Ext.Date.add preventDst bug

[stevepn]

The default value of `preventDst` is false, according to docs. Right? Right, in Ext 6.7.0.

Now, guess the result of this:

const startDate = new Date('10/1/2019 00:00');
const added27Days = Ext.Date.add(startDate, Ext.Date.DAY, 27);
const added648Hours = Ext.Date.add(startDate, Ext.Date.HOUR, 27*24); //It is 27 days, yes
console.log('It should be true, but ', Ext.Date.isEqual(added27Days, added648Hours));

Oh, wow. Let's break this down.
You broke a documented behaviour without any notification in a minor version upgrade.
It is still behaves badly in 7.0.0.
The code caused the break is the following

if (preventDstAdjust === false) {
  d.setTime(d.getTime() + value * 24 * 60 * 60 * 1000);
} else {
  d.setDate(d.getDate() + value);
}

Nice?
Let me guess the reason, the date time picker was broken badly, it had two October 27th this year(the day on DST change occurring always doubled), but instead of fixing date time picker to use your function as you documented(pass true as the last argument), you conditionally broke a core utility function without notification. Seriously asking, why are you doing this? This is not the first or the second time you make breaking change in minor version upgrades, but what kind of "fix" is that?

[mitchellsimoens]

I'm not sure what caused this change but the "fix" is not a fix. Since the argument is boolean, truthy vs falsy checking like the other cases in that switch are enough.

I've logged a request to have them open a bug to fix this and pointed out there may be other issues that will need to be resolved from reverting this code.

[mitchellsimoens]

They opened a bug for this: EXTJS-28680

[SebTardif]

I wrote many big Ext JS apps having no third party component, but after 'learning' I now use https://momentjs.com/ for anything about Date.