Which Phobos modules are expected to get a redesign?

[Jonathan M Davis]

On Saturday, August 13, 2011 00:45:15 Jonathan M Davis wrote:
> If possible, I'd like to get the actual list of modules which we're
> expecting to redesign in Phobos. The ones that I'm sure are supposed to get
> some level of redesign are
> 
> std.json
> std.process
> std.stream
> std.variant
> std.xml
> 
> Besides those, std.container isn't really going to be redesigned all that
> much from what I understand, but some level of redesign is going to occur
> to sort out memory management as well as fill out the list of containers
> that it has. Also, std.getopt looks like it's going to get some level of
> redesign to sort out its behavior, but I don't know whether it's enough to
> get a full, formal review. And there has been some discussion of
> std.encoding getting redesigned, but there are no firm plans about that.
> std.string also likely needs some redesign with regards to its functions
> which take patterns, but the module as a whole doesn't need to be reworked.
> 
> One that I'm not at all sure about is std.socket. Does it need a redesign?
> Is it the current plan for someone to redesign it at some point? I've never
> used it so, I don't particularly familiar with what may or may not need to
> be done with it. I've heard it mentioned that it might need a redesign
> though.
> 
> Are there any others besides those? Did I miss any? I'd like us to have a
> fairly firm list on what modules need to be redesigned so that we can make
> sure that it's taken care of in a timely manner and so that we can be clear
> about what we have left to do in terms of major reworking of Phobos. We
> need to be working towards having an API for Phobos which is as fully
> stable as we can reasonably get it.
> 
> Also, it would be good if we could get some clarity with regards to which
> modules currently have someone working on their redesign and which need
> someone to take on the work of reworking them.

Should std.windows.registry be added to the list? That one has been getting 
reworked somewhat, piece by piece, but it's still a bit of a mess in terms of 
what's in druntime and what's in Phobos with regards to Windows system calls 
and the D stuff no top of them. Also, std.windows.registry isn't under the 
Boost license, so we may want to try and get ahold of the original author so 
that we can fix that.

- Jonathan M Davis