[5.5] Allow to disable CREATED_AT

[hlorofos]

Exactly as we allowed to disable UPDATED_AT b4758e2
I'm allowing to disable CREATED_AT by setting
const CREATED_AT = false;
in the model

[tillkruss]

This is too risky, please re-submit with tests.

[mfn]

I thought this was already possible with overriding this in your model (or am I wrong?):

    public function getCreatedAtColumn()
    {
        return null;
    }

[hlorofos]

@mfn no its not possible, since getCreatedAtColumn is not used during timestamps update, instead it uses setCreatedAt method which blindly set created_at attribute:

$this->{static::CREATED_AT} = $value;

Thats why there is additional verification required in updateTimestamps method:

if (! is_null(static::CREATED_AT) && ! $this->exists && ! $this->isDirty(static::CREATED_AT)) {
              $this->setCreatedAt($time);
}

[Miguel-Serejo]

Wouldn't it be enough to set const CREATED_AT to null in your model?

[hlorofos]

@36864 if it was so easy I wouldn't created a pull request =)
No it will fail later in setAttribute method passing only one argument out of two required, so the model still believe it has created_at because $this->setCreatedAt($time); call from updateTimestamps.

[hlorofos]

I'll try to extend this PR with the tests as @tillkruss requested.

I'm trying to make the framework consistent, if we allowed to use const UPDATED_AT = null; why shouldn't we allow the same for CREATED_AT? For me it was the case, but plenty of others - they will never face this issue is the whole life.

[ludo237]

I don't understand this.

If you don't want to use created_at just set public $timestamps=false; or am I missing something?

[hlorofos]

@ludo237 yes, absolutely.
By using this setting I'll lost also updated_at.

In my case I want to keep model using updated_at but I don't need created_at.
Again, it might be the very rare case and nowadays developers doesn't care about a gazzilion additional column data in their DB. I'm not that one, sorry =)

[ludo237]

You could hook up the created/saved/updated events and work on that if you need an updated_at column alone.

Also an additional column is not that much of a big deal to be honest.

Donald Knuth once said:

Programmers waste enormous amounts of time thinking about, or worrying about, the speed of noncritical parts of their programs, and these attempts at efficiency actually have a strong negative impact when debugging and maintenance are considered. We should forget about small efficiencies, say about 97% of the time: premature optimization is the root of all evil. Yet we should not pass up our opportunities in that critical 3%