The Hurdle of getting a PhD

Friend x: You are now working or studying?
xman: I'm now working in full-time basis, and studying in part-time basis. :)
Friend x: Not bad huh! 8) Getting a master?
xman: No. I'm getting a PhD. :D
Friend x: Wow! PhD can do part-time 1? :o When are you going to finish that?
xman: mmm.... may be 2 years later. :? (it's just a random number, I wouldnt know if I'm able to complete it at all. :( )

Theory vs Practice: Part I
Theory says: Assuming this and that, ignoring those and all those, we can show that method A works much faster than method B. 8) Hooray! :) We rigorously proved something. 8) My theoretical results looked beautiful like art, covering much wider range, and can sustain much longer than your experimental results. When you upgrade your machines, every 4 years??, you'll have to re-run all your experiments. :( What a boring task for a short lifespan of results! :?

Practice says: Given this set of data, information, given an old machine, my experiments tell method B works up to m times faster than method A for all the data set I've tested :o . It seems that we should use method B instead }:) , luckily I didnt blindly following your theory! }:) Obviously, your assumptions do not hold for my case, what you are ignoring can have 'butterfly effect'. What a beautiful theory, yet hardly useful in practice! :? Who cares with those mathematics in your papers. :?

Theory says: Well, your results are valid only on your old machine. New machine might show your results no longer useful. 8) How many experiments can you afford? :? Experiments on all diffferent types of machines in the world? :? How many types of methods can you afford, there might exists a method C that's much more superior than your A and B. How much time it takes to run one experiment? :o Why should I trust your numbers published in the paper? :? How do you explain the special curves in your graph plots? :o If your job is merely running experiments, why do you need to do PhD? :o }:) I can pay a low cost engineer for that.

...... to be continue ......