has_many :questions

My 4 Ds of time management: Discard. Delegate. Delay. And if nothing else works, Do.

Legal judgments aren’t decided on the basis of popularity. Then why do we use that criteria for selecting our leaders?

Interruptions kill productivity. They also help us refocus on the task at hand but my assertion is that we have too many interruptions already and anything that reduces them will help increase the productivity.

One of the major source of interruptions is people trying to contact me. Communication is important but it needs to be managed well so that it does not disturb the work. Over time, I have come up with these guidelines for people wanting to reach out to me.

Email

The first option you should try in order to contact me is email. I prefer email because (1) it is asynchronous and therefore allows me to deal with incoming messages at my convenience and (2) it keeps a written record of our conversations for future search and reference. Please keep in mind that I do not keep my inbox open at all the times. But I do reply to most emails on the same day (unless I am traveling or busy). I deal with emails in batches, about 3-4 times a day so never assume that you will receive a reply instantly. It is okay to call me on the next day (but not earlier) if you don’t receive my reply to your email.

IM

We use IM heavily internally at work so I’m available most of the time on IM. However, you should try this option only when you have waited sufficiently after sending me an email and still have not received a response from me. IM is better for discussions which are not focused. It is far more simpler to think through your question and email with specific queries. At times, my IM windows are on the second desktop, not immediately visible to me. So, my reply can be delayed by 15-20 minutes on average.

Phone Call

This should really be the last resort. In most cases, IM is good enough, allows me to not interrupt my fire-fighting work and has the benefit of saving the conversation for future reference. Call me when it is super-critical as in “people are dying out here!!”. When you call, I have no choice but to stop my existing work (I’m thinking about getting a voicemail account), and after the call is over, it still takes me 10-15 minutes to get back to the task with the same level of concentration. Please do me a favor and think whether a phone call is really necessary. Also realize that I do not always keep my cellphone around. If missed, I will rarely return the call – the assumption being that you’d call me again if it was really important.

I hope you understand that urgent and important are two different things. Think one more time before calling me. Did you email me about this when there was time and situation was not serious? Why not? You had the option of pinging me on IM too. If you do give me a call and it turns out to be a non-urgent (no matter how important) discussion, I’m going to be pissed off at you.

In summary,

Email – for everything that does not require my response within 24 hours

IM – for unfocused discussions which require heavy interaction and things that require my response within 1 hour

Phone Call – for urgent discussions only.

The underlying pricinciple behind this system is to “take care of important things before they become urgent as well“. Respect my time and I promise to respect yours.

For YouSuggest, we needed to monitor blogs, web, twitter and the traffic statistics. However, this meant opening many different sites and repeating the same searches again and again. Eventually, I got fed up with this and built a search mashup for fetching and displaying all these information nicely collated in one place. I polished it up, called it BrandPeek, and put it live at http://brandpeek.com for other folks to use. Try it and give me your feedback on how to improve this.

It is a brand monitoring tool, which allows you to listen to what is happening on the web around your (or your competitors’) brand. As of now it shows:

• Real-time Twitter search results for <brand>, <brand> rocks, <brand> sucks, <brand> should
• Google search results for Web, News, Blogs and Videos
• Monthly traffic statistics from Compete for <brand>.com

Eventually, I hope it will evolve into a full-fledged competitive intelligence tool – a one-stop page for knowing everything happening around a brand. I would like to show recent ad launches by the brand, the changes in their executive team, their sales data etc. This means that the data needs to be refined carefully to make it relevant for this specific usage and therein lies the challenge in making this valuable for the enterprise brand managers.

A book is the happiest when it is in the hands of a reader and not on the bookshelf.

A good trip is one with stories that last a lifetime.

I have been insanely busy this past week. Lots of work and Lots of travel. And still no respite. Sigh.

We are coming up with a Ruby on Rails training program shortly for enthusiasts around Noida. Actually, it is an internal training program for our new joinees but this time we decided to welcome outsiders as well. We will be talking more about this. Stay tuned at Lattice Purple blog.

Internships. Not interviews.

