383 karma · joined June 16, 2013
For a hacker like the author, anything is a good learning practice and thus good time spent.
That being said, you have a good point -- For people really looking for a serious alternative to computers, they shouldn't look too deep into Xbox 360. Thanks for putting the warning about RROD here. You've added a valuable warning sign to the original content.
> You will not use, copy, modify, merge, publish, distribute, reproduce, or create derivative works of the Software, in whole or in part, for any commercial, military, or illegal purposes.
I'm probably best known as a writer (https://lmy.medium.com), though.
Better in terms of? Style? Performance?
I've moved everything out of the paywall yesterday.
I write about tech & economics, productivity & best practices, time & knowledge management, interesting analogies, and occasionally about the comedy club I help organize.
Yes. Stand-up comedies are also huge in China:
- "Single-standing Comedy" in Beijing: http://www.danlirencomedy.cn/ and https://www.youtube.com/channel/UCeN0myzkjm2BDjaBZ5C3Qng
- "No-name Comedy" in Nanjing: https://www.wumingxiju.com/
- Fun Factory in Shanghai: https://www.linkedin.com/company/%E4%B8%8A%E6%B5%B7%E7%AC%91...
Cunningham's law at work.
# For those who end up going home
For those who end up going home, a traditional queue is still more efficient.
In a Landsburg's queue, if you decided to go home, it must be that there has been too many people joined the queue ("the queue got full"). *Their arrival takes time*. You would have waited for quite some time when you decided to bail out.
A traditional queue, on the other hand, allows those who end up going home to choose to do so immediately upon arriving at the scene. You arrive at the scene, look at the queue, judge that it would take unacceptably long period of time before you can get served ("it's currently full"), and decide to go home. You never join the queue. You wait for zero seconds.
# Arranging early arrivals is important *Filtering for fans.* In the book _The Armchair Economist: Economics and Everyday Life_, Landsburg himself stated that some businesses would prefer to serve fanatic patrons over mediocre customers. In an ordinary queue, customers can prove their loyalty by *arranging to arrive early*. On the other hand, by attaching the probability of getting served directly to the probabilistic nature of when one arrives, a Landsburg's queue does not allow/support proving patronage with early arrivals.
*An ordinary queue is an auction.* One of the main arguments that Landsburg proposed is that an ordinary, first-in-first-out queue is inefficient because it is not a market. I disagree to some extent. Let's re-examine the act of arranging early arrivals. By arriving early, you sacrifice the liberty of arranging the day's tasks liberally, often risking postponing a couple of other chores to the next day and/or losing opportunities for spending the morning somewhere else. This *cost of flexibility and opportunity* is a bid[^1] for guarantee/priority of service. Yes, it's still not a market where you trade spots in the queue, but its auction-like nature makes it a much closer implementation than Landsburg's queue.
[^1]: Well, it's technically a payment, since you don't get your money back should you lose the auction (i.e., decide to bail out from the queue). However, since it's not a game of chance, I can't call it a gamble either.
# For those who end up going home
For those who end up going home, a traditional queue is still more efficient.
In a Landsburg's queue, if you decided to go home, it must be that there has been too many people joined the queue ("the queue got full"). *Their arrival takes time*. You would have waited for quite some time when you decided to bail out.
A traditional queue, on the other hand, allows those who end up going home to choose to do so immediately upon arriving at the scene. You arrive at the scene, look at the queue, judge that it would take unacceptably long period of time before you can get served ("it's currently full"), and decide to go home. You never join the queue. You wait for zero seconds.
# Arranging early arrivals is important
*Filtering for fans.* In the book _The Armchair Economist: Economics and Everyday Life_, Landsburg himself stated that some businesses would prefer to serve fanatic patrons over mediocre customers. In an ordinary queue, customers can prove their loyalty by *arranging to arrive early*. On the other hand, by attaching the probability of getting served directly to the probabilistic nature of when one arrives, a Landsburg's queue does not allow/support proving patronage with early arrivals.
*An ordinary queue is an auction.* One of the main arguments that Landsburg proposed is that an ordinary, first-in-first-out queue is inefficient because it is not a market. I disagree to some extent. Let's re-examine the act of arranging early arrivals. By arriving early, you sacrifice the liberty of arranging the day's tasks liberally, often risking postponing a couple of other chores to the next day and/or losing opportunities for spending the morning somewhere else. This *cost of flexibility and opportunity* is a bid[^1] for guarantee/priority of service. Yes, it's still not a market where you trade spots in the queue, but its auction-like nature makes it a much closer implementation than Landsburg's queue.
[^1]: Well, it's technically a payment, since you don't get your money back should you lose the auction (i.e., decide to bail out from the queue). However, since it's not a game of chance, I can't call it a gamble either.
Website: https://ggtkx.org
I myself sometimes participate their weekly open mic sessions.
Thank you for your questions. They have been very provoking.
We are looking for a full-time Data Scientist to join us at Wharton Research Data Services (WRDS), a Wharton department that provides business intelligence, data analytics, and research support to academic institutions around the world.
## Technical Details
· Programming languages: Python! You will make use of the `pandas` module extensively in daily work. Knowing how to parallelize your computation is a bonus (we work on a 40-core server). It would help to explain your work to your fellow colleagues if you know SAS, but it is not a required skill.
· Machine learning: "Shallow" learning techniques (such as SVM classifiers with `scikit-learn`) would help a lot, while neural network packages (such as Tensorflow/Keras) would be an overkill.
· Environment: All our R&Ds are performed on a Linux server. You need to be comfortable with terminal access, Linux commands, SSH tunneling, and package management with `conda`. Our filesystem is NFS, so knowledge of optimizing IO for cache utilization would be great.
· Delivery: You need to be good at clearly summarizing your work by writing reports and documentations. By "clear", we mean "easily understandable by financial analysts who have little CS background". In terms of format, Word is good, but LaTeX or Markdown would be a delight. You will also need to pack your derived datasets and/or codes in a portable way, so that our data team can add them to our client-facing database. You can choose your own version control solution.
· Computational resources available: We have Jupyter Lab deployed on our internal R&D server, as well as a huge SAS cluster shared with our clients. Most of our numerical data are on PostgreSQL. Feel free to set up our own MongoDB, etc., if needed.
## Preferred Background Knowledge in Finance
We are in the School of Finance, after all. While not required, these background and experience will be preferred:
· experience with business/financial/accounting analytics based on large datasets,
· knowledge in finance,
· good at foosball,
· experience working with SEC data (including textual filings and numerical data), and
· experience with financial databases (e.g. CUSIP, CRSP, and Compustat).
## Apply
If you are interested, please apply through Penn Human Resources at https://upenn.dejobs.org/philadelphia-pa/business-systems-an... .