the only time the company that for years is manipulating DAILY your “wall” for its and its associates’ profit, reports back to you by publishing some moderately interesting insights about human behavior, you get angry?
People think focus means saying yes to the thing you’ve got to focus on. But that’s not what it means at all. It means saying no to the hundred other good ideas that there are
TAL (Time to Abstraction Leak) measures the time it takes for an abstraction to leak. Because eventually, all of them do.
ALV (Abstraction Leak Volume) measures how much of the inner gory details you need to figure out when the abstraction eventually leaks,
The two metrics are very important when you start using a new framework, which will (like all others) strategically improve your productivity. Unfortunately the metrics are inexistent.
… then take a break and do a comprehensive search of the related work. You would be amazed by how often another scientist stole your idea, ran with it back in time, and published it there in a prestigious conference of that time!
This is why doing a literature survey at the beginning of the PhD is better than at the end when you are writing the thesis :) On the other hand, if after doing survey the idea starts to look less original than at the beginning … read the next post to feel better.
You’re probably doing something really wrong.
I mean, think about how many stupid ideas you’ve seen already. It does not matter how stupid they are, somebody has already thought about them. If it’s hard to come up with an original stupid idea, how about something that’s really brilliant?
Modern art is, it seems to me, a domain where originality is an ideal to which people aspire. The other day I saw that a Dutch artist turned a Dead Ostrich into a Quadcopter. I guess this is quite original. (Surely, if you don’t consider that he has done the same thing with his cat before.)
At the beginning of my PhD, I remember talking to a professor that I respected a lot who was a pioneer in the domain in which I was trying to contribute to. And she told me, not to worry if what I did was not original at the moment. If I did it on my own, and if I kept going I was going to bring a new perspective to the field since it would be impossible that I did things exactly like everybody else. And she was right.
If other people have similar ideas to yours, that’s good. In economics they call it that there is a market for your product. It means there is potential in there. So keep going!
When it comes to #extensibility, this is the only interesting claim that you can make about your software. The opposite is boring as hell. And still, everybody and their grandma claims this in their paper.
The source code of software it’s extensible by definition for Stallman’s sake! At least theoretically, because I bet that you don’t have more than a dozen half baked test cases for it anyway.
In #sublimetext the default cursor style is hard to spot when switching between documents. Two tricks in the preferences file fix this:
In #sublimetext the default font is way too thick. At least on OS X. This line in the preferences fixes it:
I always forget these acronyms. So I write down a few of them. For now k = 5.
Bottom-Up: processing the low-level details before mid-level structures.
- LR - handle CFG in Linear Time. LR = from Left to right. Rightmost derivation means rightmost terminals are reduced first. Rightmost derivation can be memory-hungry. Require tokenization step first.
- Variant: LALR - Look-Ahead LR. More memory-efficient. Less powerful. Yacc. Bison.
- GLR - generalized LR. breadth-first search. splits in multiple parsers in case of ambiguity.
Top-Down: assuming a higher level of the tree. working down by using rewriting rules of the grammar. appropriate for natural language processing. more prone to backtracking.
- LL - Leftmost derivation. Left to right. LL(k) with lookahead.
- PEG - a type of grammar that looks similar to CFG but in which the choice operator always selects the first match. Each non-terminal represents a function in a recursive descent parser. Not all languages that can be expressed using parsing expression grammars can be parsed by LL or LR parsers