Ah, I see. I'm no expert, but I'd argue the response is "Maybe".
I think this is what blaze and pypy try to do with python: they compile it so that certain things are optimized. The harder part is that you have to generally hint for the compiler to truly be the best...
WRT to just why not use C, python/R/Julia are MUCH easier to prototype in than C (in my opinion, having used those languages). The "grail" is to be able to prototype quickly, and then identify the hotspots to speed up (which, in Julia's case, would be just inlining with C code).
> They can 'compile' in an 'optimized manner' using something else ...
Thats actually the issue Julia attempts to solve as highlighted in the article. There are many languages that you can use to express ideas, but when you need the idea to run as fast as possible, you have to rewrite in C/Fortran/etc and then add bindings to to the language.
Julia lets you write inline C, which I believe means that you can still express algorithms simply, but with some parts (trivially) optimized. (Not a huge julia user, but thats my take on it)
NIMBY is more "Don't use my road, use the road one block over" - in other words, road is the same but not in front of me. Thats a little different than "Don't use my road, use the road that has 3 lanes and no children on it and is labeled for commuting" I feel like there is a difference in that regard; I personally wouldn't want through traffic on my street, but I also don't want it on the streets next to mine either due to how recklessly people drive when trying to beat traffic.
I've noticed though that the perspective matters. If somebody start complaining about how I parked in front of their house, a typical response is: "I can't believe how annoyed this guy is getting over me just parking here for 15 minutes."
The homeowner however doesn't care that its only 15 minutes of your time, since for him/her its probably all day :-/
The interesting thing about a 'diamond' planet is that it would be a waste of time. Diamond's value is superficial - it can't actually be used for anything (except cutting), so mining this planet would essentially cause diamonds to become worthless.
Whats more interesting is a planet made up of rare metals - that would still cause the prices to drop (this assumes the mining can easily generate lots of metal easily, transport costs aside), but can immediately be turned around into something useful.
May sound corny, but I'd have to say "Fostering." If you show interest in something, be it knowledge about a subject or field (or entrepreneurship), they will always make sure you have all the resources you need. Part of this (and has been voiced in some of the other comments) is that they will never block you from learning or exploring something.
Completely agree. The key point is not "applying" these traits, or treating them as "tricks," its simply that the professors there embody that ideal. As an undergrad, it was easy for me to grab 10 minutes of any professor's time, even if it wasn't related to their courses or study. A lot of my classes had over 300 students in them, yet that didn't keep a professor from meeting with me either.
Another parallel I've experienced with going to the gym and coding is that its far more fun doing it with somebody else. You work out harder, and push yourself in a better way than doing it by yourself - something I've found also while I code.
Any estimate as to the time of transcoded content? Also, does the term "video" mean # movies in, or # of streams (movie at certain bitrate/width/etc) out?
There are times where its handy to have readability, albeit in one-off situations.
The most relevent one I recently encountered was working with a team and needing to get some numbers crunched - I had knowledge of scipy and was able to show the code to others who hadn't used python but could still understand it.
I agree this is an issue, and given that some of my smartest peers learnt English as a second (or third) language, I definitely don't want to make it harder.
If I realize that they are not native english-speakers, then I typically give them the option to work on the code, the stop at a breakpoint and have them explain what they did and why they did it. This seems to keep undue Stress off the applicant while still letting learn about how they think.
As somebody who has done numerous technical interviews, as well as given some, I have to disagree with some of the points he makes.
1) "I don't code out loud": neither do I, nor as an interviewer would I honestly want you too in a 'real' work environment. What I do want to know (or attempt to know) is how you think. If you can articulate your thought process as you solve a question, it shows me that a) you can think logically under pressure, and b) you can convey your thoughts and ideas to somebody who might be a fellow team mate.
2) Talking about a project you have done is cool, but I personally don't care what technologies or hacks you used. I want to know if you can code. If you can't code, then it doesn't matter what you know since I can't use you on my team. If you can code, then I'll use what I learn from (1) to determine if you have the ability to choose the right algorithm/tool/technique/etc for a job.
One might retort that (2) can be easily solved by showing them project code. Unless its a github account, I don't really care, because I can't tell what you did versus somebody else and it is far to easy to control what the interviewer sees (as you showed when discussing your "team motivation" skills).
At the same time, TSC definitely made some simple mistakes in the types of questions they asked you, especially for an on-site. A proper technical question (IMO) is one that can tell me easily if it's not a fit, but has enough different solutions that let me gauge just how good a candidate is. A question that is a no if you get it wrong, but not a yes if you get it correct (hex CSS colors) would be a wast of both our time (and would probably get me kicked off the interviewing team).
As per using amazon to scale, it depends on the services they use. EC2 is expensive, but you don't need that many ec2 instances to handle the streaming, outside of permission handling etc.
As for transcoding videos, they probably have a far more burst-like intake of movies, yet not a very large burst at that. How many tv shows are released per day? How many movies come out per day? Not too many, compared to a service like youtube that has to transcode huge amounts of information all the time. EC2 at least lets them turn on machines until they finish, at which point they can easily scale down after the burst is handled.
I think this is what blaze and pypy try to do with python: they compile it so that certain things are optimized. The harder part is that you have to generally hint for the compiler to truly be the best...
WRT to just why not use C, python/R/Julia are MUCH easier to prototype in than C (in my opinion, having used those languages). The "grail" is to be able to prototype quickly, and then identify the hotspots to speed up (which, in Julia's case, would be just inlining with C code).