Armchair Quarterbacking

Hi

Mon, 24 Oct 2011 14:50:01 -0400

Hi

Mon, 24 Oct 2011 14:49:36 -0400

Hello, NYC

Thu, 29 Sep 2011 23:21:14 -0400

Hello, NYC

Letting your users make your product

Thu, 25 Mar 2010 01:06:00 -0400

Services like Twitter are interesting to think about because from an engineering perspective, the goals around what users would initially accomplish are not so clear. Yet, at the end of the day, the service was flexible enough to accommodate the user scenarios that eventually manifested themselves, given the ability to broadcast updates (see this post for some interesting scenarios that Twitter enabled). I strongly believe that had Twitter been initially designed & marketed to ‘facilitate meetups between strangers’, it probably wouldn’t have succeeded.

Frequently, I hear from more experienced friends that selling to businesses vs consumers is very different because the consumer market is so fickle. So, an admittedly lofty goal might actually be to keep whatever capability you provide so vague, yet the use of that capability so simple that people can tell you what they want just by using your service. Making this even tougher is that it has to be flexible enough to NOT drive users away once they (or a competitor) realizes what they want and to NOT demand an about-face in strategy. In that case, you might end up not falling prey to how fickle the consumer market actually is.

So, the trifecta: simple, flexible, and powerful. Easy, right? :-)

Steel-cut oatmeal in a plain white bowl

Wed, 24 Mar 2010 23:52:07 -0400

Breakfast: The single most important meal of the day. Luckily, I’m onto a foodie’s delight with my steel-cut oatmeal in a plain white bowl:

Subtle yet complex flavors AND moist, chewy texture are qualities that few foods can qualify as having, yet oatmeal delivers on all with aplomb. However, before all of this delightful tastiness can be appreciated, a true oatmeal aficionado must be aware of the meaning of its constituent words:

oa - As in, whOA

tmeal - as in, The meal

Steel-cut oatmeal in a plain white bowl stands on its own; you’ll need no nuts, brown sugar, or fruit, as these are strictly unnecessary and could even discolor the plain white bowl.

Enjoy!

My friend Karen has a real food blog that, even if I read right after a huge meal, still makes me...

Tue, 23 Mar 2010 14:28:00 -0400

My friend Karen has a real food blog that, even if I read right after a huge meal, still makes me salivate with hunger. She’s that good! Out of jealousy & spite, I decided to start my own competing food blog, and I’ll start by reviewing one of my favorite drinks ever:

Kirkland Signature Premium Drinking Water (not to be confused with Kirkland Premium Drinking Water) holds a special place in my heart. This bottle has been sitting on my desk for a week now, but, believe it or not, the water has STILL retained its refreshing, cooling, invigorating taste! Although it’s missing that subtle lead flavor that I sometimes miss from the good ol’ tap, it more than makes up for it with a dynamic, full-bodied, in-your-face experience with no aftertaste.

On top of all that, it has the Costco connection, and who doesn’t love Costco?

Note: sample tastes of my Kirkland Signature Premium Drinking Water are not permitted.

Programming: Higher level languages vs. CPU Instructions

Thu, 18 Mar 2010 21:41:16 -0400

The instructions that the computer understands are very primitive: mostly arithmetic operations such as add, subtract, divide, and multiply, some operations that aren’t arithmetic operations, and others. Because the instructions understood by processors generally operate only on numbers, it’s a time-consuming and complex process to take operations we normally think of doing, and boil them down to the primitives that computers understand.

Therefore, there are entire research areas & software markets dedicated to the task of taking languages that are more palatable to people and generating efficient sets of instructions that computers can understand. These languages that are “more palatable” to people resemble spoken/written languages in the same way that Al Gore invented the internet: hardly. FYI, the products in this market are compilers, linkers, debuggers, and other names; Microsoft’s Visual Studio is a popular one for the Windows platform.

As a back-of-the-envelope example of how describing processes can take much more space for a computer language, consider the game Towers of Hanoi. Wikipedia explains the game in about 7 sentences. An implementation in C, a popular programming language, is about 43 lines. The number of primitive instructions that your fancy Intel CPU needs to run this program is 137 (this can vary significantly but it will always be some larger multiple of the preceding 2 numbers).

