How a Tiny Python Default Argument Bug Sank Digg v4 and What Developers Can Learn

In 2011, venture capitalist Marc Andreessen warned that "software is eating the world," a reminder that even a single bug can topple a thriving product. This article recounts the real‑world failure of Digg's v4 launch, a once‑promising tech news site valued at $160 million, which collapsed after a disastrous software release.

Background

Founded in 2004, Digg let users submit and vote on news, surfacing popular stories on its homepage. While this user‑curated model initially drove growth, internal strategic missteps and external pressures—such as Google’s Panda algorithm—began to erode traffic and revenue.

The Rush to Release Digg v4

Facing declining traffic, Digg rewrote its platform (v3.5) and prepared a major v4 overhaul. After two years of delays, the team hastily pushed v4 live without a solid rollback plan, hoping to regain market relevance.

Python Function Default Parameter Failure

Shortly after deployment, pages failed to load for logged‑in users. The team initially blamed Cassandra and expanded memcache usage, but the root cause was a subtle Python bug. The Digg API, built with Tornado, exposed an endpoint that retrieved users by name or ID, defining both parameters with a mutable default list:

def get_user_by_ids(ids=[])

In Python, default arguments are evaluated only once at function definition, so the same list instance is reused across calls. Each request appended user IDs to this list, eventually filling it with tens of thousands of entries, overwhelming memcache and crashing the service.

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

f()
# [1]

f()
# [1, 1]

f()
# [1, 1, 1]

The bug persisted for a month, forcing the team to repeatedly restart processes every four hours. Eventually they rewrote the affected API, replacing the mutable default with a safe immutable default and migrating to a sharded Redis cluster.

Aftermath

Digg finally stabilized the v4 release, but the damage was done. Within weeks a new CEO was hired, followed by a third round of layoffs, and a year later Digg was sold for $500 k to Beatworks. The episode serves as a cautionary tale about the dangers of mutable defaults, rushed releases, and inadequate testing.

Will Larson, the author of the original post, reflects that while the decision to ship was controversial, the experience underscores the importance of prioritizing robust, well‑tested code over speed.