After over 22 years of working on VM and my MailBook email software, I have decided to stop.
I recently ran across the following email that I sent out in 1989(!) that I think is worth repeating here one last time: ============================================================================ ============================== Date: Tue, 5 Dec 89 20:49:40 CST Reply-To: MAIL/MAILBOOK subscription list <[EMAIL PROTECTED]> Sender: MAIL/MAILBOOK subscription list <[EMAIL PROTECTED]> From: "Richard A. Schafer" <[EMAIL PROTECTED]> Subject: History To: MAIL/MAILBOOK Mailing list <[EMAIL PROTECTED]> I suddenly realized I had let something slip by without notice. (Those of you with the full commented source to the code, go take a look at line 11 of MAILB00K XEDIT.) For the rest of you, that line says that the original date of the code is October 16, 1984. That means I've now been working on this package for ******** 5 YEARS ******* ! For those of you who haven't been around that long, you might find a little history of the project interesting. The predecessor of the current code was written at MIT by a fellow named Dave Burleigh. That's why you may occasionally see a reference to "MIT MAIL". What follows is a note from Dave he wrote me in 1986 when I was working on a presentation to SHARE on the package. (If anyone cares, I think this would have been SHARE 66; the presentation was published in Volume I of the proceedings.) Subject: Re: MAIL history To: Richard Schafer <[EMAIL PROTECTED]> In-Reply-To: Your message of Tue 11 Mar 1986 16:59:36 CDT Glad to. I'd like a copy of your presentation, if possible, too. I began working on MIT MAIL's earliest ancestor sometime in 1981, as soon as I had a full screen terminal in my office. I was dissatisfied with the line-oriented MAIL program then in use at MIT, and wrote a simple EXEC/XEDIT macro to let me prepare my mail message in full-screen mode and send it out. That program displayed incoming mail in line mode, and didn't include much power for processing incoming mail. Since other staff members were used to finding little goodies on my disk, my mail program began to find its way onto other people's disks, and soon became the standard mail program used by the CMS programming staff. I began to get suggestions/requests/demands for extensions. I had been doing a lot with the Display Management System (DMS) at the time, and decided to put it to work in MAIL for displaying incoming mail in full-screen mode. I added subcommands for replying, forwarding, logging, etc. After realizing that the MIT-Multics machine could act as a gateway to Arpanet, I studied up on the RFC822 standard for mail headers and altered MAIL to recognize and build standard headers. And then there was BITNET. While the BITNET standard for mail headers was practically undefined at the time, the work involved in being able to handle incoming BITNET mail and Arpanet mail was starting to overwhelm me. I decided it was time to make MAIL into a legitimate programming project, and approached my supervisor about scheduling some time for me to tackle it (it had been a "spare-time" project heretofore). Since he was the author of the line-oriented mail program I was replacing, it was very big of him to endorse the project. Although REXX had just arrived on the scene, my initial test with it suggested that a REXX version of MAIL would be slow and expensive to run, and I decided to reprogram in EXEC2. I regret that decision now, but we were starting to hear complaints about how much MAIL was costing to run, and I knew that the added functionality I had in mind for MAIL was going to make it even more expensive. I began rewriting MAIL in August 1983 as a Read/Send/Menu mode system, much as it appears now. It was installed for public use in late September or early October 1983, was well received, and soon began to make its way out to other BITNET sites. It underwent a great many revisions in response to suggestions and bug reports, mostly from BITNET users. Particularly helpful and influential in MAIL's early development were Richard Schafer (RICE), Bill Rubin (CUNY), and Hank Nussbacher (CUNY and WEISMANN). In January 1984 I left MIT for a contract programming venture. I continued to work on MAIL during most of that year, but found I could commit less and less time to it. I finally had to pass the baton to Richard Schafer, author of MAILBOOK, who courageously agreed to take on the project. ******* I didn't know what form you wanted this history to take. This is quick and dirty, off the top of my head. Feel free to edit or extract from it as you like. Thanks again for all your hard work, and good luck at the presentation! Dave I got started in 1983 or so working on what at the time was merely an adjunct to the MIT MAIL code, and what became the MAILBOOK portion. This is