-----Original Message----- From: boost-bounces@lists.boost.org [mailto:boost-bounces@lists.boost.org] On Behalf Of Scott McMurray Sent: Thursday, March 31, 2011 1:34 AM To: boost@lists.boost.org Subject: Re: [boost] gsoc project
On Wed, Mar 30, 2011 at 09:57, Paul A. Bristow <pbristow@hetp.u-net.com> wrote:
But there are *some* similarities between the two. (Often you feed a string and you get a 'check' be it a single 4 bit digit, or a loadsabits digest?).
Conceptually, I agree.
But since the point of check digits is to find human-entry errors, the trade-offs and practical issues are totally different.
For example, they only make sense for things short enough that a human would be willing to type, so having an interface that encourages sequentially providing 4-KiB blocks of unsigned chars makes no sense.
Similarly, a human has provided it manually, so passing it as a wstring is a reasonable, performance-mostly-unimportant thing to do, compared to how long it took the user to type it. Especially since the checker might need to consider whether the ISBN-10 was entered with 'X', 'x', 'х' (Cyrillic ha), 'Ｘ' (fullwidth), or 'ｘ' (fullwidth).
No disagreement here (though some will also want to use the check on machine written files/databases). I'm sure anyone who does this project will be tapping your expertise on this area. Paul --- Paul A. Bristow, Prizet Farmhouse, Kendal LA8 8AB UK +44 1539 561830 07714330204 pbristow@hetp.u-net.com