From mdw@TC.Cornell.EDU Fri Apr 16 11:12:49 1993 Date: Fri, 16 Apr 1993 18:33:12 +0300 From: mdw@TC.Cornell.EDU Reply to: Louis Lagendijk To: linux-activists@joker.cs.hut.fi Subject: Rfmail released Newsgroups: comp.os.linux.announce From: etmelag@crosby.ericsson.se (Louis Lagendijk) Subject: Rfmail released Keywords: Rfmail, mail, news I have released my (working) version of Rfmail 0.5. I have ported it to Linux and hacked it quite a bit so it will work with the Binkley port already available for Linux. Rfmail is a fidonet to news/mail converter so it allows you to run your fidomail stuff under Linux. Rfmail also includes an own mailer, but I never tested it as I need an EMSI capable mailer. This rfmail version: - Runs under Linux - creates binkley style flow files - can autoamitically pack/unpack mail with ANY packer - handles multi-address setups quite nicely - creates and receives type 2+ packets.` I have rfmail uploaded as rfmail.0.5.01a.tar.z to halcyon.com and to tsx-11. After uploading I found that the sun happily (without error message) damaged the original tar, so gzip could not unpack it. I have it uploaded again with the name rfmail.0.5.01a.tar.z.again, but it should be stored under the other name. I could not reach sunsite, so I have not stored it there. Here is a part of the README on the patches I made: Rfmail 0.5.0.1 adapted for use with the BinkleyTerm mailer =========================================================== This version of Rfmail has been adapted to create Binkley Term style mail packets and flowfiles, improved packer hadling 4-d support and multi-address support. These patches are only tested on my own system (Linux 0.99), and even there they have only been briefly tested. So if this causes you to loose a lot of mail. If you decide to use this version of Rfmail you do that on your own risk, so if you loose all your mail, I will not be responsible. Binkley support --------------- In the configuration header file (conf.h) there is now a #define for Binkley. The Configure script knows about this define too. Defining BINKLEY will cause Rfmail to create flow files and mail files for Binkley, so there is no more need for fiddling with batch files. Rfmail does not yet know about Binkley style lock files. I am not sure what I should implement: Binkley style lock files (.BSY) or whether I should change Binkley to use Unix style lock files. I will probably use the Unix style. Packer support -------------- I have Rfmail also extended with support for multiple packers, so that you can receive and send mail packed with any packer. I have this working for ARC and LHarc right now (one of my bosses sends me my mail packed with zip, but gets Lharc, the other sends mail packed with old Lharc and gets mail packed with arc (yes, I know, I should clean this mess up). The idea for the packer handling comes from Squish that has a separate configuration file for the handling of packing of mail. Be sure to add mail to the archive when you call fpack again before mail is actually sent. See also the packer statement in rfmail.cf as well as the key statement. Multi address support --------------------- Rfmail now allows for using different aliases for sending mail and sending of echomail. People having addresses in alternets will like this. I also needed this as I am a multipoint: I have two normal fidonet point addresses. This option may appear to be a bit tricky: for echomail you specify the alias in the newsgroup definition (#
). The alias used for mail as well as the alias used in the packets is specified in the key statement. See the key statement in rfmail.cf. 4-d support ----------- Rfmail now can send 4-d packets. FTS-001 packets are still the default, but you can specify the packettype for different nodes or even networks in (again) the key statement in rfmail.cf. Packets according to FSC-0045 can be sent but not yet received, so this is kind of a joke. FSC-0045 has not been tested at all yet (Rfmail is actually not domain aware anyhow). Fido-style adresses ------------------- Rfmail has been changed so that adresses in the style: foo_bar@1.2.3.4 can be interpreted in the internet way, so the address would map to: foo bar at 4:3/2.1 or in the fidonet order, so that the address would map to: foo bar at 1:2/3.4. As said before, the patches are not extensively tested, so please do send comments/bug reports etc. Louis Lagendijk louis@nextjk.wlink.nl or 2:281/202.5 or 2:285/311.58