David J. Sokol

Hooray, Hacked!

Wed, 26 Oct 2011 10:53:02 -0400

Hooray, my tumblr with a super-easy dictionary word password was hacked. I’ve actually put on weight in the past two weeks due to beer and wedding food!

2011 Chick-fil-A Statistics

Fri, 29 Apr 2011 16:28:14 -0400

There have been 118 days since 1/1/2011. In those 118 days, I’ve visited a Chik-fil-A 44 times, spending a total of $218.63. The average visit costs $4.97. Assuming I only went once a day (which isn’t true), 37% of my days alive in 2011 have included a stop to Chick-fil-A.

Now I’ve found out they’re testing sweet potato fries; this does not bode well. (via @_mason_)

Spring Cleaning - Useless Cables/Electronics

Sun, 27 Mar 2011 14:28:00 -0400

After working with computers for 15+ years, I’ve gotten a large collection of now-useless parts I’ve been hoarding. Today I’m cleansing my storage bins. These are the things I’ll no longer be saving.

IDE Cables
PCI Soundcards/Network Adapters
Serial Cables
Parallel Port/Printer Cables
1x CD-ROM Drives
4GB Tapes
Cell-phones from 3+ years ago
Phone Cable/RJ-11 connectors
Fire-wire anything
Non-optical mice
USB Hubs
Anything related to my TI-83+
Drivers for a 10BaseT network adapters

Coincidentally, if you want any of the above, let me know before tomorrow.

Keyboards with these keys should be cleansed from the Earth.

Mon, 21 Mar 2011 23:24:20 -0400

Keyboards with these keys should be cleansed from the Earth.

The specially crafted keyboard I use for writing. (Hooray, made...

Sun, 20 Mar 2011 22:32:42 -0400

The specially crafted keyboard I use for writing. (Hooray, made in Paint.NET!)

Morning Playlist

Wed, 23 Feb 2011 12:57:08 -0500

Trying to get to work on time is a hassle. I stumble out of bed in a haze barely awake to the shower where I struggle not to fall asleep. I’ve tried multiple alarm clocks, cold showers, coffee and nothing has worked. Then I bought a alarm clock/dock for my Zune and put it in the restroom, so I can get amped up for the workday while getting ready. It is amazing. I’ve reduced my average time in from tardy 14 minutes to early 11 minutes in one moment. Though the numbers might change as the sample size grows larger. Morning playlists must contain music to get you moving. I also have the additional requirement of not being bass heavy (as I live in an apartment complex). This is my 2nd iteration of my morning playlist, which contains 30 minutes of music max. If the playlist runs out I know I’m about to be late, so the adrenaline kicks in.

Raise Up (All Cities Remix) - Petey Pablo
Bounce - MSTRKRFT
Ante Up - M.O.P.
Daddy Fat Sax - Big Boi
Who Dat - Young Jeezy
As I Come Back - Busta Rhymes
In My Car - Wiz Khalifa (might get removed soon)
Hell Of A Life - Kanye West (might get removed soon)

Performance Testing is Hard

Sun, 13 Feb 2011 19:06:00 -0500

Our customers started complaining that one of our sites was slow. We attempted to verify on our own QA setup, then again on their production setup, and we couldn’t replicate. It turned out to be a browser-dependent performance problem, and further that, the same browser would behave radically different if compatibility mode was enabled.

Since we were an intranet application, compatibility mode was automatically forced on by default. Two weeks of research, one setting flip, instantly happier customers. Magic happens.

Performance profiling used to be very hard: it could be hardware or software dependent, or possibly bandwidth; now it’s gotten just downright impossible due to the wide disparities of JavaScript execution speed in the different browsers. As if making it render the same wasn’t hard enough.

