[pypy-svn] r19348 - pypy/release/0.8.x/pypy/doc
cfbolz at codespeak.net
cfbolz at codespeak.net
Tue Nov 1 15:05:56 CET 2005
Author: cfbolz
Date: Tue Nov 1 15:05:55 2005
New Revision: 19348
Modified:
pypy/release/0.8.x/pypy/doc/getting-started.txt
Log:
some small fixes to getting-started: highlight the translation with thunks a
bit more, bump version number to 0.8.0 where necessary. Furthermore py.py still
seems to be 2000 times slower, so I removed the according XXX.
Modified: pypy/release/0.8.x/pypy/doc/getting-started.txt
==============================================================================
--- pypy/release/0.8.x/pypy/doc/getting-started.txt (original)
+++ pypy/release/0.8.x/pypy/doc/getting-started.txt Tue Nov 1 15:05:55 2005
@@ -10,11 +10,7 @@
Just the facts
==============
-.. XXX:
- 0.7 -> 0.8
- still 2000 for py.py?
-
-getting & running the PyPy 0.7 release
+getting & running the PyPy 0.8 release
-----------------------------------------
*please come back tomorrow when the following information
@@ -23,20 +19,20 @@
Download one of the following release files and unpack it:
-*pypy-0.7*
+*pypy-0.8*
* download one of
- * `pypy-0.7.0.tar.bz2`_ (unix line endings) or
- * `pypy-0.7.0.tar.gz`_ (unix line endings) or
- * `pypy-0.7.0.zip`_ (windows line-endings) and unpack it
+ * `pypy-0.8.0.tar.bz2`_ (unix line endings) or
+ * `pypy-0.8.0.tar.gz`_ (unix line endings) or
+ * `pypy-0.8.0.zip`_ (windows line-endings) and unpack it
* alternatively run
- * ``svn co http://codespeak.net/svn/pypy/release/0.7.x pypy-0.7.x``
- (the 0.7 maintenance branch)
+ * ``svn co http://codespeak.net/svn/pypy/release/0.8.x pypy-0.8.x``
+ (the 0.8 maintenance branch)
-then change to the ``pypy-0.7.0`` or ``pypy-0.7.x`` directory
+then change to the ``pypy-0.8.0`` or ``pypy-0.8.x`` directory
and execute the following command line::
python pypy/bin/py.py
@@ -48,15 +44,14 @@
CPython, it runs around 2000 times slower than the
original CPython.
-However, the major news with the 0.7.0 release is
-that you can use PyPy to `translate itself to lower
-level languages`_ after which it runs standalone
-and is not dependant on CPython anymore.
+However, since the 0.7.0 release it is possible to use PyPy to `translate
+itself to lower level languages`_ after which it runs standalone, is not
+dependant on CPython anymore and becomes faster.
.. _`90% of CPythons core language regression tests`: http://codespeak.net/~hpk/pypy-testresult/
-.. _`pypy-0.7.0.tar.bz2`: http://code2.codespeak.net/download/pypy/pypy-0.7.0.tar.bz2
-.. _`pypy-0.7.0.zip`: http://code2.codespeak.net/download/pypy/pypy-0.7.0.zip
-.. _`pypy-0.7.0.tar.gz`: http://code2.codespeak.net/download/pypy/pypy-0.7.0.tar.gz
+.. _`pypy-0.8.0.tar.bz2`: http://code2.codespeak.net/download/pypy/pypy-0.8.0.tar.bz2
+.. _`pypy-0.8.0.zip`: http://code2.codespeak.net/download/pypy/pypy-0.8.0.zip
+.. _`pypy-0.8.0.tar.gz`: http://code2.codespeak.net/download/pypy/pypy-0.8.0.tar.gz
Svn-check out & run the latest PyPy as a two-liner
--------------------------------------------------
@@ -262,7 +257,8 @@
It is interesting to note that this lazy-computing Python extension
is solely implemented in a small `objspace/thunk.py`_ file consisting
-of around 100 lines of code.
+of around 100 lines of code. Since the 0.8.0 release it is even possible
+to `translate PyPy with the thunk object space`_.
Running the tests
+++++++++++++++++
@@ -337,9 +333,6 @@
>>> t = Translator(test.is_perfect_number)
>>> t.view()
-.. >>> from pypy.translator.translator import Translator
-.. >>> from pypy.translator.test import snippet as test
-
trying out the type annotator
+++++++++++++++++++++++++++++
@@ -364,9 +357,9 @@
>>> t.specialize()
>>> f = t.ccompile()
-The first command replaces operations with variables of types that are
-avaiable in C (e.g. int) with low level versions. This can also be ommited
-although it is not recommended. To try out the compiled version::
+The first command replaces the operations with other low level versions that
+only use low level types that are available in C (e.g. int). This can also be
+ommited although it is not recommended. To try out the compiled version::
>>> f(5)
False
@@ -485,15 +478,19 @@
garbage collector`_ or our own reference counting implementation
(as we have seen Boehm's collector is the default).
-.. XXX:
- --stackless ? it's own section?
-
You can also use the translate_pypy.py script to try out several smaller
programs, e.g. a slightly changed version of Pystone::
cd pypy/translator/goal
python translate_pypy.py targetrpystone
+
+.. _`translate PyPy with the thunk object space`:
+
+Translating with the thunk object space
+++++++++++++++++++++++++++++++++++++++++
+
+
It is also possible to experimentally translate a PyPy version using
the "thunk" object space::
@@ -503,6 +500,10 @@
the examples in `lazily computed objects`_ should work in the translated
result.
+Translating using the LLVM backend
+++++++++++++++++++++++++++++++++++
+
+
To create a standalone executable using the experimental LLVM_ compiler
infrastructure::
More information about the Pypy-commit
mailing list