Correct: if adding a variable for a literal used as a parameter for a function call, the argument name in the function definition is used (provided we can figure out which function was called!).
There are multiple name suggestions when:
(1) Multiple items are being replaced and each replaced item had a different name.
(2) There would be a name collision. To resolve the collision, either one or the other variable can be renamed.
Warner/Chappell has submitted a response arguing that it's not quite a smoking gun.
Copyright divestment by failing to put a proper copyright notice must happen with the "consent of the copyright owner". The song in the 1922 songbook said, "Special permission through courtesy of The Clayton F. Summy Co." without the appropriate copyright notice. The lack of proper copyright notice would have released it to the public domain (proper notice was included on other songs in the songbook). However, Warner/Chappell claims that Summy didn't own the copyright in 1922: the songbook asked the wrong people for permission.
"There is no evidence that the Hill Sisters (Jessica or Patty) granted anyone the right to publish the Happy Birthday to You! lyrics until 1935. The evidence instead shows that Summy sought and obtained a license to publish the Happy Birthday to You! lyrics from Jessica Hill in 1935."
As circumstantial evidence, Warner/Chappell also points out that in earlier court battles in the 30s and 40s the copyright holders never indicated that there was any authorized publication before 1935.
Warner/Chappell doesn't address the filmmaker's claim that a 1922 publication would put the song into the public domain in 1997. I'm confused here: Would the publication have to be authorized for the 1997 date to hold?
Most browsers these days implement window.crypto.getRandomValues() so you could use that and do the generation client-side in Javascript. That way, I don't have to trust that you're not storing these passwords in preparation for a later attack.
Author here. There are other services that do this sort of thing, but none that accomplished all these goals:
1. Browser-encrypted so the server cannot decrypt it. The decryption key is stored in the URL fragment after the # so the server never sees the decryption key. (Fragments are not sent in HTTP requests.)
2. Secrets never touch disk. SneakyNote.com even goes to lengths to clear the in-memory buffers on storage/access.
3. Verified delivery. If you're going to send the link over an insecure channel, you can't prevent someone else from getting to it first. But you'd like to know if they replaced your link. The verification code step solves this problem.
4. Open source.
Incidentally, this is written in Go because it of its good HTTP support and because Go is a memory-safe language that will also allow you to clear memory.
Police are trained only to shoot to kill because pulling the trigger needs to be a big event. If you think that "shoot to wound" is an option, then of course you're going to pull the trigger more often. You'll become more comfortable shooting people and, presumably, you'll also end up killing people more often.
Other reasons for "shoot to kill" relate to effectiveness: in the energy of the situation if you aim to wound you may either kill by accident, or not hit at all.
As I understand it, police are generally taught specifically to shoot center of mass because aiming there is the most effective way to reliably and quickly end the immediate physical threat. This also ends up pretty close to maximizing the killing effectiveness of shots, but is not the primary focus.
The reason is that you are more likely to hit the target rather than what some propose, that they should just wound him in the kneecap, an almost impossibility.
TF–IDF ranking doesn't seem to be too complex a thing to implement. Maybe this is an opportunity for someone here to contribute to the open source project.
As the IRFY creator, I freely admit that site needs some love.
Also, technically, it's a Ruby benchmark. The Rails version is going to be locked to 3.1.3 for as long as possible. The benchmark does not reflect speed improvements in the Rails codebase.
Comparing across languages was originally part of the long-term goal. However, I haven't yet come across any good, multi-language, real-world benchmark.
The variance is sometimes due to RVM troubles. The benchmark is partly a test of the entire Ruby ecosystem. If RVM is bugging, then people are installing sub-optimal implementations. For example: https://github.com/wayneeseguin/rvm/issues/1418
There are multiple name suggestions when:
(1) Multiple items are being replaced and each replaced item had a different name.
(2) There would be a name collision. To resolve the collision, either one or the other variable can be renamed.