Renamings, and fix duplicate code paths and actually use all of relays
for the status check instead of just one (like the FOLLOW does)

And since WebUI payload used it, this was not reported as 'unused'

Make sure we seamlessly handle 'convert' for the number and the string version.
And since it is a two-way map, update 'serialize' to use it as well instead of
either a simple static_cast<int> or duplicating the same strings used in 'convert'

Only string literals or static vars can be used in constexpr context,
but those can still be shoved into the flash region via PROGMEM.
Notably, PSTR(...) inside of a lambda is not a constexpr.

Some quirks to work out
- we don't 'enumerate' things through compiler, enum values may go
  missing since it is not a switch-case
- 'get' default value via query still requires us to know the settings
  key in the first place. and it still needs an explicit call to
  'serialize'
- sensor units are stringified as their display value.
  but, this also avoids two different 'string' versions of those
- EnumOptions struct instance may also be in PROGMEM, but one needs
  to be very careful to only allow aligned access to it's members
  (which currently means we can't use 8bit or 16bit 'enum class'es)

If the endpoint receives no header, treat it as '*/*'

Don't fail prematurely, since asyncwebserver handler needs to claim the
request to issue a custom response code.
ref.
- https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/406
- https://datatracker.ietf.org/doc/html/rfc7231#section-6.5.6

cstring's *cmp* return result is integer, not boolean
Pulse processing should only trigger when *current* state is not normal

Remove the ticker dependency, and make sure 'save' callback does not
override itself when at least one relay does not set the 'persist' flag

Also fixing accidental copies of the timer object.

Don't introduce our callback to the type system, continue to bind the
ticker instance as the timer's arg and just use the lambda to pass the
argument to the reset function

```
libraries/Ticker/src/Ticker.h:136:41: warning: cast between incompatible function types from 'void (*)(CustomResetReason)' to 'Ticker::callback_with_arg_t' {aka 'void (*)(void*)'} [-Wcast-function-type]
  136 |         _attach_ms(milliseconds, false, reinterpret_cast<callback_with_arg_t>(callback), reinterpret_cast<void*>(arg));
      |                                         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
```

Don't fail when the header also has charset=...
Also, fallback to text/plain when there's no Accept

Attempt to either parse with newer syntax, or fallback to the floating
point seconds as default. Settings also return a 'result' instead of the
default zero, fallback to build value otherwise (which is still floating point, though)

Update /pulse API endpoint to report the actual pulse timer value that
is active right now, not just the value attached via the setting

Unused variables, bogus const-qualifiers and some strict switch-case rules with -Wextra

Make sure we don't try to strlen some empty space or neighbouring
segments, that may or may not contain nulls

Also tweak the function to warn about large messages, since it is not
(yet) clear whether we actually want those. At least something to log.

Localize build constants & settings use, no need to prefix where
namespacing already does this (and also removes some duplicates)

Also calm down clangd, mark template specializations as unused

Make sure to properly clamp delay times, as the recent issue had shown
it is possible to go below the update limit. SNTP app does not do any
runtime checks, so it's up to us to enforce these limits

Update internal random number generator from raw `secureRandom()` to a custom
'std::random_device'-look-alike, where we could also use it with STL functions
(current implementation is `os_random()` aka `esp_random()`)

Not as clean as other modules, but at least prepare for the further changes.
Implement settings for float duration conversion, and test it out with pulse
and flood window time values.

More safe-guards for an externally-defined pulse time, make sure it is
not either inf or nan and is an actual finite number

Also use EnumerableConfig for the WebUI output, and make sure to use
internal _relays.size() instead of doing a function call

Saves ~100bytes in RAM

Get rid of 'assumed' units and use strong types instead
Also update HomeAssistant conversion code as float -> uint durations

No need for specific delay for each instance, just use a shared one

Make sure it could be moved further along, and the values inside of it
can be safely moved as well.

Also fixup commands that were missing ctx in OK / Error.

No need to duplicate argv::size()

Use internal duration types, store time in seconds and propogate as
uint32_t when it is necessary. Update tick values as well.

Instead of generic `toInt(), overload between `strtol` and `strtoll`.
Current implementation of `settimeofday` rejects `struct tz` pointer,
make sure it's nullptr.

Sync is a simple enough routine... But, expect an already synced system
before triggering the sntp restart.

As noticed in logs for 'Post Cache PlatformIO'
> Unable to reserve cache with key ..., another job may be creating this cache.

For this, keep things as milliseconds (at least for now)

Make sure to use __builtin_strlen in constexpr context
Remove `has*()` functions that are no longer used

Update WebUI callback to use EnumerableConfig

Don't assume the size <-> type, utilize type matching in the template
specialization to retrieve the actual type <-> format pair that we need

There's no need to move a temporary which is already an rvalue

Clean-up some redundant attributes and function in settings.
Make sure C types are proxied to fixed-sized ones, not the other way around.

Clean-up webprint implementation
- use typed duration for the backlog
- don't duplicate members and types for config, just re-use the struct as-is
- don't go over-the-top with const members, just proxy through a method

ref. https://github.com/esp8266/Arduino/pull/8393