🎨 Fix spelling, whitespace

This commit is contained in:
Scott Lahteine
2022-05-31 12:43:44 -05:00
parent 7dd34848b6
commit 58ce5182c2
18 changed files with 28 additions and 28 deletions

View File

@@ -29,7 +29,7 @@
* Specifically optimized for 3D printing G-Code, this is a zero-cost data compression method
* which packs ~180-190% more data into the same amount of bytes going to the CNC controller.
* As a majority of G-Code can be represented by a restricted alphabet, I performed histogram
* analysis on a wide variety of 3D printing gcode samples, and found ~93% of all gcode could
* analysis on a wide variety of 3D printing G-code samples, and found ~93% of all G-code could
* be represented by the same 15-character alphabet.
*
* This allowed me to design a system of packing 2 8-bit characters into a single byte, assuming
@@ -38,7 +38,7 @@
*
* Combined with some logic to allow commingling of full-width characters outside of this 15-
* character alphabet (at the cost of an extra 8-bits per full-width character), and by stripping
* out unnecessary comments, the end result is gcode which is roughly half the original size.
* out unnecessary comments, the end result is G-code which is roughly half the original size.
*
* Why did I do this? I noticed micro-stuttering and other data-bottleneck issues while printing
* objects with high curvature, especially at high speeds. There is also the issue of the limited