The following is a list of features under consideration for the next release of Snarf. If you have a feature you'd like to see that does not appear here, or if you have some comments to make on these features, please send WWIVnet e-mail to Graham Mainwaring #1 @9970. * Have Snarf process file descriptions as well as just files. This is a big one. When a user uploads a file, they will have to specify a description for it somehow. Also, if Snarf understands file descriptions, users will be able to do things like download a list of files which are new since a certain date. In addition, this feature will allow Snarf to update WWIV's file records for the number of times a file has been downloaded. The problem with this feature is that many WWIVs are modified and do not use standard file formats. Some method must be provided for making Snarf work with modified userrecs and directoryrecs. * Give the sysop the ability to break into chat with a user while they are Snarfing. (To be fair, perhaps also put a user option in SNARF CONFIG for 'refuse to accept chats from sysops'.) * Become compatible with any !-@REMOTE@-! interface standards that develop, or if none do, create one. (If such a standard exists, I would like to hear about it.) * Allow protocols besides DSZ Zmodem. Possibly even provide support for bimodem. However, this would require a method for negotiating between BBS and user as to which protocols both systems understand. * Provide Snarf companion utilities which generate and manipulate file lists. Perhaps even provide a menu & mouse driven shell for Snarf which allows downloading files chosen from a multiple-select list box. Possibly do the user interface using Turbo Vision. * Add some more security features. For example, allow password-protection of some directories. Another option would be requiring a valid WWIV username and password before allowing Snarf downloads. Adding this feature would allow sysops to maintain file ratios and restrictions for Snarf downloads as well as regular downloads. However, since this would require reading userrecs, implementing this feature would require a solution to the modified-file-structure problem.