The first half of the history of computing has been about the server side, with dumb clients.  The second half is going to be about intelligent clients and severely dumbed down servers.

It is instructive to look at this trend from the lens of the browser. The defining technologies on the client side were HTML and HTTP. The browser knew how to fetch HTML, and could display it. Javascript was a much later add-on.

So engineers flocked to the server side. Server side jocks were born.  The bearded, suspender wearing sysadmins of the ’80′s migrated to the shorts and T-shirt crowd of J2EE toting server side gun slingers. The job of the browser was to render HTML. The server’s responsibility was everything else – database lookups, service integration, authentication, validation… – everything.

Soon, Javascript came along. And it was treated with disdain, as a toy. Server side engineers expressed their disdain of a bratty little interpreted language. No self respecting engineer would work on the client side.

But Javascript continued to grow. In spite of all its warts, it continued to grow. (I highly recommend watching this video of Doug Crockford talking about Javascript – the good parts). And all its warts got covered up by very capable frameworks like jQuery. Browser compatibility continues to be an issue to the day, but tools like jQuery do a good job of managing the differences. Soon, this evolved to client side MVC, reactive programming, two way binding – things like backbone, knockout  and other frameworks evolved to meet the needs of more complexity on the client side. And HTML5 provided the ability to manage data, visualization, and state much better on the client side.

Soon, there was a realization – we can offload more of the work to the client. Kicking and screaming, server side work was being dragged to the client side. The Servers role is relegated to providing the data – JSON, typically – to the client ( I acknowledge that I am simplifying things a bit). The client is responsible for determining how to display it.

I still remember a conversation in 2011, where a server side engineer was aghast about the notion of the client side doing anything other than rendering HTML. The attitude was that the browser, and hence the customer, is a necessary evil.

Let’s look at this from the lens of a consumer of data. There is no question that a device, a browser, is necessary. That is how a consumer interactions with your rendition of the data. The fact that a server exists is inconsequential to the consumer. To the developer, the server is a necessarily evil. You MUST have the browser, the mobile app. You NEED the server because the client cannot do everything. Or can it ?

The server as a necessary evil

So what do servers do exactly?

- Serve up data

- Handle connections from lots of clients, to pull up data

- Do analytics on the back end (preprocessed or at runtime)

- Provide security functionality

In reality all of this can absolutely be done on the client side. The problem is about space, bandwidth (network and CPU).

In fact, that is the direction that new frameworks like meteor are going. JS on the client side, JS on the server, like Node.js. But wait, there is more ! Direct access to the database (Mongo) from the client. The data is stored as BSON in Mongo, allowing rapid conversion to JSON, the data format of choice in the browser.

To be clear, there is still a server layer in Meteor. The trend is clear, however – Increasingly, the role of the server side is continuing to be marginalized. It is all about the client and access to data. Dramatic advances in computation power and power consumption on client devices means that more work can be offloaded to the client.

In a decade, the balance of power will flip completely between the server and the client. The server will do niche jobs and large analytics. New applications build on clients will be powered by client side technology – be it JavaScript, Cocoa or client side Java technology.  The smug looks on server side engineers will be run over by smirks on the faces of client side technology experts.


After 4 months of waiting, I finally got my new Watch: The Basis. You can read more about it here. It has a whole bunch of sensors in the form factor of a watch, that makes it really convenient to monitor your body.  It continuously captures heart rate patterns, motion, perspiration and skin temperature. The website contains fascinating data about a lot of stuff – your temperature, prespiration, how many calories you burned during the day, with an hour by hour breakdown. Awesome for data geeks.

Of course, the first thing I wanted to know is my vital signs during ashtanga practice at my awesome shala, Yoga is Youth in Mountain View, CA.

The most interesting insight was how Ashtanga impacts perspiration. I know I sweat a lot during practice, but the closing sequence cools us down, and the final shavasana (corpse) posture relaxes us. But this is what – consistently – the basis band showed. In the chart below, my practice was done around 9 AM. As would be obvious, I started at 7:30. I started the closing sequence at 8:50 – but I really did not stop perspiring till 11:30 – when I went for a shower! So what I’ve been told by teachers is actually true – wait for atleast an hour before having a shower after practice.

Perspiration - asthanga primary sequence completed at 9 AM

In case you think this is an error, here’s the reading from Thursday, when I started at 6:30, and went for a shower at 9. Perspiration continues till I had a shower.


Heart Rate

The heart rate capture is a little bit annoying. As you can see from the chart below, Basis seems to have some trouble capturing the right heart rate when you are perspiring (I think). So suddenly your heart rate drops from 110+ to 30.

