caused by slow bandwidth to the server. A couple of times we paused it; this also had no negative impact on the playback. It paused and resumed like playing any movie might.
Picture quality was not too bad, given that they are sourced from a DVD. There were no noticeable deinterlacing issues. Most things were smoothly blurred for the 1920x1080 resolution we were playing at compared to native DVD at 720x480. Now, the non-full screen player window looked much smaller than 720x480, so I don't know if it's actually being streamed at DVD resolution anyway. (Update: It was 720 wide, but not 480 high. Instead, it was 390 high which made it a 1.85 aspect ratio screen. This means it's probably movie dependent. Click on the first screen shot to see the actual size.)
The biggest issue with the quality, though, was the color depth. Faces and skies ended up being made of only a few color bands or so. Many other things that had slowly changing colors had also being downsampled to only a few color bands. This made these things, which included many faces, look like rough topographic maps. When we were within only a few feet of the screen this was actually a very distracting issue. Jungle scenes and scenes where everything was moving were movie or just busy were actually much better looking. Sure, they weren't sharp, but they didn't have the coloring issue. When back farther from the screen, only the worst of this would show.
I hope they can figure out a way to make this better. I also don't know if it was just that movie, though. As we watch more, we'll have no choice but to keep an eye on it because it really is bad enough to notice and, like I mentioned, actually be distracted by it.
Overall, though, this does make a great personal library supplement and filler for when we don't have any Netflix DVDs or HD-DVDs on hand. They seem to be adding new movies fairly quickly, too. Maybe the day will come when we'll only be renting HD-DVDs and playing anything else back online. We'll see. ;)
February 13, 2007
Using Massive Amounts of Computational Performance
So, what is one to do with a teraflop desktop CPU? Intel recently demonstrated their 80 core CPU that can do floating point math to the tune of a trillion operations a second while using a mere 63 watts of power.
Already, there are some projects to leverage to available computational power of modern video cards, which greatly exceeds that of the general purpose CPUs. This is being done by some of the distributing computing projects, such as Folding@Home. The current line of NVIDIA GPUs enables the use of 128 stream processors for this type of processing. This is already more specific purpose cores than the 80 core demonstration by Intel, although Intel is planning on that being 80 general purpose cores. ATI was the first to be supported by Folding@Home. This led to some amazing performance gains.
Sure, all these new CPUs can be used for distributed computing where a single CPU will never keep up (by definition, since thousands will always perform better than one). There will always be scientific and research projects that could benefit from very high end calculation performance (as defined at the time it's needed -- meaning the performance level of that definition goes up every day).
Ok, so we'll need some of the new performance to process data off of our ever increasing network speeds (wireless or not). We'll always need more performance for our video games. So long as any game still chugs along on the current highest end monitor resolutions, there will be room for improvement in games.
Things like fingerprint recognition are here now. Maybe in 5 years we can use the built-in webcams (that are hopefully better than 1.3MP resolution) for retina recognition in real-time.
Microsoft Windows Vista tries to guess now what you'll do and when you'll do it. Maybe in 5 years it'll be better at that. (I assume we'll likely still be using Vista, but probably SP2 or SP3 by then.) Perhaps other applications will join in on this predictive computing to reduce average latency of answering your questions or responding to your queries by guessing what you'll do next -- and doing it (of course, it won't be applied until you actually ask for it).
This reminds me of playing chess where each person tries to predict what the other will do and protect against it before it happens. Speaking of chess, it's been nearly 10 years now since Garry Kasparov lost to Deep Blue. The web site is still very much the same as it was back in 1997, too, which is odd. It talks a lot about "last year's" machine, which was actually Deep Blue 1996. In 1997, Deep Blue could look at 200 million positions a second. That was enough to just barely beat the highest ranked human chess player, who still holds that title. According to the Wikipedia entry, Kasparov called for a rematch that was denied and Deep Blue was retired (Kasparov is also retired now). The point of this, though, is that it's been 10 years. Would there even be a possibility of a human player to beat a modern equivalent of Deep Blue anymore? Aside from the special purpose chess chips, Deep Blue "only" ran at around 11 Gigaflops, at it's best.
