This is essentially the clock used by the Femto-4. Please use this instead of the old version made for Dustin Harris' 16-bit computer (which was directly ported from an old version of the Femto-4) for efficiency reasons. 

What has changed?:

The old system used a pair of monostables running into each other to produce the pulse loop. This was bad, since each clock loop involved up to ten circuit updates per edge. The new system now only has a loop of three components, reducing the number of updates. The new loop is also naturally unstable, and no longer requires a pulse to start it, making it generally more flexible and stable should you wish to alter execution patterns. 

Why should I use the new system?

The new system causes less updates. This has three primary benefits: a) You can overclock to higher speeds because less updates are involved, b) it causes less lag for the same overclocking speed, improving real time performance, and c) it is less fragile than the old clock and is more resilient to variations in how it is disabled. 

What other alterations can be made to it?

The clock currently does not pulse on both edges of the default clock. Using monostable circuits can be used to do that, doubling the execution speed. Alternatively, two clocks, one which uses the negation of the clock, can be used like the Femto-4. 

How does this work?

Please refer to: https://circuitverse.org/users/4699/projects/circuitverse-delay-introduction