HNHacker News
TopNewBestAskShowJobs

dkokelley

3,934 karma · joined January 28, 2008

My name is Danny.

How to contact me:

danny@[remove]dannyok.com

@dkokelley

submissionscomments
dkokelley··on The Dentist Office Software Story
In order to get a return, someone with lots of money (or many people with some money) must believe that there is a future in the company. Yes, technically a VC cares most about the point between initial investment and liquidity event, but in a rational market (HA!) the liquidity event is most valuable when the company has a solid foundation and growth path.
dkokelley··on Ask HN: Is Your Income above Your Country's Median Income?
I am right at the median household income for the U.S. Consider though that I'm measuring my individual income against the median household income, which puts me at above average for individual income. I'm also young (26), and finishing up grad school, so I expect my individual income to rise significantly over the next 5-10 years.

There are lots to consider when trying to gauge your income to the median. Median for your city or state could be very different than your country's median income. You also have to take individual, family, and household income into consideration. There are many other factors, too. Age, occupation/industry, and educational background are all heavily correlated to income.

dkokelley··on Finding the World's Economic Center of Gravity
Agreed to a point. The "middle of the map" theory only applies if you assume economic activity is relatively evenly spaced (which it mostly is). Unfortunately this representation won't tell us much about where economic activity really is.

There are better ways to represent the information presented. A 3D map would put economic activity somewhere in the middle of the earth, but you would eliminate the "starting center" bias.

dkokelley··on Ask HN: 16-hour work week jobs?
A: Was it an even trade of days off for pay?

B: What is your unpaid time off policy? Could your employees take 4 weeks off and receive their regular 2 weeks vacation pay? (Specifically, can they do this upon request. I understand that your engineer is doing this, but I don't know if that's available to everyone as a rule.)

If I was given the option of making 100% pay with a relaxed unpaid time off policy, or making 80% pay with a 20% PTO policy, I would choose the former option. That way I can decide exactly how much vacation/unpaid time as I want.

dkokelley··on Why Budget Airlines Could Soon Charge You to Use the Bathroom
> By my calculations you lose money if approximately 1 ticket holder in 1000 doesn't return their $250 iPad intact

Why not install the iPads in the seat backs? It doesn't have to look any different than another IFE, so long as it saves on weight. I'm not sure that it would be worth the cost of retrofitting, but for a new plane it would definitely be worth looking into, provided that the upfront investment is comparable.

dkokelley··on FAA bans the use of drones to deliver packages
I think the implication is that a military drone is flying over the enemy. But taking a more realistic approach, a drone failure/crash in any current military theater is unlikely to hit anybody. We're not doing much flying over heavily populated/urbanized areas as far as I'm aware.
dkokelley··on Starbucks to Provide Free College Education to Thousands of Workers
I don't think that's the connotation of the OP's sentence.
dkokelley··on Uber protest gridlocks London
I upvoted you, but I disagree with your conclusion. Certain jobs get made redundant all of the time. Your argument is based on the presumption that taxi drivers are as a whole unable to do any other work, so they must protect their source of income or be provided with a basic income if/when their need is obsoleted.

In America we have a form of basic income through social security, welfare, unemployment compensation, food stamps, etc. I won't go so far as to say these programs are near perfect, but their intent is correct. These programs are designed to A) provide temporary assistance for individuals who are unable to support themselves due to work changes (i.e. layoffs due to redundancy), and B) to provide long-term support for individuals who are permanently unable to care for themselves (i.e. work-related injury/disability).

The basic income as you've described it presumes taxi drivers to fall under class 'B'. I would put the drivers under class 'A'. They will need support if they are made redundant, but they can, should, and will retrain eventually and find new employment.

dkokelley··on Uber protest gridlocks London
> Why on earth wouldn't normal taxi regulations apply to Uber? Surely it should be counted as essentially a cab company with a fancy app, and this would solve all of these issues?

In many areas, taxi supply is regulated. The fact is Uber can't be like a taxi because the laws restrict or prohibit more taxis. The analogy Uber would prefer you use is that they are a private car service, like a toned down limo with an app.

dkokelley··on TrueCrypt suggesting migration to BitLocker?
Not quite, though. The page says "The development of TrueCrypt was ended in 5/2014 after Microsoft terminated support of Windows XP." We can infer that the two are connected, but it would be equally valid to say "The development of TrueCrypt was ended in 5/2014 after Snowden interviewed with NBC."

