[issue3393] `cd Mac make installmacsubtree` fails on Mac OS X 10.5 because of r63955

2008-07-18 Thread Ronald Oussoren
Ronald Oussoren [EMAIL PROTECTED] added the comment: The current trunk should be correct, there are no explicit calls to the arch command left, all go through the configure replacement magic. To test if everything is correct now someone should test the following scenarios: * Build

[issue3393] `cd Mac make installmacsubtree` fails on Mac OS X 10.5 because of r63955

2008-07-18 Thread Ronald Oussoren
Ronald Oussoren [EMAIL PROTECTED] added the comment: BTW. There is a --with-framework-name argument to configure that makes it possible to have several indepenend framework installations ___ Python tracker [EMAIL PROTECTED] http://bugs.python.org/issue3393

[issue3393] `cd Mac make installmacsubtree` fails on Mac OS X 10.5 because of r63955

2008-07-18 Thread Trent Mick
Trent Mick [EMAIL PROTECTED] added the comment: The current trunk should be correct, there are no explicit calls to the arch command left, all go through the configure replacement magic. $ svn info Path: . URL: svn+ssh://[EMAIL PROTECTED]/python/trunk/Mac Repository Root:

[issue3393] `cd Mac make installmacsubtree` fails on Mac OS X 10.5 because of r63955

2008-07-17 Thread Trent Mick
Trent Mick [EMAIL PROTECTED] added the comment: Similar change in Mac/IDLE/Makefile.in: - --- python/trunk/Mac/IDLE/Makefile.in (original) +++ python/trunk/Mac/IDLE/Makefile.in Thu Jun 5 14:58:24 2008 @@ -42,7 +42,7 @@

[issue3393] `cd Mac make installmacsubtree` fails on Mac OS X 10.5 because of r63955

2008-07-17 Thread Trent Mick
Trent Mick [EMAIL PROTECTED] added the comment: Alternative potential solution: use the ARCHPREFERENCE environment variable as described in the Mac OS X 10.5 arch man page. Ronald, if you could test if that works for you on 10.5, then presumably setting that environment var would be safely

[issue3393] `cd Mac make installmacsubtree` fails on Mac OS X 10.5 because of r63955

2008-07-17 Thread Ronald Oussoren
Ronald Oussoren [EMAIL PROTECTED] added the comment: This patch is wrong, it drops the call to 'arch' entirely even when the call is needed. The suggestion for ARCHPREFERENCE won't work though, arch doesn't take arguments at all on 10.4. BTW. This is a duplicate of issue 3381, which