Personal Catalogue: Django & PostgreSQL — Chapter 1, Exercise 2
====================================================

TASK
Explain the real, documented difference between psycopg2-binary and
plain psycopg2, and why this project's own deadline makes the binary
package the right choice for now.

SOLUTION

Both packages are Python adapters that let Django talk to a PostgreSQL
database. The real difference is in how they're built and distributed.

psycopg2 (the plain package) is compiled against your own system's
installed PostgreSQL client libraries at install time. That means it
needs those libraries, and a working C compiler toolchain, already
present on whatever machine you're installing it onto — genuinely fiddly
to set up correctly, especially on a machine that doesn't already have a
PostgreSQL development environment configured.

psycopg2-binary bundles its own precompiled copy of those same client
libraries directly inside the installed package. It requires nothing
extra on the target machine beyond running `pip install psycopg2-binary`
- a real, meaningful reduction in setup friction, which is exactly why
it's the officially recommended choice for local development and quick
prototyping.

The real, documented tradeoff is that psycopg2-binary is not recommended
for production deployments. Bundling a self-contained copy of the client
libraries can produce subtle version mismatches or duplicate-library
conflicts in a production environment's own more tightly controlled
dependency setup - a risk not worth taking once real, paying reliability
is on the line, which is why the plain, system-linked psycopg2 package
is the documented correct choice there instead.

Given this project's own real deadline, the setup-friction savings from
psycopg2-binary directly serve the goal of getting a working first
release running quickly during development - with the plain psycopg2
package picked up specifically once Chapter 9's own deployment step is
reached, not before.

WHY THIS WORKS AS AN ANSWER
----------------------------
It explains the actual technical difference (bundled vs. system-linked
client libraries) rather than treating the two package names as
interchangeable, and ties the choice directly back to this project's own
real time-pressure constraint rather than presenting it as an arbitrary
preference.