• There is no limit to your technical capacity. Always think about how to grow yourself technically. A lot of people, including training programs, assert that senior developers are expected to have maxed out their technical potential. But my mentor strongly disagrees. I think that he is right. If I objectively look at where I stand and, for instance, where my mentor stands, there is a large gap. I need to have more depth in diverse problem spaces.
• Never question people’s motives. This is not to say you should be naive. One can be aware of motives but not factor that in. It is much better to assume the best of people and work under that assumption.
• Realize that you are smarter than most people. Be sure of yourself and trust in your judgement. This is not an advice to be egotistic or narcissistic but rather an advice to believe in oneself. It is too easy to take opinions from other people for granted without asking the right questions. Also, when people are polarized about something, ultimately you have to decide for yourself what you think is right and also what you are comfortable with building. Interestingly enough, I have heard this advice from several people, including distinguished engineers.
• Know your limits. This may seem contrary to the previous statement. But the main point here is to realize that in the end, you can only do your best every day and no more. It is pointless and counterproductive to stress over what you could not accomplish in a given day. You also have to take into account other factors. You may get randomized or you may be mentally or physically exhausted which can affect your performance. It is important to recognize these factors and not get stressed over it.
• Compartmentalize your time. You need hours of interrupted time to be productive. Compartmentalize this time and do not answer IMs or email and just focus on your work. Unless you have this time set apart, you will stress over the fact that you could not be productive in your own work. In most cases, it is better to come into work early and block off some time to yourself.
• Prioritize your life over work. Make sure you get plenty of rest. Do the things in life that make you happy and do not let work consume your life. Recognize that your worth. If you are happy in life, you will be more creative and productive at work.

Interestingly enough, my mentor did not give any specifics on how to lead a team. I will make followup posts as I learn more about being a development lead. For reference, I previously made a post where I compiled advice on being a senior developer: http://thatdev.net/2010/09/13/senior-engineer-and-beyond/.

Harvard Professor Michael Sandel’s class “Justice” is one of the most popular at the school. “Justice” is the title of his New York Times best–selling book. He joins us to talk about how do decide the right thing to do.

Well, I managed to see the whole thing as an audience member. I’ll share my thoughts & impressions after the program.

Interestingly enough, the white background works fine in iPhone and PC clients. But for some reason Kindle for the iPad has a slightly shaded white background. This bothers me to no end.

I have two desktops at work, one at home, a HTPC, two laptops, a cell phone and an iPad. You might call it an excessive list of devices but I do make use of all of those. And it is important to be able to, say, edit a document on one machine and then pick up from where I left off on another. Dropbox enables me to do that.

Here’s a setup that I use that might be useful to you. I set the “My Dropbox” folder to some location on my desktop (that is not my C drive). Then I created folders such as Documents, Music, Photos, Videos, etc. underneath. Finally I set my Windows 7 Library folders to the corresponding folders under my dropbox folder. With this setup, all of my libraries in windows are synchronized. As soon as I copy in a file or modify another, the changes are reflected on all of my machines.

This mechanism, of course, works on any program that uses files to keep its state. As an example, I configure each of my MSN Messenger instances to store conversation history in a dropbox location so that they are all synchronized. I rely on this feature to view previous conversations (I frequently close windows).

As an added bonus, you can consume (manage and view) these files from your cell phone or iPad. Do give it a try.

The Core i5-2500K appears to be the CPU to get. It is 10-50% faster than the current Core i5-750/760 on existing applications through a 25% increase in clock speed. It also costs $216, which is pretty cheap. The K version allows over clocking at a $11 premium. But considering that you can easily get 4.4GHz with a stock cooler, I think the premium is easily worth the cost. The Core i7-2600K seems a bit expensive but may be worth it if you run VMs, etc.

I do not do a lot of video transcoding that I am aware of so I’m not that excited about Quick Sync. But apparently it’s very fast and works well.

I plan to pair a Core i5-2500K with a P67 motherboard. I already down a Geforce 460X and want to over clock the CPU so P67 is a natural choice for me.

UEFI is pretty exciting. PC users finally get to configure their hardware in a nice GUI, assuming that the motherboard manufacturers do not screw up their skinning. I need to research whether it is easily possible to dual boot Mac OSX. Also, UEFI+GPT partitions will allow using harddrives with over 3TB capacities. I currently use 1.5TB data drives that are filling up pretty quickly so this is a welcome change.

Finally 6Gbps SATA is also very exciting. I am still waiting on the next generation SSDs to replace my X25-M G2 but it’s good to know that the motherboard will be prepared for it. Lack of USB 3.0 support is a bummer but I think motherboard manufacturers will use thirdparty chips to support it.

Once products supporting the new processor and chipset are announced I will do some research and post my candidate build.

My E8400 based machine has served me very well for three years. But it is finally time to upgrade to a more modern setup.

Well, I am overjoyed to say that I can stream 1080p MKV files over my powerline network just fine. And the boxee software is slick and wonderful. I’m going to spend the next several days playing with it to decide whether I really do want to go down the Boxee Box route. I just might have to click that preorder button on Amazon…

I researched wireless extenders. However, it appears that they reduce bandwidth. I also looked into wiring up the house. But connecting one important outlet (where my HTPC would be connected to) would have required a lot of drilling. Seemingly without a good solution, I pushed off fixing my network over and over again.

