gaguiar@pinocchio.encore.com (George Aguiar-Consultant) (01/06/91)
Does anyone have an opinion on the above mentioned package? Is it really better than hacking sendmail.cf directly? Is 2.1 the most up to date release? Is there any other PD software that accomplishes the same thing? ( ie. generate sendmail.cf file ) Thanks in advance, --gaguiar
rknight@gec-rl-hrc.co.uk (Roger Knight (B21)) (01/10/91)
In article <13738@encore.Encore.COM> gaguiar@pinocchio.encore.com (George Aguiar-Consultant) writes: >Does anyone have an opinion on the above mentioned package? Is it >really better than hacking sendmail.cf directly? Is 2.1 the most up >to date release? Is there any other PD software that accomplishes the >same thing? ( ie. generate sendmail.cf file ) > >Thanks in advance, >--gaguiar I used UK-Sendmail #2.1 to set up the configuration for our Apollo network and it is certainly easier than hacking sendmail.cf, it is also completly different to the .cf file. UK-Sendmail provides an "MMDF-like" interface, you write a set of "domain tables" which re-write addresses to their full form and a set of "channel tables" which specify the routes to get to certain hosts/domains. UK-Sendmail then compiles that into a customised .cf file. Unlike sendmail the distribution contains a tutorial and a set of examples to get you going. With a bit of trickery it is quite easy to get UK-Sendmail to produce a .cf file which will put your host name into msg headers for local destinations (i.e. hosts on the same ethernet) and correct domain names for msgs to the outside world. One posible problem: it was written for use in the UK JANET network which uses big-endian addresses. However there is a switch to reverse addresses so they are little-endian for the rest of the world so it might work OK. Version 2.1 is the latest release (and probably the last, the authors having left their University Depts. for "real" jobs). Unfortunatly the initial distribution had 2 minor bugs. These have been fixed on the version from Imperial College (doc.ic.uk) but may be present in other copies. I can mail/post details if requested. ================================================================== Roger Knight GEC-Marconi Ltd., Hirst Research Centre East Lane, Wembley, HA9 7PP, United Kingdom !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! DISCALIMER: I said it, NOT my company. ==================================================================
mike@chorus.fr (Mike Bayliss) (01/14/91)
>>>>> On 10 Jan 91 15:29:20 GMT, rknight@gec-rl-hrc.co.uk (Roger >>>>> Knight (B21)) said: Roger> In article <13738@encore.Encore.COM> Roger> gaguiar@pinocchio.encore.com (George Aguiar-Consultant) writes: >Does anyone have an opinion on the above mentioned package? Is it >really better than hacking sendmail.cf directly? Is 2.1 the most up >to date release? Is there any other PD software that accomplishes the >same thing? ( ie. generate sendmail.cf file ) > >Thanks in advance, >--gaguiar Roger> One posible problem: it was written for use in the UK JANET Roger> network which uses big-endian addresses. However there is a Roger> switch to reverse addresses so they are little-endian for the Roger> rest of the world so it might work OK. It's not a problem, you can generate a sendmail.cf that knows nothing about JANET addresses, and it works. (If I remember right, it works internally with little-endian addresses and the JANET switch was just to let you recognize and convert incoming JANET addresses.) mike -- Mike Bayliss Chorus systemes, 6 avenue Gustave Eiffel, F-78182, St-Quentin-en-Yvelines-Cedex, France Tel: +33 (1) 30 64 82 37 Fax: +33 (1) 30 57 00 66 E-mail: mike@chorus.fr .... order correct the in done be to has really it when - NRS
jf@ap.co.umist.ac.uk (John Forrest) (01/16/91)
In article <MIKE.91Jan14143504@drums.chorus.fr>, mike@chorus.fr (Mike Bayliss) writes: |> >>>>> On 10 Jan 91 15:29:20 GMT, rknight@gec-rl-hrc.co.uk (Roger |> >>>>> Knight (B21)) said: |> |> Roger> In article <13738@encore.Encore.COM> |> Roger> gaguiar@pinocchio.encore.com (George Aguiar-Consultant) writes: |> >Does anyone have an opinion on the above mentioned package? Is it |> >really better than hacking sendmail.cf directly? Is 2.1 the most up |> >to date release? Is there any other PD software that accomplishes the |> >same thing? ( ie. generate sendmail.cf file ) |> > |> >Thanks in advance, |> >--gaguiar |> |> Roger> One posible problem: it was written for use in the UK JANET |> Roger> network which uses big-endian addresses. However there is a |> Roger> switch to reverse addresses so they are little-endian for the |> Roger> rest of the world so it might work OK. |> |> It's not a problem, you can generate a sendmail.cf that knows nothing |> about JANET addresses, and it works. (If I remember right, it works |> internally with little-endian addresses and the JANET switch was just |> to let you recognize and convert incoming JANET addresses.) I've forgotton what Uk-sendmail does now with the problem cases: cs.man.ac.uk uk.ac.man.cs or at least what they will do with "cs" in the top domain list. There are rules in the start-up phase that try to spot a reversed (grey-book) address, and reverse it if so. I can't remember which was round ambiguities work for. Of course, maybe you can disable all this? I'd recommend you use IDA anyway. John Forrest Dept of Computation UMIST