[pypy-commit] extradoc extradoc: be nicer to JS and update pdf

[Raemi]

Author: Remi Meier <remi.meier at gmail.com>
Branch: extradoc
Changeset: r5380:f995dc84551c
Date: 2014-07-30 20:22 +0200
http://bitbucket.org/pypy/extradoc/changeset/f995dc84551c/

Log:	be nicer to JS and update pdf

diff --git a/talk/dls2014/paper/paper.pdf b/talk/dls2014/paper/paper.pdf
index 952e31bae97ac98d48739e0a955a0662d3c64fe5..f7c23752d1005cd0ef587d90e5589777cba2a6a3
GIT binary patch

[cut]

diff --git a/talk/dls2014/paper/paper.tex b/talk/dls2014/paper/paper.tex
--- a/talk/dls2014/paper/paper.tex
+++ b/talk/dls2014/paper/paper.tex
@@ -192,10 +192,9 @@
 
 \section{Introduction}
 
-
 Dynamic languages like Python, PHP, Ruby, and JavaScript receive a lot
-of attention but have not yet embraced parallelism. A parallel
-programming model was not part of the design of those languages. Thus,
+of attention but only provide limited forms of parallelism. A parallel
+programming model was not part of the initial design of those languages. Thus,
 the reference implementations of, e.g., Python and Ruby use a single,
 global interpreter lock (GIL) to serialise the execution of code in
 threads. The use of a single global lock causes several disadvantages,
@@ -844,7 +843,8 @@
 Listing~\ref{lst:rb} is easily optimisable for compilers as well as
 perfectly predictable for CPUs. Additionally, the read barrier is not
 constrained to execute before the actual read -- both the compiler and
-the CPU are free to reorder or group them for efficiency.
+the CPU can reorder or group them freely between potential safe points
+for additional performance.
 
 \begin{code}[h]
 \begin{lstlisting}