Rational numbers in D

[Jonathan M Davis]

On Saturday, June 09, 2012 18:55:14 Joseph Rushton Wakeling wrote:
> What do you think of the idea of a "priority wish list" of functionality for
> D and/or Phobos?  Do you think it would be a useful way to guide
> contributors to where they might be useful?  (I would view the priorities
> as being set by the core dev team but with some input from community.)

In theory, it's a good idea. In practice, I question how much it will affect 
things. Pretty much all of the modules which have been presented for review to 
Phobos are modules that a particular developer personally wanted to be in 
Phobos and spent the time and energy required to get it there. I don't believe 
that there's been a single case of anyone creating a module for review simply 
because a lot of people wanted that particular module. The closest that we've 
gotten to that is when there are a bunch of people who want a particular 
module and one of the people who does steps up and does it (e.g. the semi-
recent std.log proposal essentially came about this way).

On the whole, the problem is more one of having people with the time, 
expertise, and passion to write new modules for inclusion in Phobos and push 
them through the review process and not a problem of those kind of people 
looking to contribute to Phobos but not knowing what to work on. One thing 
that definitely pops up is people who have spent the time to create a module 
for their specific needs which could be adapted to the standard library given 
time and effort, but they don't want to (or can't) put in that time or effort.

Certainly, I could see it being the case that someone has been working on  
particular module not realizing that other people would want it, and a 
"priority wish list" could help inform them that it may be worth their time to 
try and get their module into Phobos. And it's obviously _possible_ that 
someone will look at the list and decide to go implement one of the things on 
the list. I just don't know how likely it is.

If you want to try and create such a list, I say go for it. It wouldn't hurt, 
and it may very well help. I just wouldn't hold my hopes up that it will 
really lead to developers going and implementing what's on the list simply 
because they now know that a lot of people want it.

- Jonathan M Davis