Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!mnetor!uunet!husc6!hao!boulder!sunybcs!rutgers!princeton!udel!mmdf From: KEN%ORION.BITNET@wiscvm.wisc.edu (Kenneth Ng) Newsgroups: comp.os.minix Subject: RE: sending source code Message-ID: <628@louie.udel.EDU> Date: Sat, 24-Oct-87 01:28:44 EST Article-I.D.: louie.628 Posted: Sat Oct 24 01:28:44 1987 Date-Received: Sun, 25-Oct-87 22:39:01 EST Sender: mmdf@udel.EDU Lines: 38 >From: Bruce Becker >Uhhh... Ken - what's wrong with using uuencode/compress? Why reinvent the >wheel? These tools are commonly available - do they not meet the need for >BitNet transport? > >Cheers - BBecker Humber College Etobicoke, Ont. If uuencode and compressed worked in ALL cases we wouldn't be having this discussion. I am not aware of the latest version of uuencode, but one early mistake is that one character they took was the space, which some mailers strip (a so-called "feature"). Furthermore, uuencode, as far as I can tell, can not be directly translated on an EBCDIC machine. It has to be first moved to an ASCII machine. Granted, object code would not be very useful, but I believe it is desirable to view source code directly without having to move the file to an ascii machine first. As for compress, the research I've read indicates that the method I'm planning to use is generally better than compress. It is more CPU intensive however, from what I gather. What I'm planning also extends the information on files that are transfered: last modified date/time and record formats for example. Furthermore, I'd like to add the capability for sending several files at once. Before everyone says: "But use share", has anyone ever been burned by a trojan share that did "funny" thinks like removed all your files, send obscene messages to everyone on the system, or renamed all the files under your directory backwards? (e.g. '.profile' became 'eliforp.') Granted I think the last one required some imagination, and it was easy to reverse (just ran the program again), but I no longer trust executing a program off the net. And besides, not everyone has direct access to shell. Why should I discriminate against those who use VM, VMS, MS-DOS, etc? Remember, I'm trying to get my programs to work with the widest possible audience on the widest possible array of systems. Kenneth Ng: ken@orion.bitnet