How a Tiny Python Default‑Argument Bug Sank Digg’s v4 Launch and Triggered Its Downfall

In 2011 Marc Andreessen warned that software was swallowing the world, a reminder that a single bug can undermine even the most promising products.

Will Larson, a former Digg engineer, later wrote about the company’s collapse after a disastrous v4 release, tracing the root cause to a Python default‑argument mistake.

Digg, founded in 2004, let users submit and vote on news items, using internal algorithms to surface popular stories. This user‑driven model helped the site grow, but also exposed it to manipulation and operational strain.

Internal turmoil—CEO turnover, senior engineers leaving, algorithm abuse, and broken development tools—combined with external pressure from Google’s 2011 Panda update, which sharply reduced Digg’s traffic.

To compete, Digg rewrote its platform from v3.5 to v4, but the project was delayed for two years. Facing a shrinking market, the team rushed the unfinished v4 into production without a rollback plan.

During the switch, the new site appeared, but most pages failed to load. The team blamed Cassandra, expanded memcache usage, and later rewrote the MyNews page using Redis, yet problems persisted, requiring manual restarts every four hours.

The decisive bug lay in a Python Tornado API service: the function def get_user_by_ids(ids=[]) used a mutable list as a default argument. Because Python evaluates default arguments only once, the same list accumulated user IDs across requests, eventually overwhelming memcache and crashing the site.

def get_user_by_ids(ids=[]):
    # logic that appends to ids
    pass

A simple illustration shows the effect:

def f(l=[]):
    l.append(1)
    print(l)

f()
# [1]

f()
# [1, 1]

f()
# [1, 1, 1]

After fixing the bug, Digg v4 finally ran correctly, but the damage was done: a new CEO was appointed, a third round of layoffs followed, and a year later Digg was sold for $500,000, a steep fall from its $1.6 billion valuation.