At Lattice Purple, we are big fans of internships. We have had the privilege of some really bright students work with us in summer/winter vacations on our projects. This gives us a chance to see and evaluate them actually at work. We get to see how they collaborate, lead and execute. Interviews have their place but just like preliminary written tests, it should be used only as a rejection test. And it works both ways. For the candidate also, internship is a better way to experience the work culture and the environment of a company, who may make strong claims of being innovative, flexible, open-minded, merit-oriented but remember, WYSIWYG.

Ignore the irrelevant.

We are startups. We have to look for exceptional candidates. Guy Kawasaki says, “A playes hire A+ players”. Such candidates being very rare, any rules like looking for a particular college degree stamp or a particular course should be thrown out. Just ignore the irrelevant. Age. Background. Gender. College. Degree. Ignore all of these. Just see what the candidate knows and how he/she has applied the knowledge in the past.

Ignore the references.

Checking references is a good way to verify the credibility of the candidate’s claims. Unfortunately, in India, references are not taken seriously and are doled out liberally. Simply ignore them.

Communication skills.

I can’t emphasize it enough. No matter how bright, if a candidate can not coherently express his thoughts in both written and spoken form of language, you are better off without him.

Do you have any tips for hiring A+ players?

Pluggd.in had a post about IPv4 crisis. The questions was whether IPv4 issue is overrated. Here are my thoughts on the same.

The real issue here is not exhaustion of IP addresses but that workarounds for address limit actually add more complexity to the network. Machines behind NAT do not have end-to-end connectivity and some Internet protocols cannot be supported. For eg. running Skype behind a NAT was (is?) a big mess.

This impact is often underestimated but end-to-end connectivity is one of the core principles of Internet. And NAT violates that. This breaks applications. Here is an example.

I will quote:

A user behind a NAT box is no longer a peer to other sites on the Internet. Since the user no longer has an externally visible Internet Protocol (IP) address (fixed or variable), there is no way (in the general case–there may be “workarounds” for specific NAT boxes, but they’re basically exploiting bugs which will probably eventually be fixed) for sites to open connections or address packets to his machine. The user is demoted to acting exclusively as a client. While the user can contact and freely exchange packets with sites not behind NAT boxes, he cannot be reached by connections which originate at other sites. In economic terms, the NATted user has become a consumer of services provided by a higher-ranking class of sites, producers or publishers, not subject to NAT.

In short, using workarounds like NAT is delaying the deployment of IPv6. But networks are actually better off without NAT. A perfect case of a temporary practical, low-cost fix blocking the shift to a better but expensive solution. In my opinion, this is a serious issue. We are breaking the Internet by delaying the move to IPv6.

Logic is the branch of study of a systematic and consistent method of reasoning to verify and extend our knowledge. It is not a substitute for philosophy – which strives for workable and useful definitions for everything: mind, matter, ideas, reality, beauty, ethics, justice and so on, science – which attempts to discover reality, and language – which defines words in order to have a universally consistent expression of ideas. Logic is not a tool for acquiring new knowledge but only a tool for verifying its internal consistency so that we don’t contradict ourselves.

The first term to define is proposition. A proposition is a meaningful declarative sentence which is either true or false. All our beliefs, opinions, facts etc. can be stated as propositions. Immediately, we feel the need to define truth. But defining truth or reality is the domain of philosophy and science. As far as study of logic is concerned, we only need to assume that true and false are two categories which are complementary sets – they are mutually exclusive and their union is the universal set. Exclusivity means that if A is true, it cannot be false and vice versa. Their union being the universal set means that a proposition A has to be either true or false. There is no other possibility.

Let us now see what relationships might exist between two propositions P and Q. Since we have defined P and Q to be propositions, P can be either true or false when considered independently. Similarly, Q can be either true or false, when it is the only proposition under concern. However, when considered together, a number of scenarios may emerge. Let us list them:

Scenario A: P is true. Q is true.

Scenario B: P is true. Q is false.

Scenario C: P is false. Q is true.

Scenario D: P is false. Q is false.

Now, if there exists a relationship between P and Q, one or more of these scenarios may not logically follow from the premises P and Q. In other words, each of these scenarios may be possible or impossible. This leads us to imagine the following 2×2x2×2 = 16 cases:

Case 1: A, B, C, D are possible.

Case 2: A, B, C are possible. D is impossible.

Case 3: A, B, D are possible. C is impossible.

Case 4: A, B are possible. C, D are impossible.