The reason I make this distinction is because continuing from a cautious/paranoid perspective, the DMS might not say "WARNING! Dead Man's Switch Activated! If you are reading this, I may have been compromised, and am no longer available to maintain TrueCrypt." It's possible that the landing page simply references a relatively innocuous event in the cyber security world to plausibly discontinue the software. The best evidence I have for this is the fact that TrueCrypt didn't shut down precisely when XP support was dropped. (In fact, according to http://www.microsoft.com/en-us/windows/enterprise/end-of-sup... official support ended in April, not May like the landing page states.)

dkokelley··on TrueCrypt suggesting migration to BitLocker?
Is it possible that this is the result of a "dead man's switch" (DMS) set by the developer(s)? Perhaps a (continually updated) process was set up so that TrueCrypt would shut itself down if the developer were unable to prove he or she was still actively maintaining the software.

I can see a couple of scenarios where this would be wise:

A) The developer passes away, leaving nobody else to maintain TrueCrypt. Zero-day 1234 is discovered which compromises TrueCrypt. The DMS activates, depreciating the software and advising users to migrate to another alternative (why BitLocker, I have no idea).

B) The developer(s) is(are) coerced into compromising TrueCrypt in some way. As a part of the coercion, the developer(s) is(are) unable to demonstrate proof of life to the DMS, so the system nukes itself.

dkokelley··on ‘Solar’ jet fuel made out of air
Where are you finding that 73B gallons/year figure? That's more than 4x the number I found at http://www.transtats.bts.gov/fuel.asp
dkokelley··on ‘Solar’ jet fuel made out of air
That might only be an upper limit on the economic potential. After all, buyers need to be incentivized to switch suppliers, and the fuel needs to be transported to where it is needed.

I used Nevada as an example because it's a low population density, highly undeveloped, sunlight-rich state, which makes it ideal for solar production plants. I'm sure other production facilities will spring up in the California desert and similarly situated regions around the world.

What excites me most about solar energy plants is that it's ideal for our empty desert regions - Lots of sunlight, lots of undeveloped land, low population.

Edit: The $4 billion figure seems off. If global demand is 60 billion liters, that's $45.6 billion at $0.76/liter. Am I missing something?

dkokelley··on ‘Solar’ jet fuel made out of air
Isn't the point that this process can be completed with the components in the air and the energy from the sun? At ~15% efficiency we're losing 85% of the potential energy from the sun, although we were previously capturing 0% of it. It should take an insignificant amount of input energy to begin the process, and then be a net energy positive for us.
dkokelley··on ‘Solar’ jet fuel made out of air
How do you mean?
dkokelley··on ‘Solar’ jet fuel made out of air
For comparison, global jet fuel consumption for major airlines in 2013 was over 60 billion liters. (http://www.transtats.bts.gov/fuel.asp)

My envelope calculations show that we would need 8,334 km^2 (roughly 3% of Nevada) of plant space at 20k liters/day to roughly cover global jet fuel consumption. (I welcome sanity checks on this math.) That's actually an encouraging and apparently achievable number.

dkokelley··on Divide Your Rent Fairly
The way we did it was to evenly split the common spaces among the renters, while accounting for room size.

individual_rent = total_rent * (room_size + common_share)/house_size)

Utilities are split evenly among roommates.

It works, although I probably would have prefered to use something like the system here were I to do it over. Things like "has a master bathroom so you don't have to share with everyone else" tend to be worth more per square foot than regular bedroom space.

dkokelley··on Urine in Portland reservoir: How dangerous is pee in drinking water?
That was my thought as well. I'm not sure what the likelihood of a disease contaminating the entire resivoir is, but regardless wouldn't the water be sterilized before it hits the tap? After all it's exposed to other wildlife.
dkokelley··on Google Revenue Jumps, But Misses Forecasts
The forecast is not made in a vacuum. In fact, Google's management has a lot of input into the process through guidance. I understand that there is no explicit agreement from Google to hit the forecasted numbers, but from another perspective, the shareholders have put out goals, and Google missed them.

