Re: [boost] [xint] Design Improvements (was Boost.Move vs Copy-on-Write timings)

Subject: Re: [boost] [xint] Design Improvements (was Boost.Move vs Copy-on-Write timings)
From: Chad Nelson (chad.thecomfychair_at_[hidden])
Date: 2010-05-05 09:39:52

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 05/05/2010 12:06 AM, Jeffrey Lee Hellrung, Jr. wrote:

>> [...]
>>> +1, and going the natural step further, build these algorithms off a
>>> view *concept*, rather then hard-coding a specific view class, for
>>> maximum flexibility.  The requirements you need *might* be as simple as
>>> existing range requirements in many cases (see Boost.Range).
>>
>> I'm afraid they'll be slightly more complex, as the view will need to
>> know the sign of the integer as well (and be able to alter it, for
>> mutable arguments). So I'll probably need to hard-code at least one view
>> class anyway. We'll see when I get to it, which I hope will be tonight.
> 
> Yes, you're right, start with a concrete view class, just with the
> additional abstraction to a view concept in the back of your mind.
> [...]

Did so. I've got a mutable_view_t and a const_view_t now, and I've
switched three interlocking functions to using them. I've got some
debugging to do on them though, and no time to finish it this morning. :-(
- -- 
Chad Nelson
Oak Circle Software, Inc.
*
*
*
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkvhdSgACgkQp9x9jeZ9/wSndACg0ORGb4ujco7SUGS1aCqgD5XP
dvUAn38VHyCMs03UZTLNYpO6S0VlCl32
=t8zN
-----END PGP SIGNATURE-----