Modern implementations only check a few million positions a second due to improved algorithms, but a teraflop-level CPU should be able to check significantly more. At some point, a combination of truly deep brute force along with improving algorithms will make the computer chess engines nearly unbeatable.
This is one possible use of advanced computational ability. Naturally, it's in games again, although not your traditional 3D game. That said, if your 3D shooter AI dudes had the smarts of some of these chess engines, they'd be better able to train you for your first online encounter.
Outside of the scientific, research, and gaming communities, what can the average, normal computer user do with this sort of computational power? Add in the fact that bandwidth could be increased by a thousand-fold in a similar time frame, there is enormous potential in the very near future for doing some very interesting applications.
What are those? How do you even going about figuring out what those could be? Can they be simulated now? What do you want to be able to do with your computer that you can't now?
Sure, I've posed similar questions before. If we really do get CPUs that perform at those levels in only 5 years, maybe sooner if competition and the need demand it, wouldn't you like to be ready for it? Why wait for it to be here? Chess algorithms of the past didn't wait for supercomputer to be able to work well. Many were written such that as a speed of a computer goes up, the playing performance would go up because they'd be able to do more within the standard allotment of time.
It's all about thinking outside the box inside today's boxes. ;)
February 12, 2007
A New Look at Future So-Called "4G" Networks
NTT DoCoMo in Japan recently did a successful test at 5Gbps. Yes, that's five times the speed of a wired Gigabit Ethernet network. That's pretty impressive.
Now, I don't think the generation can be labeled until it's in the past and can be defined. However, that's not really the issue here. The issue here is: What can we do with 5 gigabits of data moving every second to the mobile?
I say that there is a ton of stuff that we can do with that amount of bandwidth. But really, what's that come out to? Maybe 500 megabytes per second after protocol overhead? Then you factor in the mobile reality factor, which is that current 2.4Mbps networks operate at an average of 500Kbps. So, we can bring this number further down to a mere 100 megabytes per second. That's so sad, isn't it? What, that's only about 157 times the speed of my current DSL line. Uh...
This is some amazing amount of bandwidth available. As long as the latency is low enough, this could enable some very interesting applications. It's nearly as fast as some of the fastest desktop hard drives available today. So, anything you could do from a super fast hard drive, you could enable on a handset like this without having it's own local storage. This brings data and applications from the entire web to the phone in on-demand time. That is, you get what you want when you ask for it.
Think something like Google Earth, but with no waiting time for bringing in the data while you zoom around. Imagine being able to play full HD video of any movie available without having to wait for buffering and never having it stutter. All applications available by just launching them without a need to copy them locally or install them.
Don't forget about all of things that we (as a whole) haven't thought about yet. There will be new media, new applications, and new uses that will be able to leverage this sort of bandwidth. Let's also not forget that this is just the next step up. There will be more steps as big or bigger than this. What will they enable?
Sure, thinking about that is like thinking about the uses for the 2.4Mbps networks we have now back in the early 1980s. If you were sitting on a 300 baud modem back then, you probably didn't even dream of watching movies over it. After all, why would this even be a consideration? You could rent tapes and you could see TV over cable.
Following by example, what do we take for granted now as something that we can't do on the web? This would be something that you would normally think of as taking too long to do on the web or maybe that the interface isn't right or something like that. For example, if you calculated out the size of what a movie might have been and then calculated how long it might take to transfer it over twenty years ago, you'd come up with a number on the order of 30 years. Remember, part of why we can even do it now is because of the modern compression being applied to the size. Even a highly compressed, low resolution DVD contains 2 or 3 gigabytes of video and audio. Even audio wasn't really downloaded much until it was reduced from around 15 megabytes a minute for high quality audio to only a megabyte a minute.
Here's a final thought: If your mobile is working at that rate, how fast will your wired home Internet connection be? If it's that fast, imagine how fast your local storage will be? I could go on, but I won't...
