|
Boost : |
From: Daryle Walker (darylew_at_[hidden])
Date: 2004-08-30 02:47:10
On 8/28/04 3:36 PM, "Jonathan Turkanis" <technews_at_[hidden]> wrote:
> "Daryle Walker" <darylew_at_[hidden]> wrote in message
> news:BD55A2D6.E82D%darylew_at_hotmail.com...
[SNIP]
>> But why go through all the trouble to forcing an idea through a framework,
>> which is built on top of the Standard I/O framework, when the idea is simple
>> enough to go through the Standard framework directly?
>
> The components, such as null_sink and value_source, are actually *easier* to
> write than the corresponding stream buffers.
> They can also be reused in cases where streams and stream buffers are not
> necessarily the best abstraction.
When would that ever happen? (More on this in my review of the next
Iostreams library.)
> So I don't see this as a case of forcing them into a framework. As sources and
> sinks they express their core functionality concisely; as stream buffers, the
> are forced into the 'basic_streambuf framework' with its pecular interface
> (underflow, uflow overflow, xsgetn, etc, eback, pbase, ....)
But they're still forced through two frameworks. The extra indirection
isn't worth it (especially since I've gone through your library).
-- Daryle Walker Mac, Internet, and Video Game Junkie darylew AT hotmail DOT com
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk