Comprehensive Technical Guide: Setting Python 3.5.2 as Default Version on CentOS 7

The Importance and Challenges of Python Version Management

In Linux system environments, particularly Red Hat-based CentOS 7 distributions, Python version management presents a technical challenge requiring careful handling due to its critical role as a dependency for both system tools and applications. CentOS 7 ships with Python 2.7 as the default installation, and essential system utilities like the yum package manager depend on this specific version. When users need to upgrade to Python 3.x series, directly replacing the system's default Python interpreter can break core system functionality, which represents the central technical challenge addressed in this paper.

Problem Scenario Analysis

The user attempted to set Python 3.5.2 as the default version by modifying symbolic links, executing the following command sequence:

mv /usr/bin/python /usr/bin/python-old
sudo ln -fs /usr/bin/python3 /usr/bin/python

While this direct replacement of the /usr/bin/python symbolic link appears straightforward, it immediately disrupts the system toolchain. When attempting to run the yum command, the system returns the error:

-bash: /usr/bin/yum: /usr/bin/python: bad interpreter: No such file or directory

This error indicates that the shebang line (#!/usr/bin/python) in the yum script now points to an incompatible Python 3 interpreter, while yum and its dependent libraries are specifically written for Python 2.7. The deeper issue is that numerous other system tools and scripts may similarly rely on Python 2.7-specific syntax and libraries, making simple version replacement trigger cascading failures.

Solution One: Precise Symbolic Link Correction

To address the above issue, the first improved approach ensures the symbolic link points specifically to the Python 3.5.2 executable rather than the generic python3 link. The correct command should be:

sudo ln -fs /usr/bin/python3.5 /usr/bin/python

This method is more precise than using python3 because python3 might be a symbolic link pointing to other Python 3.x versions (such as Python 3.6 or 3.7), while python3.5 explicitly specifies the desired version. However, this approach still suffers from a fundamental flaw: it breaks system tools' dependency on Python 2.7. Even if the corrected link allows yum to locate an interpreter, incompatible syntax and APIs in Python 3 will still cause runtime errors.

Solution Two: User-Level Alias Configuration

A safer alternative involves configuring Python aliases at the user level without affecting global system settings. By adding an alias to the user's ~/.bashrc file, version switching can be implemented to affect only the current user session:

alias python="/usr/bin/python3.5"

This method leverages the bash shell's alias mechanism. When the user types the python command in the terminal, the shell expands it to /usr/bin/python3.5. The advantages of this approach include:

• Non-invasiveness: Does not affect other system users or system-level scripts
• Reversibility: Can be reverted by editing the .bashrc file or starting a new shell
• Flexibility: Allows configuration of different Python versions for different projects

After configuration, the bash settings need to be reloaded or a new terminal session started:

source ~/.bashrc

The limitation of this method is that it only affects the python command in interactive shells, not shebang lines in scripts or Python interpreters invoked through other means.

Solution Three: Alternatives System Management Tool

CentOS provides a more professional version management tool called alternatives, which allows system administrators to perform structured switching between multiple software versions. The complete process for managing Python versions with alternatives is as follows:

First, register Python 2.7 as an alternative with root privileges:

alternatives --install /usr/bin/python python /usr/bin/python2 50

Here, /usr/bin/python is the generic name, python is the alternatives group name, /usr/bin/python2 is the actual executable path, and 50 is the priority (higher numbers indicate higher priority).

Next, register Python 3.5.2:

alternatives --install /usr/bin/python python /usr/bin/python3.5 60

Since Python 3.5.2 has a higher priority (60) than Python 2.7 (50), the system will default to Python 3.5.2. For manual selection, run:

alternatives --config python

This command displays an interactive menu listing all registered Python versions, allowing users to select the default version by entering a number. The core advantage of the alternatives tool is that it maintains a symbolic link system that centrally manages all related links, ensuring consistency.

Best Practices with Virtual Environments

For Python development projects, the most recommended approach is using virtual environments. A virtual environment creates an isolated Python runtime containing independent interpreters, libraries, and scripts. This method completely avoids interference with the system Python version and represents the Python community's consensus best practice.

The basic command to create a virtual environment using the venv module is:

python3.5 -m venv myproject_env
source myproject_env/bin/activate

Once activated, all Python operations occur within the isolated environment, including packages installed via pip. The advantages of virtual environments include:

• Project Isolation: Different projects can use different Python and package versions
• No System Impact: Completely unaffected by the system Python environment
• Reproducibility: Environments can be precisely replicated using requirements.txt

Comparative Analysis and Recommendations

The following table compares characteristics of different Python version management methods:

For most users, a layered strategy is recommended:

1. Maintain system default Python as 2.7 to ensure system tools function properly
2. Use bash aliases or alternatives to set Python 3.5.2 for personal work
3. Create independent virtual environments for each Python project

This strategy balances system stability with development flexibility, offering a secure and efficient solution.