February 9, 2007
Does Perl Use OS Threads on Win32?
So, after reading through perlipc about how threads used to be only available for emulating fork on Windows, I became curious to know if Perl, specifically ActiveState's ActivePerl, had it's own thread manager or if it used OS based threads.
Even when reading about threads it sounded as if they may not be true OS level threads since they were called "interpreter threads." That sounded like something where Perl might have it's own manager. In fact, Perl might be required to have it's own manager on systems where OS level threads do not exist.
On a typical Linux system, doing a fork will usually show the process duplicated in the ps view. On Windows, however, I noticed that there was still only one Perl process running. So, I went about to find a tool to show the threads of a process.
A quick Google search (or is that just "google" these days?) turned up PrcView (or pv on the command line). So, I brought up the CPAN module at the command line (e.g. "perl -MCPAN -e shell") and checked it out. Sure enough, there was only one thread sitting in the process.
Then I brought up my little test app that does two forks. It ended up with 4 threads total in the process, which was great to see. Why?
Well, if an application doesn't use OS level threads then the OS just thinks of it as a single process. The reason this matters is because then all of the processing will only ever be scheduled on to a single CPU. This used to only matter for servers and higher end workstations where most people would be running applications designed for a multiprocessor environment. However, in the days of multicore chips and even hyperthreaded chips, almost anyone can have a system with multiple execution cores. Any application that does spread it's processing across multiple OS based threads will benefit from this. I'm not just referring to the performance of the application, either. The response time and smoothness will be improved.
Of course, since threads will truly be running at the same time, proper synchronization is that much more important. Even the Perl documentation talks about lower level calls not being re-entrant.
At a first glance of Apple iTunes, one would almost certainly peg it as not using OS based threads. After all, if you're downloading, or doing certain other tasks, the UI becomes completely non-responsive. You can alt-tab in and out of it, but you can't actually do anything with it. However, using this tool on iTunes right now, shows that it's using at least 17 threads. So, the real question then, is what the *bleep* are they doing to cause the responsiveness and performance to be so poor??
February 8, 2007
Making Your (Computer) Life Simpler: Automation
One of the things I like to do on computers is automate various tasks. Now, I'm not talking about things like disk cleanup, virus scanning, and other such mundane and boring tasks. I'm talking about slightly smarter things.
I often find myself create a lot of temp files. That is, I'll briefly download a file to use in a post or something like that. These files are rarely needed for more than just a few minutes. If I have them go to the desktop, then my desktop ultimately gets cluttered with stuff I don't need and stuff that I do need.
So, on my work machine, I created a directed called "Deleted Daily" where I can put temp stuff in. However, instead of deleting it myself every once in a while, I created a simple batch file:
rmdir /s /q "c:\Documents and Settings\Shane\Desktop\Deleted Daily\"
mkdir "c:\Documents and Settings\Shane\Desktop\Deleted Daily\"
The "/s" option is used to remove an entire tree. The "/q" option keeps it from prompting. Finally, the script recreates the directory. This batch file is then called every 4 hours throughout the day by using "Scheduled Tasks" within the control panel. The only issue is if you are using the directory right when it's deleted. This can be reduced by running it less frequently, but I like having it cleaned out so frequently I'll take care on what I put in there -- and what I don't.
I also have scripts that send me emails about various things. I'm more likely to read an email than I am to notice a popup when I'm not at the machine. This also allows it to run without interrupting my work.
Not all of these scripts have to run on a schedule, but causing them to do so helps simplify your life. They then become things you don't have to think about. Another way to have scripts run is to base them on triggers. That is, if a script runs every time you do action X, it can run more efficiently while also being more likely to run in a reliable method.
An example of this sort of script could be one that monitors for BMP files in a particular directory and automatically creates a JPEG out of the file. It's like running the script or action over the file, but with the added benefit that you can do it on files pasted or saved into there from another application.
All modern machines have various scripting options available to them for automating things from the mundane tasks to the advanced behaviors. Learning to use them can make your life simple and speed up your work.
