From: Steve Kelley (SKELLEY@flowcyt.cyto.purdue.edu)
Date: Wed Feb 11 1998 - 10:13:03 EST
In principle, the bounce messages should only come to me. There are headers in the mail messages that *should* be honored by the software that receives the mail on the other end, that would restrict error notification to me as the actual sender, rather than the original author. Unfortunately, the 'state of the art' in E-Mail software is absolutely abysmal, from top to bottom, first to last, simple to complex. It pretty much all stinks, because it pretty much all assumes that the users want to use their email exactly the same way the author of the software does. Some clients honor the error headers, some don't. Some handle attachments well, some don't. Some handle word wrap automatically at the reader, some try to impose wrapping at an intermediate stage, some ignore it totally. The fundamental transport agents, the programs that get the mail from computer to computer, are a little better, but the main program used by most system administrators was designed back in the late 60s and has been incrementally (perhaps accretionally is a better term) 'improved' since then. And that, I think, is the main problem with E-Mail software in general. It is accreted, rather than designed, and unless all the various authors of all the various clients start thinking similarly, we will continue to have problems. Now the specifics. The bounce problem occurs because the original author's address is in one of the headers of the message. The *only* way to completely solve the bounce problem, is to completely remove the author's address from all of the standard headers. This would make it *impossible* for any client to ever 'reply' to an individual - the reader would have to make a conscious decision to restrict his or her reply, then either type or cut-and-paste the author's address (an error-prone operation). For various reasons, we have decided to keep the author's address in a standard header, therefore, some software will continue to send a copy of bounce messages to the author as well as to me. What I will do, is to be much more aggressive over the next few weeks in weeding out addresses that fail. I have been very hesitant about pruning the address list, since many 'failures' are due to temporary conditions. I will try to be careful about which addresses get removed, but if you find yourself not receiving cytometry mail suddenly, please let me know. Steve Steve Kelley kelley@flowcyt.cyto.purdue.edu Purdue University Cytometry Laboratories (765) 494-0757 -- voice B050 Hansen LSRB, Purdue University (765) 494-0517 -- fax West Lafayette, Indiana, 47907
This archive was generated by hypermail 2.1.6 : Thu Jan 01 2004 - 17:35:13 EST
