Most floats don't end! They're just truncated for display.

float_truncation.py

import struct
from decimal import *
getcontext().prec = 40

x = 1.987654321987654321

h = int.from_bytes(struct.pack(">d", x), "big")
f = h & ((1 << 52) - 1)
e = (h >> 52) & ((1 << 11) - 1)
s = h >> 63

frac_part = sum(
    ((f >> (52 - i)) & 1) * Decimal(2)**(-i)
    for i in range(1, 53)
)
val = (-1)**s * (1 + frac_part) * Decimal(2)**(e - 1023)

print(repr(x))
print(f"{x:.39f}...")
print(val)