But recently I decided to get powerline networking a try. As with most people, I have been very skeptical of the technology. But a recent Netgear product has been getting a lot of good reviews on both Amazon and Newegg, and I decided to give it a try.

The product I purchased is the Powerline AV+ 200 Ultra Adapter Kit. The name is quite a moutful but basically the kit comes with two power adapters. When hooked up to two different power outlets, the units basically acts as an ethernet cable that could hit up to 200Mbps (follow the link and check out the diagram).

The packaging is quite minimal, which I like. I plugged one unit next to my cable modem and wireless router and was delighted to see a green blinking light, indicating a link rate of over 80 Mbps. I connected that unit to my router with the bundled ethernet cable. Then I plugged another unit next to my HTPC and connected those two with the other ethernet cable.

I then booted up my HTPC. And sure enough, the wired connection was recognized and the transfer speeds were unlike anything I have seen on that PC. It is rare that I come across a product that is simple and works so well. Some people appear to have issues with their powerline signal strength but given what I have seen so far everything works exactly as advertised.

I am already planning to purchase Powerline AV Wireless-N Extender unit so that I can plug it into a power outlet that will maximize my Wireless-N coverage. (Currently my wireless router sits in the corner of my house.) These units aren’t exactly cheap, but I think they’re worth every penny.

I would love to see Microsoft innovate in the XBox space and then use it across product offerings where it makes sense. I heard that Microsoft does not own the technology but a lot of software can be built on top of it. The company that owns the IP is smart to bet on its potential. And Microsoft competitors will take notice. With Apple pushing hard on Facetime, I can almost bet that sooner or later Macs and iOS devices will start having extra dots to take advantage of motion based interfaces.

During that time, I worked on a feature that required a lot of thinking and fairly complicated code. One afternoon, after spending hours bouncing ideas, designing, reviewing, and coding, I checked in my progress before heading home. It was one of those check ins that gave a sense of accomplishment and made me feel good. But after dinner and taking a shower, I came back to my computer to find a reply email to my check in. And the sender was no less than the top engineer of the group.

Shivering, I opened the email. It was two pages long. On a 30 inch monitor. And the screen full of letters appeared to spell out in every detail why my code was completely garbage.

My first reaction was that I had made a mistake to join the group. After all, if the top engineer thought I was a dumb ass then what chance did I have? But then I started thinking about the integrity of my work and realized that even after having read the email I stood by it. So I gathered myself and started writing a reply to every single point made in the review mail.

Soon after I sent the email, probably after  just long enough for the sender to have read through my reply, I got another response. It read, “Okay, that makes sense.” But I did have to go back on his feedback and make quite a few improvements.

We’ve all come a long way since then. And now our group is large enough that I do not get to work with that top engineer directly as much, which is my loss. In any case, I like to think that I have earned his trust to do good work. And now I remember my little “war story” with fondness and share it with people to remind myself never to get too comfortable. But it occurred to me that I never reflected on the incident with that top engineer. So I recently did, and he seemed to remember it well too. Then he told me a war story of his own.

When he was younger, in the early 90s, he got to work with some multi-threading code. I wouldn’t know first hand, but he said a lot of the multi-threading support in Windows that we now take for granted were being worked on at the time so he had to work with some super smart engineers that designed them. I actually recognized one name.

In any case, unfamiliar with what he was doing, he ran into trouble. He couldn’t debug the problem on his own and so he called up the engineer whose name I recognized. They hooked up a debugger and the first thing they did was to look at how many threads were active. Once the number popped onto the screen there was dead silence on the phone for a while. And then came the reply.

“JESUS! Hey, come take a look at what this asshole did.”

He says it was hard to take that from an engineer that he deeply respected, but it is something that he still remembers to this day very clearly. I certainly believe him, given his vivid account. And he says the experience has made him a better engineer.

It does’t matter what field you work on. You should count yourself lucky to be able to work with someone who is scary good and can point out where you can improve on. If you find yourself really comfortable, then perhaps you should find ways to challenge yourself unless you want to be stuck on where you are.

I was going through rendered images when I discovered that on a particular image, the edge of the shadows were a lot shaper than I expected. Given that I use rather large area light sources in all my images this was surprising.

Please take a look at the following comparison. These are actually poor sample images because you can barely notice the difference and the light intensity and number of samples are different, but if you look at the full size image carefully, the soft shadow on the left image is angled wrong. It almost looks as if there are two different light sources casting two different shadows. You might also notice that the lighting looks off.

Because the problem is with shadows, I knew I could probably reproduce it using a direct illumination shader (which is much simpler and is a subset of a photon mapping shader) to narrow the problem. And, as expected, the problem persisted.

I then suspected the problem might lie in sampling because sampling an area light source is a straight forward closed formed evaluation of the rendering equation and that is where the problem likely was. I looked at the code where I randomly select a point within a triangle with a uniform distribution and sure enough, there was a missing square root. And this was one code which I foolishly decided to manually refactor.