The implication of the fact that computers understand such primitive languages is that every little detail need be spelled out, much like, say, instructing a toddler to do laundry :-)

Analogies for programming

Sun, 14 Mar 2010 14:27:00 -0400

Often, I get asked by nontechnical people what exactly programming is, but it’s a somewhat difficult thing to answer concisely. When I’m thinking about this, most of my comparisons are to: running a team at work that consists entirely of toddlers.

Just like a manager might distribute a memo to his/her team indicating a process to follow in certain situations, a programmer can be considered to be distributing a document that contains a process to follow in certain situations. The manager’s memo might indicate what to do if a link to another office goes down and the team needs communication from them; the programmer’s “memo” might indicate how to modify a photo if the user decides to bump up the contrast setting in Photoshop.

A program, then, is, collectively, a bunch of “memos” that are executed by the computer when the user triggers an action (a mouse click, keyboard press, or any form of input) that the programmer has decided to associate with that memo.

Programmers usually work in units of code files which can be considered documents containing instructions for the computer to execute. It’s up to the programmer to use somewhat primitive instructions to build up documents that provide functionality we take advantage of every day.

The reason I think that the analogy to a manager running his/her team is apt is because a big part of a manager’s job is setting process for his/her team. Similarly, a huge part of computer science is the study of processes: the formal expression of what a process is, the execution of a process efficiently (The word for this which Google has made trendy is “algorithm”), and the analysis of effects of a process.

Next up: the toddler part of this analogy :-)

Cognitive Dissonance

Wed, 10 Mar 2010 14:56:04 -0500

If you’ve ever thought about why first impressions matter so much, it’s probably because of cognitive dissonance, something that Scott Adams, creator of Dilbert, writes about quite a bit (you might have surmised that he’s one of my favorite writers from the fact that this is the 3rd or 4th mention of him on my blog :-)).

Unless someone takes the time to limit the impressions they form during a first meeting (and assuming there are future meetings), cognitive dissonance indicates that it’s very hard to undo or change those initial impressions.

Another example at the other end of the spectrum is how malleable someone’s opinion of what’s important can change over time. I grew up in a household emphasizing balance but I will still pretty obsessed with making a lot of money and having prestige when I was in high school & college. It took more than a few promotions & instances of meeting job-related goals to discover that ascending along this path wasn’t everything I’d thought when I was growing up, and that it wouldn’t be the right thing for me for the future. The fact that it took a few years is probably because of my own cognitive dissonance!

Mobile ads

Mon, 08 Mar 2010 02:13:00 -0500

I went to Patxi Chicago Pizza with a friend not too long ago, and we split a one of their signature deep dish pizzas. In between shoving down slices of pure heaven, we checked in via FourSquare and received notice that we’d get a free fountain drink for checking in. Too bad our mouths were too full for us to say anything to the waitress.

Some of the other deals available on FourSquare revolve around offers to the mayor, but it seems to me that the emphasis should be on getting customers who are not frequenters to frequent more. I don’t mean to be a heartless businessman, and I am all for rewarding customer loyalty, but it’s likely that you don’t need to spend much money to entice your mayor into coming in more. In fact, in the case of bars, if someone is the mayor of a joint, they’re probably getting enough free drinks already just out of knowing the bartenders there.

Something that I envision is a check-in system, not tied to a specific establishment, but to some geographical area like a neighborhood. I’d also like to make it seamless for businesses to run promotions on a minute-by-minute basis. Then, if the restaurant A16 has determined that it’s slow and would like to offer 2-1 drink special for the next two hours, they can do so for anyone who checks in in the Marina neighborhood.

The mayor concept is fun and maybe makes FourSquare more used, but I’m not sure it should be the primary target for advertising specials.

Google Analytics provides a pretty nice way to track the...

Mon, 08 Mar 2010 01:48:24 -0500

Google Analytics provides a pretty nice way to track the visitors to your site. From a high level overview that aggregates information about your web site’s visitors, you can drill down into various stats about your users, such as whether they are new/returning, what kind of platform/web browser they are on, whether the traffic was direct or the result of clicking on a link for another web site, etc.

