I think the schmitt circuit ( Post #7) is preferable - and it's acceptable to parallel gates to achieve a marginal improvement in the o/p current (typically 20mA per gate - read Data-Sheet). Most 4000-series are several gates inone package. Indeed inputs should be defined even if the gate isn't used, to prevent oscillation troubles - Schmitts tend not to suffer from this, but "good practice..."
The circuit presumes the input source swings between zero and 7v
I would add that a buffer cct might be a good idea aas we don't know what this i/p is and it might be disconnected. A simple buffer transitor should avoid gate damage due to stray spikes.
Also, to be picky, there should be a resistor 1-10k between the capacitor and the gate i/p, so protection diode current is not excessive.