strategy: define w (as w(P) rather than w(H)

[topbloke-formulae.git] / strategy.tex

diff --git a/strategy.tex b/strategy.tex
index 97c41b9a9b0a1e55ea45e085746a3909f08b317b..2f676bb7808acc38e7e6033cf437d260fd6a178f 100644 (file)
--- a/strategy.tex
+++ b/strategy.tex
@@ -54,6 +54,13 @@ $\bigforall_{E \in \pendsof{\set X}{\p}} E \le T$
 $\bigcup_{\p \in \allpatches} \set H^{\pn} \cup \set H^{\py}$.
 All the input commits to the update algorithm.  (See below.)
 
+\item[ $\set H^{\pc^{_=/-}}$ ]
+
+The existing head commit(s) $\set H$ of the branch $\pc^{+/-}$.
+These are the heads which will be merged and used in this update.
+This will include the current local and remote git refs, as desired.
+Obtained from the function $h$ (see below).
+
 %\item[ $\set E_{\pc}$ ]
 %$ \bigcup_i \pendsof{S_{\pc,i}}{\pc} $.
 %All the ends of $\pc$ in the sources.
@@ -76,8 +83,12 @@ all of its dependencies will be updated.
 
 \item[ $h : \pc^{+/-} \mapsto \set H^{\pc^{+/-}}$ ]
 Function for getting the existing heads $\set H$ of the branch $\pc^{+/-}$.
-These are the heads which will be merged and used in this update.
-This will include the current local and remote git refs, as desired.
+
+\item[ $w : \pc^{+/-} \mapsto \set W_0^{\pc^{+/-}}$ ]
+
+Function for getting the existing local head of the branch
+$\pc^{+/-}$.  I.e., the current value of the branch ref for $\pc^{+/-}$.
+$W_0^{\pc^{+/-}} \in \set H$.
 
 \item[ $g : \pc, \Gamma \mapsto \Gamma'$ ]
 Function to allow explicit adjustment of the direct dependencies
@@ -107,31 +118,20 @@ The new tips of the git branches $\pcn$ and $\pcy$, containing
 all the appropriate commits (and the appropriate other patches),
 as generated by the Traversal phase of the update algorithm.
 
-\end{basedescript}
-
-\stdsection{ WIP tip satisfaction, reachable commits }
-
-Set of all reachable commits, O.
+\item[ $\allreach$ ]
+The set of all reachable commits.
 
 A reachable commit is one we might refer to explicitly in any of these
 algorithms, and any ancestor of such a commit.  We explicitly
-enumerate all of the input commits (U), so the reachable commits are
-originally the input commits and their ancestors.  Each commit we
-generate will have reachable commits as ancestors so when we generate
-a new commit we only add that new commit to O.
+enumerate all of the input commits ($\allsrcs$), so the reachable
+commits are originally the input commits and their ancestors.
 
-So O varies over time as we generate more commits.  We write O_py for
-the set of reachable commits at the point where we have just generated
-T_py, ie jusjt after Merge-Tip(P).
-
-We preserve/ensure
-  Tip_Py >= pendsof( O_py, Py )
-(Tip_py is computed during traversal for the patch P)
-
-We ensure this property by:
-  - we do not generate any commits for py other than
-    during Merge-Tip
-  - so at the start of Merge-Tip pendsof (O, py) = pendsof (U, py)
-  - Merge-Tip itself wip wip wip
+$\allreach$ varies over time as we generate more commits.  Each
+commit we generate will have only reachable commits as ancestors, so
+generating a new commit (only) adds that new commit to $\allreach$.
 
+\item[ $\allreachof{\py}$ ]
+The set of reachable commits at the point where we have just generated
+$\tippy$, i.e. just after $\alg{Merge-Tip}(\p)$.
 
+\end{basedescript}