Repository maintainers need to collect, analyze, and report usage statistics for several reasons:
A server program normally writes access records to a log file. In the case of http, these records contain only the domain name or IP address of the requesting machine, not the identity of the individual user. In the case of anonymous ftp, the log records contains the email address typed in by the user. It is possible for an http server to provide access via a cgi-bin program that requests information from the user, such as name and email address. Many users would consider it an annoyance to see the request on every file download. If a mechanism for bulk file download is adopted by the WWW community, it will be possible for the user to supply the information once per bulk download. Access mechanisms for restricted software that require user authentication will do more extensive record keeping, such as retaining a digitally signed transaction record.
Tools to collect, analyze, display, and report log statistics (including reporting to home sites of mirrored files) will be available as part of the NHSE's Repository in a Box (RIB).
Download counts only tell how many times a file has been requested. They tell nothing about whether or how the software has actually been used. To obtain this important information, repository maintainers will need to try to contact users who have downloaded the software. This followup might most practically be done on a per-package basis by the software developers, perhaps through a user group mailing list. Repository maintainers might provide assistance to authors in carrying out followup surveys.