Essential Python 2 vs 3 Differences Every Developer Must Know

This topic has largely quieted on forums, but Python 2 still lives in legacy internal systems, archived scripts, and un‑modernized maintenance projects.

Ask any operations or QA person and you’ll likely see Python 2 on servers or buried in deployment scripts. The language is officially "ended," yet its impact persists.

If you work seriously with Python, knowing the main differences between Python 2 and Python 3 is useful because you may need to touch both.

Basics: Syntax looks like

Printing text examples:

print "this runs"

print("this also runs")

In Python 2, print is a keyword without parentheses; in Python 3 it is a function, so parentheses are required. This small change can cause migration headaches.

Arithmetic: Unexpected results

Integer division behaves differently:

x = 9 / 5  # gives 1 (integer math) in Python 2

x = 9 / 5  # gives 1.8 in Python 3

Python 2 returns an integer, while Python 3 returns a float unless you import true division from __future__. This has caused real‑world rounding bugs.

Strings break in weird places

Unicode handling differs: word = "naïve" Without a leading u, the string is stored as bytes, leading to decoding errors on some terminals. word = u"naïve" In Python 3, strings are Unicode by default; raw bytes require a b prefix: raw = b"naïve" This makes string processing more predictable, especially for APIs, file handling, or email content.

Verbose loops & range

Python 2 offers two options: range() – returns a list xrange() – generator (memory‑efficient)

For large loops, use xrange() to avoid building huge lists. Python 3 merged them; its range() behaves like the old xrange().

for i in range(1000000):  # no massive list built in memory
    pass

Exception style

The except syntax changed:

try:
    fail_something()
except ValueError, e:
    print e

try:
    fail_something()
except ValueError as e:
    print(e)

Small but important; linters catch it, but unfamiliar code can waste time.

Input differences

Python 2: raw_input() → returns a string input() → evaluates the input as Python code (dangerous)

Python 3: input() → returns a string only

Misusing raw_input or running code on the wrong version can cause NameError or other bugs.

Dictionary iteration

Python 2:

my_dict.items()      # builds a static list
my_dict.iteritems() # returns a generator (saves memory)

Python 3: .items() returns a view that works like a generator; iteritems() no longer exists.

This difference shows up when handling large JSON files or configs.

Library support and ecosystem

Most major libraries (Django, Requests, NumPy, Pandas) have dropped Python 2 support. Older versions may be vulnerable, unstable, or lack features, making dependency resolution harder.

Stubborn legacy systems

Internal tools built around 2010‑2014 often still run on Python 2 without migration budgets. Developers inherit scripts, patch them enough to keep them alive, and eventually upgrade the stack when resources allow.

Conclusion

Start new projects with Python 3. If you maintain older code, knowing these differences saves time, avoids embarrassing errors, and gives confidence when porting tools.