the part that has a date of 1984 on it. (As I go back to Dave's note, I realize that the code must pre-date that October, 1984, but I have no clue what the true date should be.) As Dave took off for greener pastures towards the end of 1984, I agreed to take over the project, which had (if I can remember correctly) around 35 to 50 clients at the time. (To give you a feeling for the difference, the number is closer to 475 today.) For a while, I maintained his code in EXEC2, but soon realized that a total rewrite (in REXX) was absolutely necessary, and began the grand design that merged MAIL and MAILBOOK into a single piece of code, which eventually came out sometime in 1985? (Does anyone remember when? I'm not sure I know any longer.) Well it's been through a lot of work, design, thought, and mistakes since then, with each new version taking a lot of its changes from comments (and code) submitted by all of you, particularly people like Nick Gimbrone and Bill Rubin (but not to slight all of the rest of you). With any luck, the next version should be this year's Christmas present. After that, who knows? I've gots lots of ideas that I stuck off for the version past this one, since I was committed this time to doing something about the need to load the entire file into memory to work with it, and that has taken me about 8 months to do, with some *major* rewrites of large portions of the code. (As an example of that, the source update files for versions 88.01.00, 89.01.00, 89.01.0A, and 89.01.0B for MAILB00K XEDIT, the mainline code, total 1216 lines of updates. The files which take me from 89.01.0B to 89.02.0A for the same file total 2174 lines of updates!) It's been a fascinating 5 years, during which the world of BITNET (and EARN, etc.) has changed more than we might have dreamed, not to mention the world at large. I've had a lot of fun doing this, and expect to keep going a while, yet. ============================================================================ ========================= Well, that was 1989. By 1994, the program had gone through several more versions. The only ones I have records of for sure are the 8902 and 9201 versions, but there may have been more. And from 35 sites using the software, the distribution had gotten up to more than 600 sites, on every continent except for Antartica. In 1994, after 10 years of giving away the software, I completed another massive rewrite and restructuring of the code from the massively monolithic Ricemail version into a new multi-module version. Rice University graciously agreed to my ownership of the software and agreed I could commercially market the software as long as it no longer referred to Rice. So I began marketing the program under the name of MailBook and successfully convinced large numbers of the institutions who had been getting Ricemail for free to pay for a commercial license for the software. I continued to develop the code over the past 13 years since the first commercial version, with 7 major releases during that time, most recently in 2004. Now it's 2007 and the time has finally come to say goodbye and put this project to rest. While I've thoroughly enjoyed working with VM and MailBook over the years, I truly no longer have the time to devote to either any longer and feel it's time to quit. The MailBook code is mature and stable and bug reports have been rare for years. And the number of people still doing email on CMS has continued to dwindle. I am exploring the possibility of someone else taking over the software, but those talks are in their earliest stages and may not go anywhere. The past 22 years of software development work on MailBook and my involvement with the VM community have been exciting, fun, and a tremendous experience. I have learned much in the process and have enjoyed the contacts I've made with VM and MailBook users over those years. And I think I made a useful contribution to the community with this work. I put aside that time with a certain nostalgia, but with a recognition that the time has come to move on. As some of you may know, in 1995 I left Rice University and became an attorney, beginning a practice in intellectual property law where I am now Senior Counsel with Akin Gump Strauss Hauer & Feld LLP. So while this portion of my life is ending, others are continuing in exciting ways--I'm not ready to retire yet. You can continue to contact me for the foreseeable future at the same postal and email address or at my law practice address listed below. I would love to hear from any of you at any time. But for now, So long. Richard A. Schafer MailBook 6632 Fairfield Drive Houston, Texas 77023 t: 713-921-1433 f: 713-921-1366 or Richard A. Schafer Akin Gump Strauss Hauer & Feld LLP 1111 Louisiana Street, 44th Floor Houston, Texas 77002 t: 713-220-8184 f: 713-220-2384
