jw@pan.UUCP (Jamie Watson) (06/14/88)
First, please believe me, I am not trying to start a religious war here. I am likewise not trying to demean Dave Taylor's efforts in developing elm. I am simply trying to pass along a very good piece of advice that I got not long ago. I posted a question to this group asking what the latest version of elm was, and where to get it. I wasn't the first to post such a question, nor the last. I got a couple of kind offers to send 1.7 source, one of which was quite interesting because it said 'I'll send you elm source, but why don't you take a look at mush?". I accepted on both counts, he sent me elm 1.7, and I extracted the source and patches for mush from comp.sources.unix and tried it. After using mush for a couple of days, I decided not to even bother with elm 1.7. My reasons for this decision were the following: - There is a huge amount of ongoing confusion over what the latest version of elm is, and now what its legal status is. Although I make the decisions here about what software we use, I feel a considerable obligation to my employer to avoid brining in software of questionable origin. I'm not accusing anyone of any wrongdoing, but I don't think anyone could deny right now that the origins (and future status) of elm 2.0 gamma are certainly questionable. - I have very little confidence that the confusion over "official" versions, patches, new releases, etc. of elm will get better in the future. I know a search is currently underway for a single person who to take responsibility for future updates to elm. I also know that this is at least the third time this search has been conducted. Obviously, the results of the first two were less than optimal. - After trying the latest release of mush, I found that it had all of the features I personally liked in elm, and after having the other programmers here try it, I got the same response from them. I am not saying that mush does everything elm does; Dave Taylor did an excellent job of developing auxiliary programs and interfaces to elm, most of which don't exist for mush. Also, elm appears to be more usable for non-programmers or better non-Unix specialists than mush. But I don't have any of those to worry about, so that doesn't matter to me. - I find that mush is much better integrated with Unix than elm. First, as a very small example, it updates the headers of mail messages with Status: lines correctly, so my other utilities, such as checkmail/newmail, work correctly. Also, with its shell-like command syntax and pipes mush lets the user make mush better use of other unix utilities with their mail. - For those who have Sun systems, mush has a tool mode that works with the Sun windowing system. For the poorer masses, it has a curses mode that is somewhat like elm (but the command syntax is, of course, different). For experienced Unix people, mush has a shell mode that is nothing short of wonderful in its power and flexibility. - Finally, the author of mush (Dan Heller) is still very active in support, development, and *control* of his program. This counts for a lot with me, as I firmly believe that no one else will ever understand a program, its code, and its philosophy, as well as the original author. This makes for more stability, better fixes, and better development and enhancement. As a final note, I should mention that I looked at the original posting of mush quite some time ago, and decided not to fool with it because it seemed too dependent on Berklix features, and too dependent on sendmail style mail delivery agents. The newest version of mush (6.2) has solved both of these problems very nicely. The code compiles for me without a hitch on SysV.2 and Ultrix 2.0, and runs just fine. The Ultrix system has sendmail, the SysV does not, and mush works with both of them quite well. jw