Show HN: Logarithmic Flappy 2048 • Hacker News Board

Re: Show HN: Logarithmic Flappy 2048

#61

Post by gburt » Thu, Mar 20, 2014, 6:53 AM UTC

harmegido wrote: post #55

computer wrote: post #48

Earlier quoted context omitted.

I suspect you know this, but either way: you should initialize your arrays in R to the proper length-- dynamically growing them within a for-loop as you do with values, func and results in optimxWrapper is extremely slow in R, as it creates a new array every loop, copying the old one.

I feel like I'm failing at R if I'm using a for loop typically.

The for loop is a dirty hack to deal with a bug I didn't want to invest time in for one-off code. I wish I had cropped this out of the screenshot...

Re: Show HN: Logarithmic Flappy 2048

#62

Post by jjaredsimpson » Thu, Mar 20, 2014, 6:57 AM UTC

dllthomas wrote: post #39

albertjorlando wrote: post #38

Earlier quoted context omitted.

Well yes, logbase2(x), where x is your score, does equal the number of wall jumps. But that doesn't account for how the game is laid out. Your score is a result of the number of jumps you've made. The number of jumps you've made is not the result of your score. Therefore, despite these two functions being the inversely related, the correct relationship is 2^x. That is just my opinion though.

I agree "exponential" would have been a more natural choice. "Therefore, despite these two functions being the inversely related, the correct relationship is 2^x." There is no "correct relationship", there is a relationship that (depending on choice of variables) we can describe as y=2^x or y=lg(x).

Sure, score = 2^walls iff log_2(score) = walls. So, being equivalent statements, one alone can not be correct.

However, as you said, one choice of describing the relationship is certainly more natural. The exponential description assigns a score to every natural number. While the logarithmic description does not.

I would interpret correct in the way that it is sometimes said, "the correct way to think about X is..." which doesn't say other ways of thinking are wrong, but limited or not illuminating.

Re: Show HN: Logarithmic Flappy 2048

#63

Post by javajosh » Thu, Mar 20, 2014, 7:03 AM UTC

agildehaus wrote: post #24

davidgerard wrote: post #19

Awaiting Doge and Dr. Who versions.

50% complete: http://doge2048.com/

Cool. This version is more challenging since it's hard to remember the ordering of images. Also, I had to cheat to see the "final" image, the derp. That was a motivation, too: I wanted to see what the next image would be!

Re: Show HN: Logarithmic Flappy 2048

#64

Post by ____miah » Thu, Mar 20, 2014, 8:44 AM UTC

I guess things like this can't be avoided. anyway, great idea, but I think you can add features that let's the flying tile hit numbered tile that can get the score higher. just suggesting. or not.

Re: Show HN: Logarithmic Flappy 2048

#65

Post by TazeTSchnitzel » Thu, Mar 20, 2014, 9:12 AM UTC

shmageggy wrote: post #11

I would describe this as multiplicative rather than logarithmic. 2×2×2×2... Or maybe exponential if you consider your score to be 2^(number of walls you've jumped). I don't get logarithmic, though.

I chose the name without any thought, really.

An early version of the code did use log2() on the score internally, but now it does pow(2) externally.

Re: Show HN: Logarithmic Flappy 2048

#66

Post by gibybo » Thu, Mar 20, 2014, 10:07 AM UTC

kyro wrote: post #36

Quite a coincidence that I see this tonight. I was on the bus today and noticed someone playing a game on their phone, something simple where you have to trace over a pattern as closely as you can. It was frustrating to watch because every time the user wanted to repeat a level or move on to the next, they were forced to tap through a couple of intermediary screens with flashy stats that took up as much time to get t…

I was thinking the same thing after I failed the first time. Most people might not think of this sort of thing because 'hey, the interstitial shows you important information and it only takes a second', but the ability to just continue immediately after messing up makes the game tremendously more enjoyable.

Re: Show HN: Logarithmic Flappy 2048

#67

Post by nnnnni » Thu, Mar 20, 2014, 10:22 AM UTC

collinvandyck76 wrote: post #52

philippotto wrote: post #17

Earlier quoted context omitted.

Where can I find the any-key?

all i see is catarl, pig-up, and esc

I just used one of those little water-bird things to play the game and won!

Re: Show HN: Logarithmic Flappy 2048

#68

Post by nnnnni » Thu, Mar 20, 2014, 10:22 AM UTC

This is actually significantly better than Flappy Bird because it has ACCURATE hit detection.

Re: Show HN: Logarithmic Flappy 2048

#69

Post by Kiro » Thu, Mar 20, 2014, 10:28 AM UTC

nnnnni wrote: post #68

This is actually significantly better than Flappy Bird because it has ACCURATE hit detection.

Why do you think the hit decetion is off in Flappy Bird? I think it's very accurate and fair and that that's one of the reasons it became such a hit.

Re: Show HN: Logarithmic Flappy 2048

#70

Post by gone35 » Thu, Mar 20, 2014, 10:53 AM UTC

gburt wrote: post #2

Welp. http://cl.ly/image/13370z0l053H

Nice. I thought I was the only one using end-of-line semicolons in R! Don't listen to the haters/purists that tell you against them.

Note though you missed the first two statements of 'somaWrapper' and in the return statement of 'optimxWrapper'.