[pypy-commit] extradoc extradoc: merge heads

antocuni noreply at buildbot.pypy.org
Fri Jan 13 15:27:14 CET 2012


Author: Antonio Cuni <anto.cuni at gmail.com>
Branch: extradoc
Changeset: r4026:56bab9df185f
Date: 2012-01-13 15:26 +0100
http://bitbucket.org/pypy/extradoc/changeset/56bab9df185f/

Log:	merge heads

diff --git a/blog/draft/tm.rst b/blog/draft/tm.rst
--- a/blog/draft/tm.rst
+++ b/blog/draft/tm.rst
@@ -82,17 +82,14 @@
 of hacks that you do and forget about.  In this case, the dependencies
 might be simpler, but you still have to learn and use subtle locking
 patterns or a complex inter-process library, which is overkill for the
-purpose.  I would even argue that this is similar to how we might feel a
-priori that automatic memory management is unnecessary in small programs
---- but of course anyone who wrote a number of 15-line Python scripts
-knows this to be wrong.  This is even *so* wrong that the opposite is
-obvious nowadays: it makes no sense whatsoever to manage object
-lifetimes explicitly in most small scripts.  A garbage collector is not
-overkill; it is part of the basics that you expect.
+purpose.
 
-(I think the same will eventually be true for using multiple CPUs, but
-the correct solution will take time to mature, like garbage collectors
-did.  This post is a step in hopefully the right direction ``:-)``)
+(This is similar to how explicit memory management is not very hard for
+small programs --- but still, nowadays a lot of people agree that
+automatic memory management is easier for programs of all sizes.  I
+think the same will eventually be true for using multiple CPUs, but the
+correct solution will take time to mature, like garbage collectors did.
+This post is a step in hopefully the right direction ``:-)``)
 
 Events in Transactions
 ----------------------


More information about the pypy-commit mailing list