Re: [boost] [gsoc 2013] chrono::date

Subject: Re: [boost] [gsoc 2013] chrono::date
From: Vicente J. Botet Escriba (vicente.botet_at_[hidden])
Date: 2013-04-26 22:54:20

Le 27/04/13 01:12, Anurag Kalia a écrit :
> I realize the fact that this is my second message and possibly very irritating. Sorry! (really)
>
> But I had an epiphany just now. Digging through boost docs, I saw in 'Appendices' > 'Future Notes' that they intend to release chrono::date on the basis of a proposal written by you! First of all, wow. Second of all, how concrete are those plans? They are included in boost docs themselves so I have doubts now if there is any room to modify the public interface.
All this is ongoing work. Nothing is fixed and nothing is delivered on 
Boost.Chrono.
And yes I would like to have Chrono/Date library in Boost, either based 
on my current interface, the one of Howard or yours if you reaches to 
get to finish a whole implementation with better interface. So no, the 
interface is not closed, but any proposal you could do would need to 
show it is a better interface.
>
> When I saw the given idea on boost gsoc page, I thought it was a reference point to go further. Now, do I have to implement the proposal as it is without minimal changes? Because, if yes, then I would have to change my proposal drastically, even after the fact that it is quite influenced by your paper itself.
As always it is up to you to make your proposal. You had the entry 
points up to you to make a convincing proposal. Propose an interface, 
compare it with the existing ones, argument why do you go this way, ...

Once thing is clear I would not mentor a project if I'm not confident 
with the competences of the student on the domain, know the problems and 
some solutions and has a clear vision of how the final library could 
look like. Of course during the project, things can and always will 
change, but we need a direction, a goal.

HTH,
Vicente