Because of problems at both FRMOP11 (RSCS V2 swallowing a few NJI codes from the file) and FRORS31 (being a JES3 node), Spain cannot reach FRECP11 (and probably a few other nodes). We had to transfer this file via MSGXFER... *-----------------------------------------------------------------------------* Date: Wed, 10 Sep 1986 15:55:44 ABC From: Jose Maria Blasco Comellas <ZCCBJBC@EB0UB011> Subject: Open subscription and private rev To: Distribution list <LSTSRV-L@FRECP11> If a list is defined as Subscription= Open, setting it to Review= Private (or Stats= xxx,Private) should have no effect. Consider the following dialog: tell listserv at uiucvmd stats REXX-L R; FROM UIUCVMD(LISTSERV): * You are not authorized to review statistics for list REXX-L tell listserv at uiucvmd subscribe REXX-L anyname R; FROM UIUCVMD(LISTSERV): * You have been added to list REXX-L tell listserv at uiucvmd stats REXX-L R; FROM UIUCVMD(LISTSERV): File "REXX-L STATREPT" has been sent to you tell listserv at uiucvmd signoff REXX-L R; FROM UIUCVMD(LISTSERV): * You have been removed from list REXX-L (The same stands for the REView command) If the list has been defined as Notify= No (which is the case for list REXX-L at UIUCVMD) the only means of detecting such an unauthorized (?) access is inspecting the LISTSERV's console log. What to do? -- either define the list as validate= All Commands or set it to Notify= Yes, so that this kind of access to the list can be discouraged or at least detected. In any case, I think that Subscription= Open, Validate= Store Only, Notify= No should have priority over Review= Private and force it to Review= Public. Any ideas? J. M. Blasco *-----------------------------------------------------------------------------* My reply: DIRMAINT checks the operand of a CMS command to make sure that it can not cause any VM READ state. "If command = 'SORT' or command = 'FORMAT' then reject". Now if I do a EXEC ZAZA with ZAZA EXEC containing a &READ VARS, here I go. My opinion on all that kind of problems is that if the person who sent the command/installed the server/configured the list specified stupid options, then the result will be stupid. You can configure a list with review= postmas- ter, send= public, subscription= open, files= no while formcheck= yes, and even a notebook on the D-disk (t-disk), confidential= yes but validate= store only and pw='', etc. However, I might indicate in LISTKEYW MEMO the sets of options that are "stupid", if that may help. Any comment? Eric