DBR 14.1 Pyspark Join on df1["col1"] = df2["col1"] syntax fails

[Trifa]

Hello
After upgrading my cluster from DBR 12 to 14.1 I got a MISSING_ATTRIBUTES.RESOLVED_ATTRIBUTE_APPEAR_IN_OPERATION on some of my Joins

df1.join(
    df2,
    [df1["name"] == df2["name"], df1["age"] == df2["age"]],
    'left_outer'
)

I resolved it by changing the syntax to:

df1.alias("l").join(
    df2.alias("r"),
    [col("l.name") == col("r.name"), col("l.age"]) == col("r.age")],
    'left_outer'
)

Is the second syntax the new standard or am I missing something with the first one ?

Thanks