February 7, 2007
An Amazing Clock!
The story over at this persons site about him building a clock entirely out of a rare wood from Australia is just amazing. The person had never built a clock before. It sounds like they were pretty good with wood already, but the sort of challenge building a clock presents is far beyond almost anything else you can build out of wood. The wood chosen, Western Australian Sheoak (as opposed to maleoak?), turned out to very hard.
Now, I don't know anything about clocks (well, physical ones, at least) but one of the pieces, the escapement, looks truly amazing. The building of it is also quite amazing. They heated the wood to make it harder. This removes moisture from the wood and if cooled in the right shape, it becomes harder than it was.
The claim is that it took 3,000 hours of labor over nearly three years. That's really amazing to me. You can build small airplanes from kits in that amount of time. Granted, part of the time was rebuilding an escapement that broke -- and they took 3 months a piece.
I've always been fascinated with physical engineering feats like this. Sure, to some people, the behavior of software is like magic. To me, though, these sorts of devices are like magic. Sure, all of the physics makes sense, but the level of patience, extreme levels of precision, and just shear beauty of the gears and wood pieces is really impressive and fascinating to me. It's one thing to build a precision clock out of metal and buying pieces and putting them together. It's a completely different thing to build it entirely out of wood and with all pieces self build and measured.
Thanks to the MAKE: blog for the link to this.
February 6, 2007
The Death of Phlog
I was recently looking through my links and thought I would look at my older blogs. One of them was a photo blog hosted over at phlog.net, a small site run by an individual that sprang up in the very, very early days of camera phone use. It was very friendly and had a great community going. However, it's clearly dead or dying now. So, I posted the following on the forum over there and thought I'd put it here, too, for anyone interested:
So, I moved away from phlog to my own hosted blog quite some time ago because I was a bit concerned about having my images hosted by someone else -- this applies for me to any of the big sites, not just phlog (I don't really use flickr, or others, either). I always really liked phlog because, as a service, it was very friendly and so was Alan.
Clearly, though, phlog has basically died and been overrun with spam (even the forum has).
When I migrated off of phlog, I created some Perl scripts to grab all of my images as well as the descriptions to go with them. If the site layout hasn't changed a whole lot since then, they would still work.
If anyone is interested in these scripts, please email me -- "kf6nvr" at "gmail" dot "com". I can either provide the script for download or even try to run it over an account and provide the results via a zip file.
It's a pretty sad situation this site got in to. Hopefully this won't be drowned out in the spam, too, for anyone who might find it useful. It is certainly meant as no offense to all of those who helped support phlog over time.
The story sounds like many sites and services. Company A buys Little Company. Little Company employees only work on Company A stuff. Little Company workers get tired and leave. Meanwhile, Little Company service is left to rot yet Company A won't give up Little Company services are even sell them back to the founders.
I'm certainly in no position to try to help the service itself, but I can help out people that want help trying to get their images back if they don't have the originals. At some point, the server will crash or go away and never come back up -- unless Company A actually tries to revive it, which seems unlikely at this point.
UPDATE: It seems that, as of March 10, 2007, the original creator of phlog.net, Alan Bradburne, as regained control over phlog.net. He's now moving everything off of the old server and on to the new. This is great news for the site and my offer above doesn't really apply so long as it's around. Not that I wouldn't if requested, but I'm no longer "recommending" it.
February 5, 2007
My Trusty Dell XPS M1210 Laptop
I haven't really talked much about the laptop I'm using on a daily basis and the laptop that we took around the world on our honeymoon. However, it's been a really great laptop. It's strikes a wonderful balance between performance, weight, functionality, size, and battery life.
It's in the XPS line of gaming laptops, which basically means it has the option (currently) of a NVDIDIA 7400 256MB TurboCache video card instead of the standard GMA950 of the Intel chipset. The card has 64MB of built in video RAM with the rest being shared with the system. This gives it pretty good performance, but not great. However, at 1280x800, it's not bad at all. I still have to play stuff like NWN2 at lower resolutions, but that's even true on our QX6700/7950GX2 system. With a dual core processor it's rather responsive and speedy at most things.
