Difference between bulk update with FROM VALUES and without it?

Nobody has claimed this yet.

```python3
await conn.executemany(
"""
UPDATE user SET login=$2, name=$3, age=$4, sex=$5
WHERE id=$1
""",
users,
)
```

```python3
await conn.executemany(
"""
UPDATE user SET login=actual_user.login, name=actual_user.name, age=actual_user.age, sex=actual_user.sex
FROM (VALUES ($1, $2, $3, $4, $5)) AS actual_user (id, login, name, age::integer, sex::integer)
WHERE user.id=actual_user.id
""",
users,
)
```

The first is cleaner. But I worry about performance. I know the second is performed as one request. But the first is the same?

Is there any reason to use the second way?

Start with asyncpg's executemany entry point and PostgreSQL's UPDATE ... FROM VALUES semantics, comparing both SQL examples and their request behavior. Done would be a clear documented explanation of whether the approaches differ in performance or purpose, and when the second form is useful.

Written by the indexing model from the issue text.