Ignoring these losses, the chart shows what I’d expect: A gradual rise in heart rate, and then a slow down as the closing sequence is done.

The peak was reached at around 8:45, of 133 beats/sec. Right when I was doing supta kurmasana , Titibasana and the jumpback, I bet. (check it out on youtube if you don’t know what that is)


I do not buy what Basis is telling me – that essentially, the days I do 2 hours of a really physical practice of ashtanga, I burn essentially the same number of Calories as I do on non-Ashtanga days.

I suspect I know the reason for this. I have noticed that doing Ashtanga does not add significantly to the step count for the day, which presumably is how a major portion of calorie consumption is computed. Here is a monday compared with Sunday



I have not found the skin temperature to be very useful. It clearly increases during practice, and then goes back down after. Just what you’d expect

Finally, the step monitoring is actually very useful – I found it quite motivating to try to slip in some extra walking during the day. The feedback cycle really works.

The strap

I really hate the strap. If you fasten the strap too tight, the sensors will start digging into your skin. I do not know if this is because I am not used to wearing a watch, but I find myself adjusting the strap every 15 minutes because it is uncomfortable. I wish Basis would offer the option of a cloth and a steel strap. Plastic sucks, IMO. 

In Conclusion

Any device that gives feedback on how you’re doing is bound to be interesting, useful and motivating.  I highly recommend trying it out. Oh, it is also waterproof, though I prefer removing it when I have a shower




I ran into this problem recently. It was pretty hard to correlate the error to  the solution.

hitting my /users/new method had suddenly produced this error

undefined method `model_name' for NilClass:Class
in my view on this line
 <%= form_for(@user) do |f| %>
The obvious thought was that somehow my controller did not define @user – but it did. I put a debug statement in the new controller which did not execute. 
The real reason for the error? A syntax error elsewhere in the file. There was an unbalanced ‘end’ statement in a different method in the users controller.  I would’ve expected a syntax error, but no such luck. Fixing this resolved the problem


As usual, I discovered I knew something because someone asked me how I do something they had observed me doing.  This has become an interesting pattern for me: I am unaware of things I am competent at. This is not a good place to be because it means you cannot teach it to others. Being consciously competent, I have found, sometimes requires someone to ask me a question.

Let’s put that aside for a moment. The question I was asked was about thinking strategically. I was really taken aback to see  my version of the answer slip off my tongue effortlessly: It is then that you realize that you are pretty good at something without having known it.

So here’s my framing of how to think strategically:

Imagine that you are in a room with a bunch of people. Someone has opened up an idea and there is a free form discussion in the room about the idea.

So what happens in meetings? Someone says something. And there is a reaction to it. And yet another reaction. And so the discussion continues, usually bound by a quantum of time. 1 hour? Have a relaxed discussion for 45, and demonstrate a sense of urgency in the last 15.

The problem with these discussions is the focus on action-reaction.  Reacting to something results in narrowing of focus, and an examination of a path laid out in front of you – and disregarding other avenues that might exist. Indeed, this is the reason that brainstorms are ineffective. They narrow the focus down prematurely.

Narrowing is appropriate in some instances and very inappropriate in others. The problem is when we do not take the time to distinguish between the two.

Quite generally, I have observed three kinds of people in discussions

  • The Take Charge guy. This person will lead from the front, be fearless in saying what is on their mind, and will be the person to get  things closed out and keep an eye on the time.
  • The timid minority – They will barely contribute to the discussion, and when they do, it is because they are very sure of the facts. Being vulnerable is not a strength. They would go to great extents to avoid being corrected.
  • The Quiet listener – this person usually will not speak for the first 5-10 minutes and take the time to absorb what is going on. But when they do speak, they have a significant impact on the discussion, and often it will be a revelation to others – Why did we not think about it? This is your strategic thinker. They garner respect and everyone waits for him or her to speak up. They are able to do it because they have a slightly different frame of reference from others.

What this person does is apply the meta-thinking principle. Stop, uplevel, understand the bigger picture. Often, what this person finds is that everyone is missing the big picture.

If the room represents a bounds on the limits of the discussion, what this person finds is that there is a door that leads to another room, and well as one that takes you to the courtyard. They first zoom out to this bigger picture, understand it , and then either zoom back down to the smaller level – or help others understand  the bigger context.

This can sometimes work against this person – Strategy and tactics go hand in hand.