The weight is a bit over 4.5 pounds so it's very easy to carry around, which was especially useful when traveling around the world. The included power supply is also relatively light, especially compared to the XPS Gen2 brick. It's very light on the lap, has an intake on the bottom and exhausts on both sides. It gets warm when active but right now it's hardly spitting out any heat at all.
Mine doesn't have the built-in webcam (erf) or the EVDO card, but those are both options allow it to be very functional. It's also at the right size to still have a full DVD burner as well as basically a full size keyboard. It has a MediaDirect button that does a few different things. If the machine is on, it launches the Dell media experience application. If the machine is suspended or hibernated, it wakes it up out of hibernation and launches the application. However, if the machine is off, it starts directly into a special boot mode that is the Dell media experience but without the rest of the OS around it.
The battery life has been great, although rapidly getting worse. When we first got it, I would often get 6 hours on dim setting with web browsing over WiFi. Now I can barely get 4 hours with the same usage patterns. Sure, the machine has been through a lot and the battery is used often since it does still last for quite a while. However, I am still disappointed by this. This one has the 1.66GHz Core Duo, while new ones start with the 1.86GHz Core 2 Duo. I suspect it uses slightly less power than the new ones, but the faster ones would be nice, too. ;)
In the end, it's a great machine for blogging and browsing in front of the TV. It also connects up to the VGA input of the TV for a full 1920x1080 experience with ease. For gaming, it's adequate at the native resolution of the screen. It could certainly be much worse. I would have preferred to have the built-in webcam and Vista, but I got it long before the Vista promotions for free upgrade. But, regardless, I'm quite happy with the machine.
February 4, 2007
Age of Steam vs Sid Meier's Railroads!
Age of Steam, a board game that we played for the first time Saturday night. I really enjoyed playing it, although I felt a little bit lost. I want to get it so Laurie and I can play. Even though it lists 3-6 players, having played it we don't really see any reason why two couldn't play. Some of the turn order stuff wouldn't be nearly as competitive, nor would be board fill up as fast, but that could be solved with other little tweaks.
One thing struck me, though, throughout the game. It was very familiar to me.
The basic game involves selling shares to get money to build tracks.You then use these tracks to hopefully make more money back than you spent, or you can make a bigger train. You then collect money based on certain metrics on how you used your tracks and then you pay for expenses.
This reminds of a computer game that came out recently and that we have: Sid Meiers Railroads!
Sid Meier's Railroads!, a computer game that is basically a railroad simulator in interesting settings starting around the time railroads were just being put in within the US. This game is very fun and I enjoy it greatly.
In the game, you start with a little bit of money and a railroad company. You can start building tracks and putting trains on them. You can also sell shares in your company to have extra cash to develop your railroad company faster than the competition. Sound familiar?
All of this is, of course, in real-time. Besides just doing the build, you also need to do the routing of the track, which is made more difficult based on the game settings. Then you have to design routes over the track based on what cities produce and what other cities want. In addition to this, you also can buy various industries in the cities to make even more money when stuff is produced with them.
As I mentioned, you can sell shares to get some extra cash on hand. However, you have to be careful because one of the game winning conditions is owning all of the competing railroad companies. Yes, that means you can also buy shares, should you have extra cash on hand. You can buy the shares in either your own company, to keep other companies from buying them, or buy other companies' shares so you can try to take them over. When you do, you get all of their assets should you choose to keep them.
There are plenty of other complications within the game, but it's actually fairly easy to play. Performance is pretty good on our machine (a QX6700 with 7950GX2 video card on a 2560x1600 display) but when a lot of rails are around with a lot of trains it tends to start slowing down. (It doesn't appear to be a multithreaded app, though.) It's also crashed a fair amount, too. I could avoid this be restarting from a save that wasn't too far back, but the last time I played I couldn't recover from the crash condition. Maybe a future patch will fix the problem?
