I don't think people in this thread are insisting on a downloadable package. Just that the algorithm published in the write-up accurately describe the work that

I don't think people in this thread are insisting on a downloadable package. Just that the algorithm published in the write-up accurately describe the work that was actually done to achieve the stated results. If even that threshold can't be met, IMO the authors are polluting the literature with bad work.