Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.2 9/18/84; site brl-tgr.ARPA Path: utzoo!watmath!clyde!burl!ulysses!allegra!mit-eddie!genrad!panda!talcott!harvard!seismo!brl-tgr!wmartin From: wmartin@brl-tgr.ARPA (Will Martin ) Newsgroups: net.news Subject: Postnews and vi -- odd behavior Message-ID: <10365@brl-tgr.ARPA> Date: Thu, 2-May-85 11:25:55 EDT Article-I.D.: brl-tgr.10365 Posted: Thu May 2 11:25:55 1985 Date-Received: Sat, 4-May-85 00:45:43 EDT Distribution: net Organization: USAMC ALMSA Lines: 31 I have noticed the following odd behavior in postnews when using the vi editor; is this standard amongst all news-running systems or local? If standard, is it intentional or a bug? Sometimes, when running postnews, I want to save a copy of my posting before submitting it. I am in vi, finishing the composition and editing. I do ":w filename" and it works OK, and the screen still shows the text of the posting. I then do "ZZ" and the system responds with: "File not modified - no message posted" and then the shell prompt appears. (I just did it again to verify this.) As far as I know, doing a "w" to another filename in vi should NOT affect the status of the editing buffer (or whatever you call the space wherein vi works on the data), and, as I said, after doing the "w filename" command, the data is still there. So maybe the pointer to the temporary postnews file is being destroyed? But I thought that vi's "memory" of the original filename wouldn't be affected by a command to explicitly "w" to another file. I can, of course, recover by restarting postnews, reading in the saved file, and deleting the duplicated header info, but it is a waste of time. I don't notice this behavior when posting from within "rn", so I got used to the practice of saving with the "w filename" command. However, postnews is less offensive in the number of questions and gorp the user has to wade through to do something, so I prefer stand-alone "postnews". Any comments? Will Martin USENET: seismo!brl-bmd!wmartin or ARPA/MILNET: wmartin@almsa-1.ARPA