GeezerVison

assertTrue(first, second) treats second as the failure message, so each of these assertions only checked that first was truthy and never compared it with the intended value. This switches them to assertEqual() or assertIs().

They were found by walking the AST of every tracked .py file for assertTrue/assertFalse calls with two positional arguments and reviewing each hit by hand; the remaining ones pass a genuine failure message.

Two of the corrected assertions would have failed as-is, because the expected value had gone stale:

• test_ssl.test_ecc_cert: TLS 1.3 is negotiated here, and TLS 1.3 cipher suite names (e.g. TLS_AES_256_GCM_SHA384) don't include the key exchange or authentication algorithm, so the ECDHE-ECDSA prefix check only makes sense for TLS 1.2. It is now done only when TLS 1.2 is negotiated. test_dual_rsa_ecc pins TLS 1.2 and keeps the unconditional check. Dropping the check from test_ecc_cert instead would also be reasonable.
• test_ttk_textonly.test_format_elemcreate: _format_elemcreate(None) has returned ((),) rather than (None, ()) since ttk.Style.element_create using incorrect tk.call syntax #68166 changed the function's return format; the expected value is updated to match.

In the ctypes tests the expected values were str while the functions return bytes, so they are now bytes literals.

Test-only change, so this should be fine for skip issue / skip news. Tested locally on Linux with OpenSSL 3.0.13; test_ttk_textonly was not run locally (no _tkinter).

🤖 Generated with Claude Code