dennis@rlgvax.UUCP (Dennis Bednar) (01/11/86)
Our SV man page for aliases(4) says that the format is name: name_1, name_2, name_3, ... Most of the entries in our file are absolute uucp path names such as nextmch!nextnextmch!lastmch!person. However, is it okay to put in names such as person@lastmch.uucp or person@lastmch.UUCP? Or does the answer depend on how our sendmail configuration file is defined (of which I know zilch about)? -- Dennis Bednar Computer Consoles Inc. Reston VA 703-648-3300 {decvax,ihnp4,harpo,allegra}!seismo!rlgvax!dennis UUCP
guy@sun.uucp (Guy Harris) (01/12/86)
> Our SV man page for aliases(4) says that the format is > name: name_1, name_2, name_3, ... > > Most of the entries in our file are absolute uucp path names > such as nextmch!nextnextmch!lastmch!person. > > However, is it okay to put in names such as > person@lastmch.uucp or person@lastmch.UUCP? 1) To forestall future comments, there is no S5 man page for "aliases(4)". The system in question is a 4.2BSD/S5 hybrid; the "aliases" file is used by "sendmail" which comes with 4.2BSD but not with S5 (although it can be made to work under S5). 2) The way "sendmail" handles aliases is that when it detects that it's about to run a mailer which does local delivery (which, in this case, it does by noting that the mailer is the one defined with the name "local", not by looking at the "local mailer" flag), it checks to see whether the address being delivered to is in the aliases file. If so, it gets what that alias expands to and pretends, in effect, that it was just given mail to deliver to that expansion. As such, it is processed just as if it had been given those addresses in the first place. As such, putting in names like that should work as well as sending to such a name in the first place. Guy Harris
steve@jplgodo.UUCP (Steve Schlaifer x3171 156/224) (01/13/86)
> Our SV man page for aliases(4) says that the format is > name: name_1, name_2, name_3, ... > > Most of the entries in our file are absolute uucp path names > such as nextmch!nextnextmch!lastmch!person. > > However, is it okay to put in names such as > person@lastmch.uucp or person@lastmch.UUCP? > I have successfully done just that and it works fine > Or does the answer depend on how our sendmail configuration > file is defined (of which I know zilch about)? > -- > Dennis Bednar Computer Consoles Inc. Reston VA 703-648-3300 > {decvax,ihnp4,harpo,allegra}!seismo!rlgvax!dennis UUCP A handy way to find out what happens is to try it and then do /usr/lib/sendmail -bt which will then prompt for a "rule number" and a "name" to which you can respond with 0 person@lastmch.UUCP and watch it go through its various contortions and finally print the line it would use to invoke the mailer to be used. ...smeagol\ Steve Schlaifer ......wlbr->!jplgodo!steve Advance Projects Group, Jet Propulsion Labs ....group3/ 4800 Oak Grove Drive, M/S 156/204 Pasadena, California, 91109 +1 818 354 3171
dennis@rlgvax.UUCP (Dennis Bednar) (01/14/86)
In-reply-to: your article <3140@sun.uucp> > 2) The way "sendmail" handles aliases is that when it detects that it's > about to run a mailer which does local delivery (which, in this case, it > does by noting that the mailer is the one defined with the name "local", not ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > by looking at the "local mailer" flag), ... > > Guy Harris Sorry, but I did not understand the phrase I put up-arrows under. -- Dennis Bednar Computer Consoles Inc. Reston VA 703-648-3300 {decvax,ihnp4,harpo,allegra}!seismo!rlgvax!dennis UUCP
bzs@bu-cs.UUCP (Barry Shein) (01/14/86)
>From: guy@sun.uucp (Guy Harris) >> From: someone else >> Our SV man page for aliases(4) says... >1) To forestall future comments, there is no S5 man page for "aliases(4)"... Well, it might be splitting hairs but I believe that AT&T's SYSV for the 370 has sendmail as a standard utility (at least that's what their brochures said.) I agree that's a 'Berkeley enhancement', but so is VI and once AT&T puts it on a standard distribution tape...well, I am not sure what it becomes. I suppose it is an Amdahl enhancement. (This is the product directly from AT&T, not Amdahl tho same derivation.) I agree, I strongly doubt this is the version in question, just thought I would mention that this lack of sendmail in SYSV may not go on much longer or even be completely true. The useful point is would people *PLEASE* try to put the O/S, version, release and machine type on queries, I have often been about to reply with what I thought was the answer to someone's problem when the lack of that info suddenly shook my confidence. Something like: System: SYSVR2V1 with V6 enhancements Machine: Cellular Multi-Warp 7755 And even RCS or SCCS ids if pertinent. Partial info is ok, but at least try (I can't always figure out which Version in a SYSV is meaningful either, we just went through that confusion on AT&T 7300s on another list.) -Barry Shein, Boston University
