[pypy-commit] extradoc extradoc: latest version
arigo
noreply at buildbot.pypy.org
Mon Nov 14 12:38:11 CET 2011
Author: Armin Rigo <arigo at tunes.org>
Branch: extradoc
Changeset: r3966:37cc0ee778a9
Date: 2011-11-14 12:37 +0100
http://bitbucket.org/pypy/extradoc/changeset/37cc0ee778a9/
Log: latest version
diff --git a/talk/fscons2011/author.latex b/talk/fscons2011/author.latex
--- a/talk/fscons2011/author.latex
+++ b/talk/fscons2011/author.latex
@@ -5,4 +5,4 @@
{Armin Rigo}
\institute{FSCONS 2011}
-\date{November 13 2011}
+\date{November 13, 2011}
diff --git a/talk/fscons2011/example/demo1.py b/talk/fscons2011/example/demo1.py
new file mode 100644
--- /dev/null
+++ b/talk/fscons2011/example/demo1.py
@@ -0,0 +1,7 @@
+
+def f(n):
+ print "running demo1..."
+ i = 0
+ while i < n:
+ i = i + 1
+ return i
diff --git a/talk/fscons2011/example/demo2.py b/talk/fscons2011/example/demo2.py
new file mode 100644
--- /dev/null
+++ b/talk/fscons2011/example/demo2.py
@@ -0,0 +1,11 @@
+
+class A(object):
+ def __init__(self, value):
+ self.value = value
+
+def f(n):
+ print "running demo2..."
+ i = A(0)
+ while i.value < n:
+ i = A(i.value + 1)
+ return i
diff --git a/talk/fscons2011/notes.txt b/talk/fscons2011/notes.txt
new file mode 100644
--- /dev/null
+++ b/talk/fscons2011/notes.txt
@@ -0,0 +1,105 @@
+
+
+thanks Laura
+
+
+
+$ pypy
+
+differences: irc topics
+prompt
+(multi-line editing)
+
+
+def f(n):
+ i = 0
+ while i < n:
+ i = i + 1
+ return i
+
+
+
+class A(object):
+ def __init__(self, value):
+ self.value = value
+
+def f(n):
+ a = A(0)
+ while a.value < n:
+ a = A(a.value + 1)
+ return a
+
+
+
+hack/3d/test6.py
+
+
+...for maximum effect;
+in truth less impressive
+on usual programs
+(but still)
+
+
+<slides>
+
+gitdm
+ - data mining tool
+ - reads the output of
+ ``git log``
+ - generate kernel dev.
+ statistics
+ - ...3x
+
+MyHDL: VHDL-like lang
+written in Python
+now competitive with
+"real world" VHDL
+and Verilog simulators
+ - 6 to 12 times faster
+
+Random largeish program:
+ - depends on 3rd-party C
+ extensions
+ - may try to install
+ them for pypy
+ - (pypy setup.py install)
+ - may or may not work
+
+ai: jumped 3x last week
+but in general, constant
+slow(?) progress
+
+
+<slides>
+
+partially public. funded:
+ - EU, EU countries
+ - still open source
+ "mindset" at its core
+
+Open Source
+ - not GPL
+ - like CPython
+ - intense discussion of
+ about 15 seconds
+
+
+<slides>
+
+Java or .NET
+ - large pieces of codes
+ no direct control
+ - designed and optimiz.
+ for some class of
+ languages != Python
+
+
+<slides>
+
+CPython 2.7 "deprecated"
+
+PyPy 1.x won't be anytime
+soon
+
+future: support both
+PyPy (2.x) and PyPy3
diff --git a/talk/fscons2011/talk.rst b/talk/fscons2011/talk.rst
--- a/talk/fscons2011/talk.rst
+++ b/talk/fscons2011/talk.rst
@@ -9,20 +9,19 @@
--------
+Speed
+---------
+
+.. image:: speed.png
+ :scale: 45%
+ :align: center
+
Speed
---------
.. image:: progress.png
- :scale: 40%
- :align: center
-
-
-Speed
----------
-
-.. image:: speed.png
- :scale: 40%
+ :scale: 50%
:align: center
@@ -61,7 +60,7 @@
* It is easy to implement a new language with PyPy
-* Better suited to dynamic languages
+* Suited for *dynamic* languages (preferrably)
|pause|
@@ -109,7 +108,7 @@
* Pyrolog, a Prolog interpreter, is fast too
-* Haskell and a number of other experiments
+* Haskell, GameBoy, ...
|pause|
@@ -121,7 +120,7 @@
* Tracing JIT Compiler
-* Not unlike TraceMonkey for JavaScript in FireFox
+* Not unlike TraceMonkey for JavaScript in Firefox
* But two levels
@@ -168,6 +167,38 @@
PyPy 1.x <------> PyPy3 1.x
+PyPy's future?
+--------------------
+
+.. sourcecode:: plain
+
+ CPython 2.7 -------> CPython 3.x
+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ ^ written in C ^
+ | |
+ | |
+ | |
+ V V
+
+ PyPy 1.x <------> PyPy3 1.x
+
+
+PyPy's future?
+--------------------
+
+.. sourcecode:: plain
+
+ CPython 2.7 -------> CPython 3.x
+
+ ^ ^
+ | |
+ | |
+ | written in |
+ V Python 2.5-7 V
+ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+ PyPy 1.x <------> PyPy3 1.x
+
+
Contacts, Q/A
--------------
More information about the pypy-commit
mailing list