Mostly Software

    • Was reached out to by citadel (via njf), a hedge fund. They have offices all over the world, including LA and SF. Scheduled a phone call for Monday.
    • Listen to griz most of today. Could listen to horns and other wind instruments with drops all day.
    • Lots of newsletters today.
    • UTM info should not be in url params. It’s annoying. It makes the link bloated when copying. It is information the user does not care about. It’s only for the service. There should be a place in metadata. I don’t care if it’s a field in the http headers or elsewhere, but it should not be in the full url.
    • Didn’t read it, but a good summary of cryptography: https://www.garykessler.net/library/crypto.html.
    • The most recent version of unicode is 12.1. It has 137,994 characters.
      • Remember the utf-8 and utf-16 encodings are variable-width. utf-8 can use 1-4 bytes (each 8 bits) to represent a char. utf-16 can use 1 or 2.
    • Good list of db lectures: https://www.youtube.com/playlist?list=PLSE8ODhjZXja7K1hjZ01UTVDnGQdx5v5U.
    • Ordered much better heat gloves for the smoker.
    • Watched a couple “day in the life” videos from devs at larger companies to get the internal feel.
      • Instead of having employees recollect their opinions once a year for reviews (and company survey), amazon asks a daily question instead. Do you find your work important? What can be improved? etc. Some will repeat to get multiple data points. This is much better.
    • Sean Lee’s video is always a good one. It’s geared toward college/entrylevel, but he still speaks impressively.
    • The basic sorts (dumb ones like bubble) are bad in time complexity (quadratic) but good in space (constant).
    • Mergesort. Divide into smaller pieces, sort those, then reassemble by iteratively comparing the first elements of the pieces.
    • Heapsort is pretty great. A little slower than quicksort, but better worst case. Is not stable. It’s in-place, so constant space complexity.
    • Created a gdoc to summarize the structure of the ds&a interview, and to prepare.
    • Smoked 8lbs of curry/mesquite beef ribs.
    • I don’t think GitLab is very likely. All of their vacancies are explicit about ruby/rails experience as a hard requirement. Gonna hound the people who have emailed me, since this is a top choice.
    • Made 2lbs of roasted pecan butter and 2lbs of roasted hazelnut butter.
      • Burned a hole through the bottom of the food processor because I transferred the nuts straight from broil.
      • The hole (gash) was maybe 2x2x30mm. I didn’t collect any of the butter from the bottom to be safe. This is a small amount of plastic. The body will just pass plastic in about a day because it’s not digestible. The worry is the chemical (pthalates), which can be cancer causing. The amount in the batch is small, and the batch is about 1-2 months worth, so I’m not too worried about the ingestion.
      • I did throw the food processor away. It’s old and I needed a new one anyway.
    • Downloaded pdfs of Cracking the Coding Interview and Programming Interviews Exposed.
    • Reviewed some old notes on complexity theory. P vs NP. Easy to solve, easy to verify, hard to solve, hard to verify.
      • Cryptography (and other one-way functions, and jigsaw puzzles) are implicated by NP (hard to solve, easy to verify).
      • Np-hard and np-complete. While most in np are verifiable in polynomial time, some aren’t. These are the most difficult problems in np.
      • Exptime = hard to solve, hard to verify.