loading...

Friday Blast #48

horia141 profile image Horia Coman Originally published at horia141.com on ・2 min read

Programming in the adult entertainment industry is broken (2018) - fun read about all the broken stuff in the adult entertainment industry. The root of the issue is that payments are hard when there is a very high percentage of refunds.

How to upper-bound the probability of something bad (2018) - this is mostly useful for people who need to write proofs as part of their day to day work. Still, it’s a good refresher for some normal techniques (Markov’s Inequality, Chebyshev’s Inequality etc), which one might find in non-advanced texts. I also find it a good peek into the mind of somebody doing high-level research. And even they have to look stuff up on Wikipedia :)

Enhancing the quality of Uber’s maps with metrics computation (2018) - this is something quite relevant to the sort of stuff we do at Taxify (btw, we’re hiring), and for many issues Uber already did it. The base idea is how to ensure that the geographic data used is actually high quality and corresponds to reality. A lot of nifty data processing is used for this.

An introduction to hashing in the era of machine learning (2018) - some time ago a nice paper from Google and MIT circulated where they “learned” an index structure for database access, which was competitive with B-Trees and their variants. This was somehow expected (apply ML to everything!), but from another POV it came out of the blue. That’s CS101 stuff right there - the stuff which is supposed to be airtight - well understood problems, good algorithms, theorems etc. None of that pesky perception or reasoning stuff. And here ML was competing with it. Big news and a lot of people got excited. The present article deals with how hashing would be affected, given that it’s also the same sort of lookup problem ML was good at attacking. The bottom line was that learning to hash is good, but Cuckoo hashing is better. Still, super nice read.

Becoming a dramatically better programmer (2017) - sage advice and some directly applicable skills. In my view, the biggest thing is a mindset of trying to become better and holding yourself accountable to that, rather than coasting along. After that, each person will have their own way of doing things. Biggest thing here is how good deep work really is. I need to get to Deep Work - The Book at some point as well.

Posted on by:

horia141 profile

Horia Coman

@horia141

Leading the Bolt teams in Bucharest. We're working on cool products like food delivery, and interesting and challenging platforms like geo, A/B testing, user accounts and route tracking.

Discussion

markdown guide