In any case, Age of Steams was very familiar because it has a lot of similar features. Sid Meier Railroad games have actually been around longer, but both are unique in their play and style (and not just because one is a computer game and one is a board game).
February 3, 2007
The Television Networks are a Pain!
So, we started watching Heroes a little while back online. About a week ago, we went to watch the 6th episode, but NBC had removed episodes 1-7. We were pissed! Today, we went back to just start at episode 8 but, alas, they had removed that one, too! So, instead we started back up again on episode 9. However, we got to episode 12 only to find out there was an error and it couldn't be played ("file not found" flashes in the player). Grr...
So, we go over to ABC to watch more Ugly Betty, which they kept proclaiming they had the full "Season to Date" online. Well, guess what? They don't! They removed all except the pilot and the last four episodes. We got lucky, though. The episodes they left happen to be the only ones we haven't seen (pilot aside).
All of this time, I was hoping that the networks would start adding more episodes and maybe even more seasons. This would allow the to monetize on back catalog and get more viewers throughout the season of people that haven't seen it at all in the season.
Oh, and I must add that the streaming quality on NBC is wretched compared to ABC. The ABC streams look absolutely wonderful and smooth. The NBC ones are horribly artifacted and not all that smooth.
Sure, all of this stuff is new to them. It seemed like they were getting it, but now that the seasons are at their halfway point, I think we're starting to see the true colors of the networks. That's really too bad, though. In the end, it doesn't seem like this is going to allow us to stay on the current seasons of all of the shows we want to. We'll probably still have to rent most of the shows on DVD later. Well, I guess that gives more money to the publishers and less to the networks. So there. ;)
February 2, 2007
Netflix: I can Watch Now!
Well, as promised, I thought I'd talk more about the Netflix Watch Now feature. It is completely active on my system. However, it's not yet quite all I'd hoped for.
First, as it turns out, they don't support Firefox for playback of the movies. Not a terrible problem since I have both browsers available, but annoying in any case.
Second, there aren't actually that many movies available yet. There are both movies and TV shows, which is great.
I haven't had a chance to try playing anything yet. They claim that the quality varies automatically based on your bandwidth. Anything over about 1.6 Mbps will get you the high quality, but that's not your link speed but rather the current bandwidth available and usable for the client.
I'm still very curious on how the TV shows work since they usually include multiple episodes. Do they just string them all together or what? I'll have to update when I find out more about this. For now, though, the selection is more for augmenting our current set of DVDs rather than watching, say, the next disk of Alias before we can get more disks back. However, we likely will be using it on weekends and other times when we run out of Netflix disk. I just hope they keep adding lots of movies, otherwise we'll exhaust the streaming collection real fast.
February 1, 2007
Netflix Streaming Movies
So, after the announcement about the streaming movies on Netflix, I went to my account hoping to find it. But in the account section, the only thing I could find was some text that stated I could do it once I see a "Watch Now" tab on the home page. Of course, this wasn't there.
Well, I was going to blog about that. However, when I went to the same section in the account to get the text to blog about how I wanted to have it, you know, right now, I clicked on the link and it added a tab and brought up "Congratulations! You now have access to our instant movie watching feature." Whoo hoo!
I'm a big advocate of Netflix because it saves us a lot of money, allows us to watch TV shows without cable or satellite, and is rather fast. We've been with them long enough to have the 4 disk version at a good price. We also now get HD-DVDs through them, at no additional charge. And now we have streaming movies. But what else? That's the question... I'll have to explore the service a bunch. However, I can't right now for obvious reasons. ;)
More later...
The Obsolescence of Incandescents
So, a friend sent me this article about the possible banning of the traditional incandescent light bulb by 2012 in California. My first thought was that it would be nearly impossible to do that. Except that I thought to think about my own house and every light in our house is already either fluorescent or LED that isn't on a dimmer switch.
And it's not because of wanting to save the environment and such. That's a nice side effect, of course, but the most immediate need was to save a few greenbacks (or whatever color they are these days). Out here in California, PG&E has relatively high electricity rates, especially for those who actua