CraigRo

Should have done this 11 months ago

I deduct 2x the added charges from the usual and customary tip and leave a note on the receipt. The goal is to incentivize the servers to agitate against the restaurant.

I particularly detest sneaky 'charges' that are not embedded in the prices.

She's allowed to make money legitimately until she exhausted her appeals. She still owes a lot in restitution

The notes in the Oxford study bible are a good introduction. Lots of college courses cover this

The idea is good. But there are already plenty of companies building modular homes in factories, and rvs in factories. They have been doing this for 60 years.

This startup could have focused on design sales permitting etc and built a couple as test, but it seems vastly more efficient to contract with an existing player than to try to everything from first principles

So many people here do not know the details of NYC rent regulation, and are posting generalizations and falsehoods.

Basically, you cannot raise rents aside from annual increases that are below inflation, and you are limited to a relatively small amount of renovations, and cannot increase the rents much if you do.

So you either commit to a potentially multi lifetime below market tenancy that will bankrupt you eventually, or you keep it vacant and hope to demolish it or combine units or pray that the law will change.

Or sell it to a guy who will run air bnbs until the city puts him out of business at which point it will go in the tax lien sale. In any event, nobody is investing in bringing these units back online in this legislative environment.

There are different kinds of MS degrees. Some take people with no real cs knowledge, and some are stepping stones to a PhD. The latter is a very valuable credential, as it indicates deep technical and theoretical cs knowledge. The interview should be able to elicit this

I reviewed an oreilly book. The process was well run, and most of my suggestions were acted upon by the editors/authors. Everyone was quite professional.

Compensation was de minmus, plus a copy of the book.

I strongly suspect that tech books do not sell as well as they did 25 years ago, and most tech books make little if any profit. There's an advantage to the publisher in having a wide variety of titles for a subscription service (I loved having a safari subscription), but, at least for me, the days of buying a giant and expensive tech book for just a few relevant chapters are over.

So the addressible market is smaller, and compensation is generally lower, though there are still indirect rewards

He's generally avoided polemicists in my experience.

According to DeWitt himself, the DeWitt clause came about because he figured out that Oracle had no query optimizer.

However, the response of a lot of database vendors was to then include precomputed results for entire benchmarks, and also to have code that tried to figure out if you were running one of them. This obviously adds no value to the user, but is great for marketing.

Arthur is a non-nonsense guy, and, according to him, there's a ton of hand-tuned assembly language code in the intepreter. To make the most of it, however, you need to know how to set up the machine properly. Most people don't, and First Derivatives and Shakti don't have enough marketing people or consultants to set up machines for test runs or to explain away bad results.

In any event, k/kdb/q are wickedly difficult languages to program, but compared to modern C++ they are far faster to program and more flexible, and compared to pandas, they are faster and more likely to not have some massive performance hit due to a scalar function. I'm sure that they miss some optimizations that a good C++ compiler targeting a processor with some of the newer instructions can do, but they allow you to write fairly performant code that operates on vectors relatively easily.

Conway was a legend in the math department. Possibly because he actually lived there -- at one point he got kicked out of his house, and took the office opposite his as a bedroom. A few months later, an artist showed up to paint portraits of famous mathematicians around the border of the ceiling.

Apart from that, the login to his computer required calculating the correct day of week of 10 dates in the last 2000 years. I believe that the record was around 10 seconds. He showed me the algorithm (which isn't actually that hard), but he had a calculating ability that was truly impressive.

He also bet his salary on a series of backgammon games that were played in the common room -- he won, but just barely.

Apart from that, he'd sometimes grab an undergraduate and give them a very brief tutorial on cutting edge mathematics just so he could try to explain a new theorem or paper -- he could be amazingly generous.

On the other hand, he could be astoundingly mercurial -- there was a rumor that he used the stair method for grading, or that the grade was proportional to the length of the answer -- he loved math so much that he couldn't conceive of why anyone cared about their grade in his classes.

Amazingly beloved by everyone who came in contact with him.