The picture above shows where visitors of my blog come from; I’ve obviously gotten no international love going in the last week. Unfortunately, there’s also no Alaskan love. It’s just highlighting it as part of the lower 48 :-)

One metric they track is the ‘bounce rate’; this is the percentage of users who come to the first page (the “landing” page) and don’t click on any other links. For blogs, this tends not to work so well, because people might be subscribed via RSS, or, even if not, they would come to the blog to read the most recent entry, which would mean that they’d never click on links to see older entries or other parts of your blog!

Reading between the lines

Mon, 08 Mar 2010 00:35:00 -0500

When I started my first job after grad school, I was starry-eyed, young, eager, and ready to change the world through kick-ass software that would set a new bar. I still am, but I’m also a little less impatient for it to happen yesterday :-)

There were some pretty high and some pretty low times in my first job. But, none of it was bad experience. At times, I felt like I was in the wrong field, or that I was horrible at my job, or that a way different skill set was required of me than anything I’d been exposed to, and I was never going to be able to learn this skill set before people gave up on me.

Some of those skills weren’t really anything related to delivering software. What I’m referring to are the skills related to teamwork. I remember being particularly bad at conflict, which, of course, happens all the time, in the workplace and in personal interactions. I’d get upset and walk away in a huff, wondering how the other person couldn’t possibly see what I was saying.

I was fortunate that my first manager was an exceptional manager; he could see something from almost anyone’s point of view and cut through the disagreement with crystal clear explanations. Without delving too much into details, I’d like to say that teamwork & collaboration were pretty new skills to me coming out of college, and it very jarring that the “point, counterpoint” culture of science & engineering in the university setting did not really apply to the teamwork involved in building & shipping actual products, or personal relationships, for that matter.

Being born on 3rd base

Sun, 07 Mar 2010 01:29:00 -0500

My friend Robert has a saying: “Some people are born on 3rd base and think they hit a triple.”

The saying applies to people who might not recognize the opportunities they had growing up. Sure, someone might have slaved away to get a 4.0 at an Ivy league school, but, if their parents paid for it, the circumstances they faced growing up still nurtured them at least partly into that environment.

I was, in fact, born on 3rd base. My parents, who did not grow up well-off, immigrated to the US & worked very, very hard to provide a comfortable, almost lavish, upper-middle-class lifestyle for my sister & me. They paid for college and bought me a nice car when I finished grad school; although, out of some now-laughable rebellion-without-a-cause, I paid for grad school by taking out loans, just to make a point. Perhaps most importantly, though, they set a high bar for what I should expect out myself

The point isn’t whether you were born on 3rd base or not; it’s whether you steal home once you’re there :-)

Startup Advice

Thu, 04 Mar 2010 19:01:00 -0500

Startup Advice:

I found Mark Suster via Fred Wilson; in particular his series on startup advice (linked above) is very worthwhile reading!

Dodgeball

Thu, 04 Mar 2010 12:49:09 -0500

I was chatting with a co-worker and Dodgeball came up; it was unfortunate that the service was more or less killed after the acquisition.

However, alternatives like FourSquare have popped up, and that’s awesome, because now I can find out who might be a “competitive customer” at Mo’s Grill, where I’ve been the mayor for several months now :-)

For any potential would-be political ascendants, just a warning: please heed the incumbent and his appetite for the French burger.

I wonder how much the idea of being a mayor has to do with the success of the app. It kind of reminds me of something Randy Pausch wrote about in his book The Last Lecture, which I highly recommend. He writes about a concept called the head-fake. When you’re trying to learn/do some task, it helps to have a tangential goal that depends on you getting that task done. Specifically, when he was a kid, he talked about learning self-confidence through having a coach that taught him football.

Personally, I think the head-fake comes up all the time, and being able to head-fake yourself to get something you want to do done is a very useful skill to have. For instance, as a software engineer new to a team/code base, a manager might have them fix a bunch of low-priority bugs, but the real value out of that is not just the bugs, but the knowledge the engineer gains in the code base. But if you told him or her to “go learn the code base”, the lack of clarity, deadline, or clear expectations around the goal would make it an endeavor that could stretch on forever.

