Here's what I had in mind: The files in question are going to be fink.info, fink-mirrors.info, and the like.
If foo.info is already present, but foo.info.bak is absent, then foo.info will be moved to foo.info.bak and a new foo.info will be created. The user will be told that this has happened. If both foo.info and foo.info.bak are present, the user will be warned that foo.info.bak will be overwritten, and asked if it's OK to continue; if the answer is "no", the program bails. (Actually, to follow a suggestion which pogma made to me last week in a different context, perhaps the backup file should be called foo.info.finkbak .) These actions will *never* be triggered by running fink itself. The only way to trigger them will be to do a CVS checkout of the relevant directory, and run "./inject.pl" within that directory. That, to my mind, is quite similar to a user's own action, and is something only likely to be done by developers. I guess we could reinforce this by having ./inject.pl ask the user at the beginning if its OK to write in the local/main tree. Maybe even give the user an option to choose a different tree? (which would then be created by the program, by the way) However, if this is too intrusive, I am willing to go back to having a tree whose sole purpose in life is to receive ./inject.pl-created .info files. -- Dave ------------------------------------------------------- This SF.Net email sponsored by Black Hat Briefings & Training. Attend Black Hat Briefings & Training, Las Vegas July 24-29 - digital self defense, top technical experts, no vendor pitches, unmatched networking opportunities. Visit www.blackhat.com _______________________________________________ Fink-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/fink-devel
