HN user
The quality of the text rendering could really be improved by using subpixel render https://en.wikipedia.org/wiki/Subpixel_rendering. Most desktop apps do this. Your browser does it. Your terminal probably does it. VSCode does it. Even the canvas tag in browsers supports this.
Basically, every place you're likely to read code is going to look better than the images generated with this.
They provide antifraud protection via chargebacks, which account for a sizeable portion of their fees. Fees are about 3%, fraud is about 1-2%.
I'll also plug my own canvas rendering 3d library: http://seenjs.io/
There is scope for considerable refinement in the choice of the second factor p(life|HZ). [...] In our simplified treatment, this constant value has no effect on dP(t)/dt since its contribution is also cancelled by the normalization factor N.
Though this paper does provide some needed formalism around the Drake Equation, I don't understand how its results can be meaningful if it "cancels out" one of the largest sources of uncertainty.
The NSA ANT catalog of spy devices (leaked by Snowden) also includes cheap analog retro-reflectors: https://en.wikipedia.org/wiki/NSA_ANT_catalog
This is pretty incredible, and congrats to the research team!
But, could you please not use phrasing like "Reduces Power 10,000x". I don't think I'm the only that has an aversion to this: http://timesless.com/
Neato! Verified with:
echo "H4sICBPHnlYCA2EuYmluegDt0LENQQEUQNHnK7xINF8rLIAVbKH88VU0ZlATK4gt1DQ6vURMoBIDiN4CknOa29/BNrN72U+LZj2eL1fxPDU6EXE+3CbvUTEMAAAAAAAAAODfPfq9/DZfx6petw2B32Z5XZT31mYXH9vDv0RTIwAA" | base64 --decode | gzip -d | gzip -d | xxd"It needs to be available for research, to be put on display."
It seems like this artifact studying Dr. Jones is really trying to avoid saying "It belongs in a museum!"
If this affects only the most distant observations, I wonder if this can already be detected in our observations of the CMB.
I appreciate the minimal interface, but a little CSS could go a long way to making this look a lot nicer.
Sans serif fonts and some margins and spacing would help.
How does this differ from the other javascript NaCl implementations?
Sorry, there was a glitch in the lighting. I fixed it. The next tier of values should now look like this:
The tiles should start to cycle through the colors again after 2048. From 2^12 - 2^22 it does triangular prisms, from 2^23 and beyond it does pyramids.
Are you only getting white shapes after 2048?
pokpokpok also suggested this. I created a branch where you can try it out yourselves!
git clone git@github.com:themadcreator/seen.git
cd seen
git checkout vary-heights
npm install
npm run site
it looks like this: http://snag.gy/RdOtD.jpg
Alright, for those of you that are experiencing nausea while playing this version of 2048, I've added a "toggle nausea" button to disable the rotation of the board.
A lot of people are commenting that the colors are too similar. This is partially due to the prismastic color scheme we chose. The specular phong shading model throws them off a bit as well. Here are the current color values:
#ffffff
#ffff00
#9aCD32
#2faa2f
#20B2AA
#1088fF
#7b68ee
#9932cc
#800080
#8b0000
#333333
I got the idea for using colors only from http://prism.clay.io/. I agree -- the colors become intuitive after just a few minutes.
Okay, I updated the colors so the dark green is darker.
There's an ordered palette of colors below that matches to the 2^N values from the original. Plus you get used to it after a couple minutes =).
Wouldn't the occlusion of the shapes behind it be confusing?
It's exactly the same under the hood as Gabriele's original 2048 from https://github.com/gabrielecirulli/2048
Cool idea! I plan on adding some rigid body physics to this library.