Document that default offset is Model 100, not Tandy 200

Author: hackerb9

README.md

@@ -25,12 +25,12 @@ single byte. Not only is this more compact, but it loads much faster.
 Programs for the Model 100 are generally distributed in ASCII format,
 but that has two downsides: ① the user must LOAD and re-SAVE the file
 on their machine to tokenize it as only tokenized BASIC can be run and
-② the machine may not have enough storage space for the tokenized
-version while the ASCII version is also in memory.
+② the machine may not have enough storage space to tokenize if the
+ASCII version is also in memory.
 
 ### The solution
 
-This program solves that problem by tokenizing on the host computer
+This program solves that problem by tokenizing on a host computer
 before downloading to the Model 100. Additionally, this project
 provides a decommenter and cruncher (whitespace remover) to save bytes
 in the tokenized output at the expense of readability.
@@ -90,7 +90,7 @@ Flex creates the file lex.tokenize.c from m100-tokenize.lex. The
 
 ## Usage
 
-One can either use the tokenize wrapper or run the executables manually.
+One can either use the `tokenize` wrapper or run the executables manually.
 
 ### The tokenize wrapper
 
@@ -133,12 +133,12 @@ m100-decomment --> m100-crunch --> m100-tokenize
 m100-decomment --> m100-tokenize
 ```
 
-| Programs used                                                                   | Effect                                     | Same as     |
-|---------------------------------------------------------------------------------|--------------------------------------------|-------------|
-| m100-sanity<br/>m100-tokenize                                                   | Identical output as a genuine Model 100    | tokenize    |
-| m100-sanity<br/>m100-jumps<br/>m100-decomment<br/>m100-tokenize                 | Saves RAM by removing unnecessary comments | tokenize -d |
-| m100-sanity<br/>m100-jumps<br/>m100-decomment<br/>m100-crunch<br/>m100-tokenize | Saves even more RAM, removing whitespace   | tokenize -c |
-| m100-tokenize                                                                   | Abnormal code is kept as is                |             |
+| Programs used                                                                   | Effect                                         | Same as     |
+|---------------------------------------------------------------------------------|------------------------------------------------|-------------|
+| m100-sanity<br/>m100-tokenize                                                   | Identical byte-for-byte to a genuine Model 100 | tokenize    |
+| m100-sanity<br/>m100-jumps<br/>m100-decomment<br/>m100-tokenize                 | Saves RAM by removing unnecessary comments     | tokenize -d |
+| m100-sanity<br/>m100-jumps<br/>m100-decomment<br/>m100-crunch<br/>m100-tokenize | Saves even more RAM, removing whitespace       | tokenize -c |
+| m100-tokenize                                                                   | Abnormal code is kept as is                    |             |
 
 <details><summary>Click to see more details about running these
 programs manually</summary><p>
@@ -264,6 +264,17 @@ work for any of them. However, the NEC family of portables -- the
 PC-8201, PC-8201A, and PC-8300 -- run N82 BASIC which has a different
 tokenization format.
 
+### Checksum differences are not a compatibility problem
+
+The .BA files generated by `tokenize` aim to be exactly the same, byte
+for byte, as the output from tokenizing on a Model 100. The Tandy 200
+stores the first BASIC program in RAM at a slightly different location
+(0xA000 instead of 0x8000). This has no affect on compatibility, but
+it does change the line number pointers in the .BA file. The pointers
+saved in the file are _never_ used as they are recalculated when the
+program is loaded into RAM.
+
+
 ## Why Lex?
 
 This program is written in