It's analogous to a salesman failing to meet his or her sales goals. You don't (typically) say that the sales manager missed the sales estimate. On the other hand, if the sales manager (or in this case the analyst and shareholders) DOES give Google an unreasonable goal, you have the option to bet against the poor analysis.

dkokelley··on Rails 4.1.0 released
I'm not 100% sure, but I believe you can upgrade without a lot of rewriting by using the protected_attributes gem (https://github.com/rails/protected_attributes).
dkokelley··on Virtual Reality is going to change the world
Nitpick: If EA's revenues are $3.7B, their value is much more than that.

Generally you could look at market cap, or derive the present value of future expected cash flows (meaning how much would you have to invest today in order to receive $3.7B in interest payments annually).

dkokelley··on Who Needs a Boss?
It's important to make the distinction between a boss-owner and a boss-manager. There are many employee-owned businesses that still utilize some leadership structure. The difference is that the leadership structure has been decided on by the employee-owners.
dkokelley··on Box files for $250M IPO on full-year revenue of $124M, net loss of $168M
From Ballmer's Wikipedia entry:

"In 1973, he attended college prep and engineering classes at Lawrence Technological University and graduated from Detroit Country Day School, a private college preparatory school in Beverly Hills, Michigan, with a perfect score of 800 on the mathematical section of the SAT... In 1977, he graduated magna cum laude from Harvard University with an A.B. in applied mathematics and economics... He scored highly in the prestigious William Lowell Putnam Mathematical Competition, an exam sponsored by the Mathematical Association of America, scoring higher than Bill Gates."

I'm not sure if he coded, but Ballmer definitely has a mathematical/technical background. I'm not sure technical background is necessarily an indicator of success as an executive.

dkokelley··on Box files for $250M IPO on full-year revenue of $124M, net loss of $168M
I'm curious, is that really as big of a deal as the tech startup makes it out to be? When the team is young, small, and scrappy, I understand that technical skills on the founding team are highly valued, but after growing to nearly 1,000 employees surely the core leadership can hand off the technical day-to-day and platform architecture to a technical SVP.
dkokelley··on A software engineer's website
Tab completion works for me.

$ls pro[tab]

==

$ls projects/

Google Chrome on Windows 7

dkokelley··on Thieves Break Into Cars Using Mysterious 'Black Box'
I believe these thieves are looking for valuables left in the car, as opposed to the vehicle itself. The "black box" unlocks the doors, but doesn't necessarily start the car.
dkokelley··on Mt. Gox Files for Bankruptcy Protection
Emotions run high when something of this magnitude happens. Judgement gets clouded. A reminder is very important as a way to intercept those without a clear head before they do something rash.
dkokelley··on The Firing Squad Problem
I think the purpose is to solve the scenario where each node has a perfect internal clock, but no external reference clock. A rough solution is to initiate a count from the origin, starting from 0. Each node hears the previous neighbor's count, increments it by 1 and sounds off. When the end node is reached, it knows how far the signal propagated. At this point a reverse count is initiated, terminating in 0 at the origin. Each node along the way back hears the adjacent node begin its countdown, and starts its own countdown from its relative position. The return signal reaches the origin node just as the remainder of the nodes reach 0, and all node execute simultaneously on 0 (with appropriate modification for any fencepost problems).

No external reference clock is required, and assuming each node counts at exactly the same rate, all nodes act in unison.

dkokelley··on Ask HN: What's the worst you've ever screwed up at work?
I hadn't heard about this so I looked it up. Chilling.

http://en.wikipedia.org/wiki/Therac-25

dkokelley··on How to validate your SaaS idea with the power of hustle
I'm curious how the "Here's a problem we think you have. We plan to solve it with software." approach compares to the "What problems do you have? Would $software_solution solve it?" approach.

This is a roundabout way to ask which product development strategy is better:

- Identifying a potential problem/solution match and then validating it via conversation

- Choosing a potential client/market and discovering a problem/solution match via conversation

The reason I pose this is because my partner and I are currently attempting the second method. We have interviews lined up with potential SaaS clients, and we are going in with the mindset of "What are your pains? What could we build to remove this pain?". Now I am wondering if it is better to go to prospects with a solution in mind and get more focused feedback.

Does anyone who has gone through the SaaS product/market development process care to weigh in?

← PreviousPage 6 of 34Next →