guy@sun.uucp (Guy Harris) (01/16/86)
> > ...noting that the mailer is the one defined with the name "local", not > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > Sorry, but I did not understand the phrase I put up-arrows under. The "sendmail" confguration file contains rules for parsing and modifying addresses. When an address has been fully parsed and modified, the message is sent out; "sendmail" recognizes that it's done with an address by a rule which includes an indication like $#local $:$n on its right-hand side. This means that the mail should be delivered by the mailer defined with the name "local"; a mailer is generally a program, and "sendmail" runs that program. Mailers are defined with lines in the configuration file like: Mlocal, P=/bin/mail, F=rlsDFMmnP, S=10, R=20, A=mail -d $u where "M" is the first character on the line, and the token that follows it is the name of the mailer, so this line defines a mailer with the name "local". For further quesions, read the "sendmail" documentation. Guy Harris
roy@phri.UUCP (Roy Smith) (01/18/86)
> For further quesions, read the "sendmail" documentation. > > Guy Harris Was this meant to be a joke? Yes, the "sendmail" documentation certainly does provide a source for questions; unfortunately it doesn't provide much in the way of answers. -- Roy Smith <allegra!phri!roy> System Administrator, Public Health Research Institute 455 First Avenue, New York, NY 10016
warren@pluto.UUCP (Warren Burstein) (01/19/86)
> However, is it okay to put in names such as > person@lastmch.uucp or person@lastmch.UUCP? > > Or does the answer depend on how our sendmail configuration > file is defined (of which I know zilch about)? Yes, it does depend on your sendmail.cf file. I hacked mine to recognize these addresses and give them to drilltech!sob's uumail program which would figure out a path for "lastmch". The case of the domain (uucp) seems not to matter. Unless you have this program your sendmail will probably do one of two things Translate person@lastmch.uucp to lastmch!person and attempt to deliver it with uux. Fine if your site talks to lastmch. Think the address is a local user, discover there is no one named "person@lastmch.uucp" at your site and return the mail. -- The Maxwell R. Mayhem Institute for Quandary Requiem and Maternal Sciamachy Accept no substitutes.
guy@sun.uucp (Guy Harris) (01/22/86)
> > However, is it okay to put in names such as > > person@lastmch.uucp or person@lastmch.UUCP? > > > > Or does the answer depend on how our sendmail configuration > > file is defined (of which I know zilch about)? > > Yes, it does depend on your sendmail.cf file. I hacked mine to recognize > these addresses and give them to drilltech!sob's uumail program which would > figure out a path for "lastmch". The case of the domain (uucp) seems not > to matter. Unless you have this program your sendmail will probably > do one of two things > Translate person@lastmch.uucp to lastmch!person and attempt to deliver it > with uux. Fine if your site talks to lastmch. > Think the address is a local user, discover there is no one named > "person@lastmch.uucp" at your site and return the mail. Or 3) feed it to another similar program instead, such as "uuroute"/"uubang" (written by, I think, Mark Horton) which does something similar - "rlgvax" runs "uubang". The way it runs "uubang" is that the UUCP mailer entry feeds the message to "uubang" rather than "uux"; "uubang" then translates it if necessary and hands it to "uux". The sendmail configuration file doesn't treat bang paths as different from @site.UUCP paths; it hands them both to "uubang". (And it's unlikely that it will think "person@lastmch.uucp" is local, unless it's a very odd configuration file, since there is a host name there after the "@" and it's not the name of the local site.) Guy Harris
honey@down.FUN (Peter Honeyman) (01/26/86)
uubang/uuroute should not be called in place of the uucp mailer, it should be called much earlier. here, we run addresses through a router before we allow sendmail anywhere near them. this admits convenient addressing for hosts on any network. the alternative is to force @silly.long:winded%addresses@else.where, allowing short!hand (or hand@short) for uucp only. peter
piet@mcvax.uucp (Piet Beertema) (01/28/86)
>> ...noting that the mailer is the one defined with the name "local" ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >Sorry, but I did not understand the phrase I put up-arrows under. ...."sendmail" recognizes that it's done with an address by a rule which includes an indication like $#local $:$n on its right-hand side. Nowhere in the manual is it stated that that mailer *must* be named "local". That's obviously what's behind the "don't understand" remark. -- Piet Beertema, CWI, Amsterdam piet@mcvax.uucp