Evolving into a Platform, part Tres

Thu, 04 Mar 2010 12:15:36 -0500

When I was thinking about this last night, I went down the path of whether a company can create a platform that made it competitive for other application developers to build on that platform.

This is necessary but I think it was a bit of a roundabout approach; if your platform speeds building features that the end-user needs, then companies will come to build on you. I think that’s pretty obvious and I just needed a good night’s sleep to say that concisely :-)

Take the domain of ‘status-updating’ applications, if that’s even a domain: Facebook, Twitter, MySpace. Could you imagine a platform that made it competitive for those apps to build on it? It seems to me that one of the key advantages of any such platform would be the social network itself, which is exactly what any of those status updating sites would not want to cede control of at all :-)

As another example, take photo-storage/presentation, like Picasa or Facebook’s photo apps. I can’t count the number of times I thought it’d be advantageous to upload my photos somewhere and be able to present them on Facebook, or Flickr, or on Picasa, or on MySpace, or in email. Instead my photo collection is scattered among Flickr, Facebook, and Picasa and I keep putting off unifying it out of the realization that eventually I’ll have to put them back on a site just so my grandparents can see them, since they only know one site.

Suppose you could offer users the incentive “upload your photos here, and it’s super easy to show them on whatever site you’d want to”. Users would upload photos to your site, then go to Facebook and do all the appropriate tagging/posting on walls, etc. Which means the platform would have to support innovations that happened at the application layer (particularly, tagging, among others)

The incentive for applications to build on top of this? Well, they could save on photo storage, for one. That has to cause some of the greatest storage requirements from their users. They cede control, but they avoid turning away a user that already has all their photos uploaded somewhere.

Anyways, I am speculating that this has already been tried, and Facebook would probably pay the storage costs to maintain storage of users’ data; it’s part of what makes a social networking site sticky.

http://dilbert.com/blog/entry/knowledge_that_matters/

Thu, 04 Mar 2010 00:47:07 -0500

http://dilbert.com/blog/entry/knowledge_that_matters/:

Scott Adams talks about what kind of knowledge is important.

LOL

Wed, 03 Mar 2010 21:38:54 -0500

I used to dislike the acronym LOL. So I never wrote it. I always thought I was the only one who had an active dislike for it.

Then I started writing it. And I’ve discovered how many people dislike it. We could have formed a support group.

But, alas, no one ever told me, “Neal, I really like how you never write LOL”

Evolving into a platform, Part Deux

Wed, 03 Mar 2010 21:13:41 -0500

I think my last post tried to say too much and failed so I’d like to explore a few more things related to being a platform vs. building an app.

The first was the bit about “building a platform to for others to innovate on, rather than providing those innovations…themselves” This of course does not mean that providing a platform is not innovating; in fact, innovating by creating a novel platform design is the only reason you would go down the route of building a platform for others, because if you can’t innovate, then why would anyone build on you vs. building the platform for themself and eliminating a dependency?

The second point is related to the issue of whether becoming a platform is possible in every app domain. After tossing this around in my head today, the only thing that’s become clear to me is that some applications are semi-platforms and make this distinction blurrier than I thought. Photoshop has a thriving add-in market. Does that make it a platform? I’d say yes, but of course it’s not a platform in the sense that Windows is a platform. A developer probably couldn’t write a Google Talk instant messaging client for Adobe Photoshop. You might ask “well, would they want to?” and I’d say that’s irrelevant; there are plenty of things that developers want to do on the iPhone that they can’t because of sandboxing restrictions, but it’s definitely still a platform. Perhaps the difference is whether being a platform gives a company a competitive advantage; in Microsoft’s case, having a ton of apps and being the enabler for those apps through awesome development tools is a competitive advantage; for Adobe, their competitive muscle comes mostly from the expertise that is embedded in Photoshop itself, and the platform aspect is secondary.

The third point, and hopefully the one that makes the above two a little less academic, is whether they give any insight into current applications and whether they can grow more valuable to users by being a platform. I’m going to save it for another post because I’m practicing brevity :-)

music nostalgia

Wed, 03 Mar 2010 14:37:59 -0500

Bivouac - Jawbreaker

I was given a copy of this album when I was 14, and to this day, I can’t really listen to it while I’m working, because I just stop working…and listen :-)