Start prioritizing todos into v3 and v4 stuff

I could sit here and fiddle with this shit forever,
start choosing what I actually need to release v3 so I can put it out of mind for a while.

Readme.md

@@ -99,13 +99,39 @@ Known Issues
   This code will be wrapped. But using the value is syntactically invalid in Ruby, because it constitutes a "void value expression" (aka a nightmare for anyone working with code).
   I can't easily check it to seee if it's void since it's not in the parsed AST.  But it's so edge taht I don't think it's worth worrying about.
 
-Shit that will probably never get done (or if it does, won't be until after 2.0)
-================================================================================
+Version 2
+=========
+
+Feature complete, I'll fix bugs in it until version 3 is released, though
+
+Version 3
+=========
+
+These need to be done before release:
+
+* Add default to number of captures (1000), require user to explicitly set it to infinity
+* Update changelog with all changes
+* Push the event stream up to the top level so we can expose it via the interface.
+* Get it working on JRuby and Rbx
+* Order frog stickers
+* Figure out how to handle markers
+
+Version 4
+=========
 
 * How about if begin/rescue/end was able to record the result on the rescue section
-* How about if you could configure which kinds of results ou were interested in (e.g. turn on/off recording of method definitions, and other results)
-* What about recording the result of a line inside of a string interpolation, e.g. "a#{\n1\n}b" could record line 2 is 1 and line 3 is "a\n1\nb"
-* If given a file with a Unicode character, but not set Unicode, inform the user
+* How about if you could configure which kinds of results you were interested in
+  (e.g. turn on/off recording of method definitions, and other results)
+* What about recording the result of a line inside of a string interpolation,
+  e.g. "a#{\n1\n}b" could record line 2 is 1 and line 3 is "a\n1\nb"
+  This would require smarter annotators.
+* Allow debugger to take a filename (ie debug to a file insteaad of to stderr)
+* `--cd dir` cd to that dir before executing the code
+* `--cd -` cd to the dir of the file being executed before executing it
+* `--only-show-lines` output only on specified lines (doesn't change stdout/stderr/exceptions)
+* More alignment strategies e.g. `min=40` would align to 40, unless that was too short.
+  Could have fallback strategies, so e.g. `-s min=40,fallback=line`
+* Package Ruby with the editor downloads so that they don't require you to know so fkn much to set it up.
 
 Inspiration
 ===========

lib/seeing_is_believing/binary/config.rb

@@ -35,9 +35,7 @@ def self.to_regex(string)
                    stderr:    to_regex('^#\s*!>\s*')
       end
 
-
       # passed to annotator.call
-      # TODO: move AnnotatorOptions to uhm, annotator or something
       class AnnotatorOptions < StrictHash
         attribute(:alignment_strategy) { AlignChunk }
         attribute(:markers)            { Markers.new }
@@ -46,10 +44,6 @@ class AnnotatorOptions < StrictHash
         attribute(:max_result_length)  { Float::INFINITY }
       end
 
-      # TODO: just a thought here:
-      # the binary is goind to do one of four things:
-      #   print_version, print_cleaned, print_help, and evaluate
-      #   might make sense to implement a state machine on these rather than setting all these boolean flags and having to check them down in finalize
       predicate(:print_version)       { false }
       predicate(:print_cleaned)       { false }
       predicate(:print_help)          { false }
@@ -70,17 +64,6 @@ class AnnotatorOptions < StrictHash
       attribute(:lib_options)         { SeeingIsBelieving::Options.new }       # passed to SeeingIsBelieving.new
       attribute(:annotator_options)   { AnnotatorOptions.new }
 
-      # TODO: allow debugger to take a filename
-
-      # TODO: --cd dir | --cd :file:
-      #   when given a dir, cd to that dir before executing the code
-      #   when not given a dir, cd to the dir of the file being executed before executing it
-
-      # TODO: --only-show-lines
-      #   Output only on specified lines (doesn't change stdout/stderr/exceptions)
-
-      # TODO: --alignment-strategy n-or-line / n-or-chunk / n-or-file (help-file should prob just link to cuke examples)
-      # add default to number of captures (1000), require user to explicitly set it to infinity
       def parse_args(args, debug_stream)
         as        = nil
         filenames = []