Added beginning draft of manual for 3.0

This is being kept in Scrivener format, for ease of writing.

1 files changed, 12 insertions, 0 deletions

diff --git a/doc/Ledger.scriv/221.rtfd/TXT.rtf b/doc/Ledger.scriv/221.rtfd/TXT.rtf
new file mode 100644
index 00000000..6e69ff52
--- /dev/null
+++ b/doc/Ledger.scriv/221.rtfd/TXT.rtf
@@ -0,0 +1,12 @@
+{\rtf1\ansi\ansicpg1252\cocoartf949\cocoasubrtf460
+{\fonttbl\f0\fmodern\fcharset0 Courier;\f1\fswiss\fcharset0 Helvetica;}
+{\colortbl;\red255\green255\blue255;}
+\pard\tx560\tx1120\tx1680\tx2240\tx2800\tx3360\tx3920\tx4480\tx5040\tx5600\tx6160\tx6720\sl264\slmult1\ql\qnatural\pardirnatural
+
+\f0\fs28 \cf0 All interactions with the Ledger library take place in the context of a @var\{Session\}.  You may create as many sessions as you like, though typically only one is needed during the lifetime of a script
+\f1\fs24 \{\\SCRV_FN=
+\f0\fs28 An example where more might be needed is a GUI program that opens multiple Ledger files in different windows, and creates reports for each file separately
+\f1\fs24 \\END_SCRV_FN\}
+\f0\fs28 .  Ledger provides a pre-initialized session named @var\{session\}, but others can still be created.  Creating a session is trivial, but before it may be used it must be passed in a call to @code\{set_session_context\}.  This is how multiple sessions are managed in a multi-threaded environment, where the session itself should be considered non-reentrant.\
+\
+The first operation after creating a session is typically to load the Ledger data you wish to report on for that session.  The loaded data can be reset at any time, and can also be modified.}
\ No newline at end of file