Case 5: A, C, D are possible. B is impossible.

Case 6: A, C are possible. B, D are impossible.

Case 7: A, D are possible. B, C are impossible.

Case 8: A is possible. B, C, D are impossible.

Case 9: B, C, D are possible. A is impossible.

Case 10: B, C are possible. A, D are impossible.

Case 11: B, D are possible. A, C are impossible.

Case 12: B is possible. A, C, D are impossible.

Case 13: C, D are possible. A, B are impossible.

Case 14: C is possible. A, B, D are impossible.

Case 15: D is possible. A, B, C are impossible.

Case 16: A, B, C, D are impossible.

Now we have the following information to work with: P and Q are propositions, hence either true or false, when considered independently. True and False are complementary sets, hence a proposition cannot be both true and false. And our goal is to avoid contradicting ourselves. Let us see what each of the above cases mean:

Case 1: All four scenarios are possible.

Case 2: At least one of P and Q has to be true. If one of them is false, the other has to be true.

Case 3: If P is false, Q has to be false. If Q is true, P has to be true.

Case 4: P cannot be false, irrespective of Q. This does not make any sense because it makes a categorical rejection of P being false, independent of Q,  which contradicts our premise that P is a proposition and hence can be true or false. Absurd.

Case 5: If P is true, Q has to be true. If Q is false, P has to be false.

Case 6: Q cannot be false, irrespective of P. Absurd.

Case 7: P and Q are either both true or both false.

Case 8: Both P and Q cannot be false. Absurd.

Case 9: At least one of P and Q has to be false. If one of them is true, the other has to be false.

Case 10: P and Q must be different. If one is true, the other has to be false and vice versa.

Case 11: Q cannot be true. Absurd.

Case 12: P cannot be false and Q cannot be true. Absurd.

Case 13: P cannot be true. Absurd.

Case 14: P cannot be true and Q cannot be false. Absurd.

Case 15: P cannot be true and Q cannot be true. Absurd.

Case 16: Absurd.

Let us discard all the absurd cases:

Case 1: All four scenarios are possible.

Case 2: At least one of P and Q has to be true. If one of them is false, the other has to be true.

Case 3: If P is false, Q has to be false. If Q is true, P has to be true.

Case 5: If P is true, Q has to be true. If Q is false, P has to be false.

Case 7: P and Q are either both true or both false.

Case 9: At least one of P and Q has to be false. If one of them is true, the other has to be false.

Case 10: P and Q must be different. If one is true, the other has to be false and vice versa.

Recall that we had assigned no special meaning to the category true. It is a category just like false. Thus, it is easy to see that case 3 and case 5 are similar in nature. So are case 2 and case 9.

Let us simplify our notations a bit. P means that the assertion P is true. !P is the exact opposite – it means that the assertion P is false. Let us define a term “implication” as follows. The statement “P implies Q” means that if P is true, Q has to be true. However, Q may be true even if P is not true. P is a sufficient condition for Q, but not a necessary condition.

It is easy to see that “P implies Q” and “!Q implies !P” (if Q is false, P has to be false) are the same thing. If P is true, Q has to be true. This means that if Q is not true, P could not have been true. Hence, !Q implies !P.

Now it is easy to express all 7 cases succinctly.

Case 1: P and Q are independent propositions. There is no relationship between them. Eg: P: x = 4; Q: y = 6

Case 2: !P implies Q. !Q implies P.

Case 3: !P implies !Q. Q implies P.

Case 5: P implies Q. !Q implies !P. Eg: P: x = 4; Q: x*x = 16

Case 7: P implies Q. !Q implies !P. Q implies P. !P implies !Q. P and Q are equivalent propositions. It is absurd to imagine one of them being true and the other one being false at the same time. Eg: P: x = 4; Q: x + 6 = 10

Case 9: P implies !Q. Q implies !P.

Case 10: P implies !Q. Q implies !P. !P implies Q. !Q implies P. P and Q are complementary propositions. It is absurd to imagine both of them being true or both of them being false at the same time. Only one of them can be true. Eg: P: x = 4; Q: x != 4

Thus, we have the 7 possible relationships between the two propositions. The basis of this were our definitions of true and false, declaring them to be mutually exclusive but collectively universal sets. Everything else has just logically followed.