Making Things Faster Tips

Any 3rd party framework generates a mass of JavaScript to do cool things. This JavaScript will run slow in IE7. It will continue to run slow in IE8 unless you enable standards mode, which is disabled by default for intranet applications.
Chrome and IE use XmlHttpRequests differently; IE seems to load in parallel, Chrome does not. This impacts debugging performance.
ViewState isn’t a huge bitch on intranet applications. But it kills mobile, especially on upload. Use the new ASP.NET 4.0 features to disable that crap if you don’t need it.
Look up Cache Busting. It helps amazingly to test cold-startup of web applications.
Cold starts can be very rare. Don’t optimize the initial startup time if the application restarts very infrequently.
Always check your logs to see what the actual load is. Just because a customer says it’s slow doesn’t mean the sites being hammered. LogParser is the best tool for handling IIS logs, hands down.

Super Meat Boy

Tue, 01 Feb 2011 22:39:00 -0500

For Xbox Live Arcade, Steam, and probably other mediums. This is, by far, the most frustrating game I’ve ever kept playing. Each level bumps up my blood pressure a few points. Every chapter greases my controller more than a large Pepperoni Lovers Stuffed Crust pizza. Boss fights provide Ninja Gaiden II level rage, the only game that ever caused me to throw a controller.

I can’t stop playing it. The feeling of clearing a level you spent the last 1.5 hours on in 35 seconds (true story) is just too good. I can only take small doses of it now, after an all-day play-a-thon Sunday, but every ten minutes of game time is awesome.

A++, would rage this way again. (Thanks to @imjameshall/@_mason_ for the recommendation.)

Stats Roundup 2010

Mon, 10 Jan 2011 22:04:00 -0500

Well, 2010 is over. I started with a spreadsheet with one row, now I’ve got 366. (One row for headers.) Lets break it down, hammer style.

Travel

I traveled via plane 9 of 12 months this year. I visited the following cities:

4x Bermuda
2x Buffalo
1x Denver
1x San Diego
1x Savannah
1x Athens
1x Orlando

This also marked my first time traveling out of the country, to Bermuda. I spent about 3 weeks working there this year, and it’s really expensive. I cannot recommend it to anyone, unless you’re traveling there on a cruise.

Drinking

Drinking was at an all time high. I accomplished my goal of drinking for 365 days straight on August 10th, 2010; this taught me that I have stupid goals. I drank a total of 12,494 ounces of beer. This corresponds to about 1,040 12oz beers or 2.9 beers per day. On the liquor (or quicker) side of things, I had 1,370 ounces of mixed drinks and 118 shots. My liver has started up an online petition against me, but it won’t matter because online petitions don’t change shit. There are some drinking moment’s I’d like to highlight:

ComicCon 2010 - I actually lost track of my alcohol intake during ComicCon and for a few weeks after. When you drink that much, you forget how to make spreadsheets.
Mellow Mushroom (Emory) - We go for Trivia every week and I average between 1 to 3 pitchers. Dirt cheap pitchers will do that.
Cinco De Mayo - 7 shots of Tequila on top of 1 pitcher of Bud Light. That’s how we do.

It should also be noted that the figures quoted are not exact. Drinking in excess blurs the memory to the point where recording anything other than “lots” would be pointless. There were 9 or 10 such days last year.

Health

My number of yearly pushups incrased 27%, from 2,643 to 3,374. Everything else went down. A measly 1,800 crunches, 30.3 miles run/walked, and 33 p90x workouts. This probably explains why my average weight went from 185 to 192.

Professionalism/Code

I was later than usual, from 6 minutes late to 10 minutes late. I directly attribute this to my shrinking job satisfaction, and you can see a significant increase in year-end tardiness. I’ve been writing more custom code, with 627 github commits on my primary project and about 50 commits on other projects. I used four different source code control systems in 2010: VSS, TFS2008, TFS2010, and Git. Git is by far the winner.

Next Year

One goal: Create. We’ll see how that manifests itself.

Science is a verb now!

Sat, 23 Jan 2010 16:25:12 -0500

Science is a verb now!

