Weeknotes: 013

Ph.D

  • After much procrastination and needless delay, I’m almost done. This week I fixed up the abstract, conclusions, and future work—none of which took particularly long once I sat down to do it.

    I’ve also updated the title page to say “January 2019”.

    For next steps, I’ve got a couple of things to check:

    • That all the code snippets I give have type signatures.
    • That each of the three “contribution” chapters says how it applies beyond Haskell.
    • That I haven’t accidentally broken the university’s formatting requirements.

    Then I can resubmit!

    I’d also like to know if this is it, pass or fail, or if there’s room for another round of revisions if the examiners have comments. If the former, I’ll make sure my internal examiner is very happy with it before formally resubmitting.

Work

  • We did some more load testing, this time focussed on the dynamic content, to see what effect the changes made since the previous load test have had. It’s much better now.

  • This week I’ve been on call for out-of-hours support since 17:30 on Wednesday, and will be until 09:30 next Wednesday. The way it works is that there are three people on call: the least experienced is the “primary” (this is me), there’s a more experienced “secondary”, and then someone from the management team beyond that. When something goes wrong, the primary gets called; they can acknowledge the problem and start working on it, or escalate to the secondary, who can then acknowledge or escalate. If it gets to the third person I assume they start ringing up other people who could help. The reason the less experienced person is the primary contact is because, if it were otherwise, they’d never learn anything as the more experienced person wouldn’t ever call for their help.

    While there are many things that can go wrong, only a few will call someone out of hours. So, of course, I naturally got pulled into something on Friday at 17:30, before leaving the office. It’s been uneventful since then though.

Miscellaneous

  • For Advent of Code day 11 I had the opportunity to use a summed-area table, a neat data structure I’d not come across before. If you have an n by m matrix, you can build a summed-area table in O(nm) time, which then lets you calculate the sum of all the elements in any arbitrary submatrix in constant time.

  • I made lemon and poppy seed muffins on Saturday. They turned out pretty well for the first thing I’ve ever baked. It’s a good thing my measuring spoons include the weird US measurements like “cups” though.

  • I started reading Tribe of Mentors, by Tim Ferriss. In which he sent 11 questions to a bunch of successful people, and published the best answers. I quite like this one by Susan Cain:

    When you feel overwhelmed or unfocused, what do you do?

    I love espresso and would happily consume it all day. But I only allow myself one latte a day, and I save it for when I’m doing my creative work—partly because it jump-starts my mind almost magically, and partly because this has trained me, Pavlovian style, to associate writing with the pleasure of coffee.