“Strategy without tactics is the slowest route to victory. Tactics without strategy is the noise before defeat.” – Sun Tzu

Being stuck in the big picture land without the ability to ferret out the path to take is a problem many leaders face – especially those who excel in being able to see the big picture. The true leader is the person who can go broad first and then help narrow down, while keeping the big picture in mind.

My advice: Actively cultive the big picture thinking. Resist the urge to jump directly into conversations till you have understood the big picture. The best way to do this: prepare beforehand and understand the big picture. Become the guy that everyone waits to hear from because they are likely to have a deep thought to share with you that could completely upend the discussion. And then be the person who will help zoom back in and propose how we should solve the problem – and actively narrow the scope of discussion.


My 2007 macbook pro was too old to take the mountain lion upgrade. Fine, I took the plunge and forked over the money for the new retina macbook pro.

You don’t normally use words like ‘breathtaking’ when talking about laptops, but I am tongue tied. This laptop is like nothing you might have used before. I’m tempted to say ‘and ever will’, but I know that’s not true.

If you are like me, you have seen this laptop and walked away saying ‘Meh’. Trust me, once you start using the Retina display, there is no going back. More than pictures, it is text that I still gawk at. Brilliantly clear and sharp. Writing code on it goes from fun to downright pleasurable. The form fact is just perfect. So much thinner and lighter. I struggle with using the previously beautiful apple 27″ monitor. Suddenly, everything else has glaring imperfections.  Going from retina to non-retina is a struggle. I have never heard the fan (even when I am running flash).

It reminds me of what Job’s said -’ Design is how it works, not how it looks’. This laptop is a real winner. The only problem is all the drool landing on the keyboard



The Shopify blog put together a list of 12 talks for entrepreneurs. Here is my list of the major points of the talks:

Dan Ariely:  Are we in control of our own decisions ?

(read his amazing book, predictably irrational)

We are not always in control of our decision making – and how choices are presented to us influence us heavily – and we are completely unaware of it

Rory Sutherland: life lessons from an ad man

Seth Godin: How to get your ideas to spread

I found this having a lot of interesting overlaps with Seth Godins presentation:  You have to either build a remarkable product (Godin) or figure out a way to make it remarkable to people (Sutherland). It is either about having enough implicit value for people that they talk about it (‘no one talks about a cow by the side of the road – but they will if it is purple’) or creating creating intangible value -potatoes became a staple in German food (the kings garden started growing it and guarded the crop after the citizens refused to eat it)

Simon Sinek: How great leaders inspire action

Entrepreneurs make a mistake by focusing on the ‘what’ and the ‘how’. Great leaders focus on the why. Inspire people. Martin Luther  King said ‘I have a dream’ not ‘I have a plan’.  Apple employees believe that they produce great, beautifully crafted products, and consumers can see the love in it. They buy into the products emotionally

(to be continued..)


A recent conversation with a colleague left me thinking about decision making, and the effectiveness of decision making.

How decisions are made impact the effectiveness of the decision, and it is not obvious that everyone thinks through this before deciding.

Imagine that you are a manager. You can

  • Mandate a decision
  • Make a decision, but consult others before making it
  • Delegate a decision or influence the outcome of a decision

From top to bottom, the difficulty level of decision making increases. From bottom to top, the value of the decision decreases.

Let’s consider a decision you have mandated (‘Everyone checks in code at the end of the day, period!’ you yell in an email). The decision is quick. But no one else on your team has skin in the game, and so there is no incentive for anyone else to actually follow this. In fact, it is quite likely that someone on the team is bitching about it behind your back (‘ What a clueless guy! How can we check in code that does not compile? We will then be held responsible for breaking the build !’)

A better option is to consult the team before deciding. Solicit opinions. Use them to come up with a resolution for the underlying problem you are trying to solve. It will take longer – but the decision has more value, because others have contributed to it.

The hardest kind of decision that you will make is one that you do not make. This happens when you delegate a decision or influence someone else – a peer, for example – into making the right decision. This is hard stuff, and it takes time. Yet it is the most valuable kind of decision – because someone else made the decision. They will be committed to it, more so than you. They will go out of their way to get to the right outcome for that decision. These are the kind of decisions that leaders make, by influencing outcomes that they do not control to get team buy in. You know a good leader when they do not throw their weight around, but ask others to make decisions and influence the outcome


Like many others, I mourned Steve Job’s demise. I chanced upon an article which described how he quit Altair because he wanted to visit India.

The obvious thought after reading this was to indi-fy his picture.

Of course, this work was done on a mac, using pixelmator from the macos app store.


