[Cosmo-dev] cosmo.datetime.Date questions

[Matthew Eernisse]

Excellent. Clarifications and questions for next steps below.

Bobby Rullo wrote:
> Yes, some more (not a terrible amount more) would be helpful. Maybe a 
> link to the original JS date stuff, and a warning that you CAN use 
> direct access of properties, but you get no validation.

Yes, if we want to make it easy to serialize/deserialize to JSON or 
whatever without going through extra steps, we have to make those 
public. Noting clearly in the documentation that you can shoot yourself 
in the foot if you touch the raw props seems like a sensible precaution.

>> At least right now, it seems like the only thing that is broken is the 
>> service call, because it's using passing no args to the constructor 
>> and then setting attributes one-by-one with setters -- instead of 
>> passing the attributes in as params the way you would instantiating a 
>> native JS Date.
>>
> 
> The service call is not broken - it's not using the setters any longer. 
> I changed that before I reported this bug.

Yes, I noticed we got February back. :) Maybe 'broken' was not the right 
word -- I should have said 'impacted by this change.' And I guess the 
tests are actually the other part.

> +1 on #2

Good deal. Does anybody else have an interest or opinion on this?

What changes would we need to make in the tests so that they reflect the 
current behavior of the code?


Matthew