3 ms·
Simply reading stuff is like passively watching TV/youtube. It evaporates fast. Active learning involves your hands. Two techniques work well for me. First, I
by sriram_malhar 4y ago
Simply reading stuff is like passively watching TV/youtube. It evaporates fast. Active learning involves your hands. Two techniques work well for me.
First, I write code to understand the essence of a piece of software (I have written basic TCP/IP/ARP stacks, a Java compiler, a virtual machine, a database, a neural net library etc). It is a joy to write code for an audience of one when you don't have to worry about error handling!
The second is to add an entry to a diary recapping (in your own words)your own understanding of the topic. In my PhD days, I put everything in a single Word doc; I chose Word because I could draw diagrams in it, search with regexes, and the file was available offline. By the end, the file was thousands of pages long. Here's what I had in it.
1. For every paper or blog post I read, I put the paper's name, the URL, author's name (in BibTex form, but that's not germane to your issue). I captured my understanding of the paper (about half a page) and put together a list of questions for which I needed answers. There were some papers that went right above my head, so that entry was all questions. I had back and forward references to other entries in the form of embedded links. This exercise (it turned out) helped me smoke out scenarios that the paper had not mentioned. That becomes the basis for critiquing and comparing.
2. When my understanding of a topic improved, I would go back and fix some of the outstanding questions, along with a meta comment about what it was about my previous understanding that I couldn't answer the question in the first place.
3. I had design ideas, little bits of code, benchmark results etc all in a single linear order. When my advisor asked what I was up to, I could quickly look up the log of entries.
It is ok to forget; no one can remember it all. Just capturing the essentials and being willing to work from first principles is good enough for most interviewers.