Deprecations for the next major release

[sidepelican]

mockolo has a long history and includes some features that are not widely known.
With the advancements in Swift, some of these features are now likely impractical or obsolete, yet they are still being maintained.
Some features also require a lot of effort to keep working.

Currently, we do not have enough people or time to maintain mockolo, and we are struggling to keep up with the latest features in Swift.
Therefore, for the next major update, we would like to deprecate some old or undocumented features that seem to be unused.

The features scheduled for removal are as follows:

• --allow-set-call-count flag
  This will always be treated as true. There shouldn't be any cases where this causes issues.
• --use-template-func flag
  Is anyone actually using this? It is not well tested, and there may be many situations where it doesn’t work properly. Furthermore, in modern Swift there are alternative approaches such as function body macros.
• Parent mock feature
  This is an undocumented, hidden feature with very complex behavior. In short, it allows you to provide part of the mock class implementation from existing source code. While it is a unique feature, I don’t believe anyone is using it.

Deprecation Plan

First, we display a warning in the console when these features are used. Afterwards, unless there are strong objections, we will proceed to remove these features.

[alanzeino]

Looking at the uber internal Swift codebase, we're using all three features. While I'm not sure if we still need them, we definitely still pass the first two flags and use the last feature as well.

[alanzeino]

This is a bit of a tricky situation to navigate. While I can understand the need to modernize the tool and not support things that are difficult to maintain today, the purpose of this library was to support Uber's needs first and then to keep up with Swift as it evolved and gained new features.

For us, if things we rely on are removed in a future major version (assuming you use semantic versioning), it wouldn't immediately affect us but as soon as we come across some problem with Swift syntax that is fixed in the theoretical future major version that removes the above features, we would be forced to reckon with a difficult migration to that newer major version instead of it simply being a quick version bump.

If you provide a migration guide this might be okay, in the case of the second and third feature. Note that the original author of mockolo left Uber a long time ago so if you remove all of these features it will fall on someone inside Uber who is not intimate with how mockolo works to have to migrate away from those features.