This issue tracker has been migrated to GitHub, and is currently read-only.
For more information, see the GitHub FAQs in the Python's Developer Guide.

classification
Title: make html in doc fails because Makefile assigns python to PYTHON
Type: Stage:
Components: Build Versions: Python 3.1
process
Status: closed Resolution: wont fix
Dependencies: Superseder:
Assigned To: Nosy List: MLModel, georg.brandl
Priority: normal Keywords:

Created on 2009-04-01 21:09 by MLModel, last changed 2022-04-11 14:56 by admin. This issue is now closed.

Files
File name Uploaded Description Edit
unnamed MLModel, 2009-04-02 00:24
unnamed MLModel, 2009-04-02 14:36
unnamed MLModel, 2009-04-02 17:36
Messages (7)
msg85094 - (view) Author: Mitchell Model (MLModel) Date: 2009-04-01 21:09
Mac OSX, py3k 70991
    cd doc; make html
fails due to an old style except clause in Doc/tools/sphinx/__init__.py

The make works in release30-maint.

The difference is that the generated doc/Makefile in release30-maint 
assigns python2.5 to PYTHON, but the py3k doc/Makefile assigns python. I 
have things set up so that 'python' is Python 3. 

I understand how to set the variable from the command line, and I can 
build the html doc, but this should be fixed, either by continuing to 
specify python2.5 in the sphinx Makefile or by changing __init__.py and 
whatever other code uses exception binding to the new syntax.
msg85113 - (view) Author: Georg Brandl (georg.brandl) * (Python committer) Date: 2009-04-01 23:11
We don't recommend calling Python 3.x "python".  In fact, the executable
will always be called "python3" in Python 3.1+.
msg85122 - (view) Author: Mitchell Model (MLModel) Date: 2009-04-02 00:24
OK, makes sense. Just out of curiosity where does a user find that 
information? a developer?

There is still an inconsistency between the Makefiles in the two 
versions. That seems dangerous.
-- 
-- 

         --- Mitchell
msg85147 - (view) Author: Georg Brandl (georg.brandl) * (Python committer) Date: 2009-04-02 02:59
Why is a change to the Makefile an "inconsistency"? Anyway, once 3.1 is
released the naming scheme will be announced.
msg85204 - (view) Author: Mitchell Model (MLModel) Date: 2009-04-02 14:36
It strikes me as an inconsistency because I wouldn't expect the build 
process to differ from Python 3.0 to 3.1 except for fixing bugs.  A 
developer, or someone trying to keep up with the most recent 
versions, who had been successfiully making the html doc in a 
subversion copy of Python 3.0 and wanted to move over to Python 3.1 
finds this surprise if the command python is bound to Python 3. I 
grant you that this is a lot of ifs, and a transitional issue if 
we're all going to call Python 3 python3.

It might even be worth mentioning in the build instructions that 
"python" should be bound to a Python 2 executable, not Python 3, 
because a few things, such as make html in doc, use tools that 
haven't been updated.

Not a big deal. Just fighting for the "Principle of Least Surprise", 
as in many of  the documentation issues I've submitted. My GUI 
background showing.
-- 
-- 

         --- Mitchell
msg85208 - (view) Author: Georg Brandl (georg.brandl) * (Python committer) Date: 2009-04-02 15:06
Even Python 3.0 didn't install its built executable as "python".  We
just made that decision, which was done for Python 3.0 because it
couldn't be considered stable, permanent.
msg85227 - (view) Author: Mitchell Model (MLModel) Date: 2009-04-02 17:36
OK
-- 
-- 

         --- Mitchell
History
Date User Action Args
2022-04-11 14:56:47adminsetgithub: 49908
2009-04-02 17:36:45MLModelsetfiles: + unnamed

messages: + msg85227
2009-04-02 15:06:16georg.brandlsetmessages: + msg85208
2009-04-02 14:36:49MLModelsetfiles: + unnamed

messages: + msg85204
2009-04-02 02:59:19georg.brandlsetmessages: + msg85147
2009-04-02 00:24:32MLModelsetfiles: + unnamed

messages: + msg85122
title: make html in doc fails because Makefile assigns python to PYTHON -> make html in doc fails because Makefile assigns python to PYTHON
2009-04-01 23:11:19georg.brandlsetstatus: open -> closed

nosy: + georg.brandl
messages: + msg85113

resolution: wont fix
2009-04-01 21:09:52MLModelcreate