Year End Statistics Roundup

Wed, 30 Dec 2009 14:49:00 -0500

This year was pretty awesome. It was my first full year as a real employee, I started contributing to my 401k, and for a brief period, I had my credit cards entirely paid off. I had a girlfriend, I started taking better care of myself, and I now have a semi-decorated apartment. Downsides: my car drastically declined in value, pushing my net worth down 15% of year end of 2008. Lets take a look at the numbers:

2009 (Numerically)

I performed 2,643 pushups, an infinite increase over the previous year.
I finished 7,160 crunches, another near-infinite increase over the previous year.
I completed 101 games of bowling; I raised my average from a 120 something to a 160-something, a 33% increase.
I was an average of 6 minutes late for work. Sadly this totals for 23 hours of missed morning work. Good thing I stay late. My goal for next year is to reduce this number to a 3-minutes-late average.
I weigh an average 184 pounds in the morning, 186 in the evening. I apparently carry about 2 pounds of easily removed water weight.
I drank a total of 346 days out of 365. In other terms, I spent 94.87% of my days last year imbibing at least one alcoholic beverage. Increased contact with family could not be mathematically correlated due to measuring techniques.
I lived in 3 different apartments in two different cities.
I had a real girlfriend for 5 months, a length that hasn’t been equaled since 2005.

Overall, I think I did pretty well. For this year, I would give myself an awesomeness of:

73.0551458% Average (C)

This is score is composed of:

Drinking – 40%
Financial Security – 20%
Fitness – 20%
Having a Girlfriend – 20%
Fudge Factor – 10% (I rated myself a 2 out of a possible 10)

Plans for 2010

These statistics, of course, only serve as a baseline. For the next year, I have different goals and different aspirations, and my awesomeness will be computed based on the following factors:

Drinking – 16.67%
Awesomeness - 33% (variable factor, 10 point self-assessment)
Girlfriend – 16.67%
Fitness – 16.67%
Professionalism – 16.67%

Using these weight factors, I hope to bring my awesomeness to around 80-85%. I’m not going to aim for the top just yet, I just want to be better than average.

Happy New Year.

I Crack Me Up

Fri, 11 Dec 2009 13:49:27 -0500

Try
  frm.Tag = lt
  frm.RefreshCharts(lt)
Catch
  ' do nothing; this operation was just a gesture of kindness anyways.
End Try

Concerning the Lexus LTA

Wed, 11 Nov 2009 10:57:44 -0500

me: i want that car
me: real bad
James: amazing shit
me: only $350k
James: thats all!
me: ain't no thang
me: i'll ring up the homies and get them to put in extra shifts on my corners
me: gotta expand that trap son
James: oh i hear ya
James: i expect my dudes to get that for me for christmas... they saw my tweet, they know what that means.. time to hustle harder
me: time to hustle harder or you're gonna hafta regulate on `em
me: just one of those things you ain't gotta say, ya feel me?
James: oh i know
James: game recognize game, and you looking mighty mirror image right now

"I Drive a Dodge Stratus!"

Mon, 02 Nov 2009 14:47:52 -0500

I love my car. The engine warning labels are written in Japanese. It is incredibly well engineered. It gets good miles per gallon. It has a keyless RFID system. It can determine its own location globally to within ten feet. It has power windows and can get radio signals from fucking space.

Daylight savings time, however, escapes its grasp.

Twitter Lists: Ready for Abuse

Thu, 29 Oct 2009 16:19:23 -0400

me: oh man, twitter lists.
me: http://twitter.com/dsokol/lames
James: doesnt work for me
me: awww
me: man
me: get on that
James: IM trying
me: (ps it's a list of just you, named 'lames')
me: lolololol
James: figured that
me: man
me: that's gonna be great
me: cause you can arbitarily name lists and add people
me: i could have a list named 'Dick Lickers' and BAM
me: everybody goes in

Cooking

Tue, 27 Oct 2009 00:11:00 -0400

Upsides of Cooking Sausage & Maceroni and Cheese with Same Lid:

Awesomness
Maceroni Tastes like Sausage
Feeling good because you’re saving water and resources like those goddamn hippies told you to.

