[pypy-svn] extradoc extradoc: minor language rephrasing

[lac]

Author: Laura Creighton <lac at openend.se>
Branch: extradoc
Changeset: r3448:903e1f1c541e
Date: 2011-03-30 16:19 +0200
http://bitbucket.org/pypy/extradoc/changeset/903e1f1c541e/

Log:	minor language rephrasing

diff --git a/talk/icooolps2011/paper.tex b/talk/icooolps2011/paper.tex
--- a/talk/icooolps2011/paper.tex
+++ b/talk/icooolps2011/paper.tex
@@ -839,8 +839,8 @@
 decimal benchmark\footnote{\texttt{http://speleotrove.com/decimal/telco.html}},
 using a pure Python decimal floating point implementation. The results we see in
 these two benchmarks seem to repeat themselves in other benchmarks using
-object-oriented code, for purely numerical algorithms the speedups are a lot
-lower.
+object-oriented code; for purely numerical algorithms the speedups are 
+significantly smaller.
 
 The benchmarks were run on an otherwise idle Intel Core2 Duo P8400 processor
 with 2.26 GHz and 3072 KB of cache on a machine with 3GB RAM running Linux
@@ -849,8 +849,8 @@
 CPython 2.6.6\footnote{\texttt{http://python.org}}, which uses a bytecode-based
 interpreter. We compare it against four versions of PyPy's Python interpreter,
 all of them with JIT enabled. The PyPy baseline does not enable maps or type
-versions. Then we have a version each where maps and versions are enabled alone
-and finally a version with both.
+versions. We then benchmarked PyPy, first using each technique separately,
+and finally using both together.
 
 All benchmarks were run 50 times in the same process, to give the JIT time to
 produce machine code. The arithmetic mean of the times of the last 30 runs were
@@ -859,12 +859,12 @@
 reported in Figure~\ref{fig:times}.
 
 Versioned types speed up both benchmarks by a significant factor of around 7.
-The speed advantage of maps alones is a lot less clear. Maps also have a memory
+The speed advantage of maps alone is a lot less clear. Maps also have a memory
 advantage which we did not measure here. By themselves, maps improved the
 Richards benchmark slightly, but made the Telco benchmark slower. Enabling both
 maps and versioned types together yields a significant improvement over just
 versioned types for Richards. XXX good explanation. For Telco, enabling both
-does not change much.
+has little effect over the gains for versioned types alone.
 
 \begin{figure}
 {\footnotesize