The technical sessions
I can’t talk much about the technical sessions, because of the gag order, but trust me.. if you are a developer, you will be salivating at getting your hands on iOS5. Tons of new api’s, and lots of fixes/missing functionality

The Truth is in the cloud
This is a remarkable makeover for a desktop/device centric company. The transition is happening at a pretty remarkable pace too. iCloud makes it easier for iOS developers and consumers. Moving data is seamless, and requires very little effort.

Google is about the web. Apple is about native

Last year’s WWDC had a lot of HTML5 centric tracks. This year, they are all gone. There was no mention of it in any of their keynotes, and there was maybe one or two session on Safari/HTML5. They are adding tons of API’s for native, zilch for services.

My Take: Apple is deemphasizing web centric development, because they want to channel everything through the app store. It is not about technology, it is about control. For example, they want a cut from their ‘NewsStand’ functionality from subscriptions. And there are attempts to bypass them by publishers who are unhappy about this, like this interesting story about the Financial Times .

But what does this mean for companies that need to support iOS as well as other platforms, like Andriod? There are no web services available. It is unclear what the terms of service will say. For many companies, unless Apple moves to define web services for iCloud (as opposed to native bindings), iCloud will be worth avoiding. Using iCloud would mean support iCloud for iOS, plus additional services for non-Apple platforms.

I suspect that many companies will take a hybrid approach, like doing some stuff in a UIWebView, while other functionality will be in a native shell, and they will be stitched together using tools like PhoneGap.

‘It just works’

This will sound like a fanboy statement, but I believe Jobs when he says this. The demos on opening day were quite astounding. Make a change on one device, and it is available on another pretty much instantaneously. They have done a terrific job of making moving data part of an OS task, irrespective of whether the app is running, suspended or otherwise not running.

It truly did work. For me personally, this solves a big hassle of having data spread across multiple devices. I am looking forward to this.

Cutting the cord

Finally. I am ecstatic. iOS 5 will not require you to connect to a computer. While the statement seemed to be driven for people whose first and only computer is an ipad or an iphone, this will be a huge help for people with multiple devices also.

Yep, it will maintain MacOS as the leading ease of use OS. I wonder about the scroll part though (which has been reversed to make it consistent with touch behavior). I tried it out and it blew my mind instantly. It will take getting used to.

The $30 price tag makes the upgrade a no-brainer. Downloads only from the app store? How the heck does someone in say, India, download 4GB’s? It will be a killer, IMO. I struggle through Xcode installs now, at 4GB, and have to let it run overnight. (My cable modem ISP is throttling me, I’m pretty sure). I read the statement as ‘we don’t really care about non-1st world countries getting Lion’. Fortunately, incremental downloads are on their way.

There is a reason I prefer Chrome over Safari. Speed and the ease of upgrade. It just happens, and I don’t really care about uprades. Apple could learn a thing or two from Chrome.

Buy once, use everywhere
My wife and I both have iphones. We both have a MacOS machines. I do not need to buy apps multiple times – buy once and use it on all the machines associated with your account. Nice


An interesting article doing the rounds is one titled ‘Why Women Rule The Internet‘. It is authored by Aileen Lee, who is a partner at Kleiner. The article is just strange. I fail to understand the premise and the conclusion it draws.

The premise of the article is
- Women are the majority of users on social websites
- Women shop a lot on the internet, more so than men
- More women are active on twitter. (Apparently, Twitter is supposed to be a ‘techie insider’ product. I am stunned)
- So Women rule the internet

It is no secret either that women are more active in social aspects of life, from relationships to parenthood. (I hope this does not sound chavunistic). Men’s magazines are about fast cars, girls and more fast cars. Women’s are about relationship, fashion and beauty. So why is it surprising that social web sites – all about relationships – attract more women than men?

How on earth is ‘Women shop a lot on the internet’ insightful? I have no interest in digging up stats on this, but really, is that so different from the experience in malls? Women shop more. The fact that they shop more on the internet – well, duh.

I find it interesting to see how ‘consuming’ is being mistaken for ‘ruling’. Consumers are not rulers: they are the suckers who purchase products and fork over their hard earned money. The producers of web properties are the rulers. They take your money to the bank. Women will rule the internet when they produce the web sites where consumers – men and women – go to consume. It is heartening to see more women entrepreneurs, as alluded to by the story. But to claim they rule the internet is silly. When the next generation of Facebook, Zynga, Groupon and Twitter has female CEO’s, there is a claim to be made. Till then, the only claims that can be made are ‘there are more women consumers of social and shopping web sites than men’ and ‘If you are selling stuff online, make sure you take female consumers into account’.