Downsides:

None

This Showcases My Conversation Skills

Mon, 26 Oct 2009 13:36:00 -0400

me: if you're that bored, i've got an article on software estimation that needs writing
James: oh word!
James: yeah lemme get to that
me: yeah
me: formula is
me: Estimate = Min x -Max + Min + ScaleFactor x Max
me: ScaleFactor being between 0 and infinity (but probably around 0.7)
me: basically centers around this distribution:
me: http://en.wikipedia.org/wiki/Uniform_distribution_(continuous)
me: which has unique properties that allow the inverse CDF input to be greater then 1!
James: i was gonna blog about good rap thats out
me: that'd probably be better then my boring shit
me: at least someone would read that.
me: "oh man numbers THBBT"
me: (but my technique allows you to easily use stratified latin hypercube sampling for determining the best estimation time!)
me: (or straight monte carlo if you want to run more trials!)
me: yeah
me: you can see why i've been just ITCHING to write about it so hardcore
James: for sure
James: totally passionate about it
me: it's gonna be extra boring
me: yeah
me: because i'm going to trash joel's theory of software estimation and debunk betaPERT as ineffectual as it doesn't handle small increments well
me: not to mention i'ts completely biased against the 'most likely' which is probably wrong.
me: i have the most boring life ever.
James: i stopped reading what you typed a while ago
James: lemme know when you're back on that rap talk

An Excellent Rebuttal of Paul Graham

Fri, 16 Oct 2009 00:07:00 -0400

I saw this a few years ago, but I’m re-reading it because it’s good. Also Paul Graham is kinda meh. It was written in response to Hackers and Painters, which was decent at best. A sample, from dabblers and blowhards:

The whole genre reminds me of the the wooly business books one comes across at airports (“Management secrets of Gengis Khan”, the “Lexus and the Olive Tree”) that milk a bad analogy for two hundred pages to arrive at the conclusion that people just like the author are pretty great.

Lines like that make a great post.

Fixing Rocketfish Bluetooth Keyboard on Vista/Win7 x64

Sun, 11 Oct 2009 17:26:28 -0400

When I originally upgraded to Vista and Windows 7, I could not get my Rocketfish Bluetooth keyboard to connect. This was incredibly annoying as the keyboard decently expensive (~$100) and pretty nice. I ended up reverting back to the crummiest cheapest $3 keyboard I could buy at Microcenter, and it sucked. Eventually, I grew sick of it and spent today figuring out how to make my Rocketfish keyboard work. After going through the usual bluetooth device connection wizard and making it hang, I ran across this thread on technet. Buried past all of the double-posts from MSFT employees was this gem, by KAMOTEQ:

When you go to add a device, right click on the keyboard icon and click on properties. in the services tab check the box that says Drivers for keyboard, mice, etc (HID). The keyboard will install automatically.

This fixed up my keyboard super-fast and everything is wonderful. I just wanted to repost it because it’s hard to find (buried in a Technet thread) and I know other people will be looking for the fix.

My Own Personal λ

Thu, 08 Oct 2009 23:13:28 -0400

I updated my main website, dsokol.com with an approximate update frequency. This λ value is the exacted number of updates per day, sampled from the last one year period. Given that my blog has had eight or so updates, over the past 365 days, my future frequency is calculated to be 0.02192. (λ = 8 / 365)

This is not an accurate prediction of frequency, as data for the past 11 of 12 months is missing. In reality, it should be sampled as (8 x 12) / 365, which gives λ = 0.2630 . This is a more accurate model for my blog posting frequency. The same goes for my github frequency, which has only been gathered for the last week or so. My twitter, reddit, stack overflow, zune profile and resume are all accurate, though the reddit number is a lot of guess work.

These λ values should be supplemented with a confidence figure; I’m more confident that my zune and twitter frequencies are around the given λ then my blog or github status. If I were to build a model to simulate my next year of internet activity (which I plan to do, after another month of data is gathered), the lack of background data for these two data series will have to be accounted for.

Why did I even post frequency? I believe it shows certain parts of my character you wouldn’t otherwise see. The fact that my resume is updated very infrequently (λ = 0.00274) shows that I’m pretty happy with my job and not looking. I listen to a lot of music. I’m a pretty steady poster on twitter (unless i spammed certain days, a weakness in my analysis), and I’m decently active in a few communities specific to programming. More importantly, it shows you where to find the most up to date me, which is what the end goal.