"All is not lost, everything is left to be got"
I believe that when somebody falls in love (whatever your equivalent for it is), you do not fall in love with a person per se, but rather with an idea of that person or a perception of that person. It is not that whatever you think you know or believe about the other person is false or wrong. A significant part of those beliefs are generally true. At least in the beginning. Obviously love starts with attraction, hopefully mutual and there are possibly multiple intermediate stages. I do not have any idea about any of these, I am just trying to relate something I realized over time. So after the beginning, you start spending time - together or apart.
If together, you start learning about the other person, changing certain previous perceptions (good or bad) and adding more facets to your idea of the other person. You can think of it in a statistical learning theoretic sense. As you obtain more and more samples, your final estimate of whatever you are trying to compute changes a bit here and there, but it generally converges (at varied rates - sometimes you find out immediately and sometimes it takes years) to what the other person actually is around you. So obviously the longer you are together, the better you start to know that person (duh!). And sometimes (or most of the times) as you get closer to true value, you realize that this is very different from your idea of your prospective partner. In that case, you either compromise and settle for a "weak learner" (something or someone in this case which is at least close to what you want more than half the time) or you move on. In my opinion settling for a weak learner is generally done to avoid going through the whole learning process again and again and the fear that you might never get a "strong learner" (your soul mate :P ) and the hope that, just once in a while, the other person (the weak learner) will be what you always wanted, and those moments should be enough. And who knows what you truly wants. Since your perception about others are so skewed, there is no guarantee that your perception of yourself is even remotely accurate. Hence, this can be thought of as a double learning problem - you learn about yourself and about the other person simultaneously. The more you sample, the more you know what you might truly want. But if you settle early, you stop learning. Consequently, your perception might not change after that, so you might never know if you actually wanted something else. So you would be satisfied with your weak learner. I guess ignorance is bliss!
However, if you decide to move on, you might feel like you are back at square one. But I believe that you are not - you learn more about what you truly want or what actually is really compatible with you. So you have made some progress towards finding your strong learner. But you do have to start at the beginning to learn about the other person. In fact you start at negative, because it might be quite some time before you actually are with some one new.
Now occasionally, after the beginning, you end up spending quite some time apart. And having a long distance relationship (or some thing even less serious) might have an inherent problem which I will try to justify by drawing an analogy from machine learning. If you have this other person in your life, but you are far apart from each other and see each other only occasionally in the beginning of your relationship (or whatchamacallit). But that does not mean you do not think about that person. You think of the all the time you did spend with that person, all the happy fun experiences. The problem here is that you made the cardinal mistake in machine learning. You learn over and over again from the same sample set. You do not get yourself new samples because you are not spending any time with the other person. So you might think you have a good idea of the other person, and many a time, you think that you like that idea of the other person. But the problem with any learning model trained too much on the same example set is that it does terribly on a new sample set. What it means is that when you do see this other person again and spend some time with that person (still thinking that you have a good idea of how that other person is), if the person does something which is something out of character for you (since you have never seen it before), you are terribly shocked. Shocked because this concrete vision you had in your head does not accept this new behavior. If you were together, the vision would be molded over time, hence generally more flexible to new behaviors. But when you are together only momentarily with extended gaps of absence, you just think that everything is wrong. However remember that this line of inference is mostly valid if you have been having a long distance relationship from the beginning itself. If you have spent quite some time together, the aforementioned line of logic does not necessarily hold. On the other hand, I do not claim that long distance from the beginning would never work - using some great and fortunate sampling techniques and settling for a weak learner are one of the many possible routes to a successful long distance relationship.
But at the end of the day, you have to be a risk taker to find yourself a strong learner. If you want to that is. I mean it is quite possible to live meaningful, rewarding lives without ever pursuing your strong learner. I mean, just think of Mahatma Gandhi or Mother Teresa (if you do not like my examples, just go with it and try to see the point than the specific examples). I do not think they ever spent any time looking for their soul mates, they spent it in service of others or maybe to the service of a higher idea. Just think about how much time a person spends in trying to woo a partner or in spending "quality" time with a partner. If only some one spent that time in doing something creative or helpful you would get so much more man-hours towards something that is more fruitful to the masses.
In short, finding your soul mate is not for the faint hearted, but then "love" might be overrated. If it were not for the survival of the human kind and people were not that goddamn sensitive about sex (the activity), the world might have been different - for better for worse? How do I know?
