On 01/03/2011 01:34 PM, Michael Stone wrote: > Pulling in a bignum library into the required list in order to > support the factor(1) command seems a little obnoxious. If someone > wants to design a coreutils-factor-bignum package that diverts > factor or somesuch I guess I'd consider it, but is there really that > much demand for this feature? (That is, would anyone really use it > for anything other than novelty value?) > > Mike Stone
For your last point: I do not know. I can imagine some boundary usage, sort of ad-hoc, but I cannot figure this being needed on a constant base -- in this case, one could always: * get the coreutils source, add the build-dep, and build & deploy it; or * get the coreutils source, and build & deploy just factor; or * write/copy/adapt any of the multiple factorisations programmes out there. I, for one, never needed factor(1) to factor anything bigger than 10**20, or thereabout. I cannot justify the cost of creating a factor-bignum package, I think it is too much effort for a low yeld; additionally, like you, I really do not like the idea of adding libgmp to the dependencies. But I considered it more wiser to ask you. Perhaps wishlist, and wait for others to consider a split? Thank you, Cheers, ..C..
signature.asc
Description: OpenPGP digital signature
