1987 - 92    P-3 Update IV

This is a story about connecting to the world and beginning to understand the fascinating topics of information organization, directories, and identity.  It is also about learning the power contained Horace's phrase: "Carpe Diem!"

Early versions of Novell Netware, out of the box, required the user to type lengthy and difficult to remember strings of characters to do the most simple things, such as printing to a shared printer, or save a file to a server.  Retrieving a file could be even more of a challenge.  Task one was to tame the network and make it easy to use.  Walt Pietrowski had started the process by writing some scripts in what were called batch files.  In effect, these added commands to DOS.  One would just type the name of your file at the command prompt, hit enter, and whatever script you had written would run.  This got rid of those long command lines for printing, but did not solve the file storage problem.  It also created a firestorm of paper waste.

Except for Lisa 7/7 from Apple, early personal computer operating systems did not provide printer drivers and a print service for applications to use.  Instead, every word processor, spreadsheet, database, or whatever, had to come with its own library of printer drivers.  Many of the apps made it fairly difficult to switch from one printer to another.  So what would happen with a high degree of frequency, is someone would setup their Netware command to print to a dot matrix printer, and then send it a Postscript stream intended for a laser printer.  The result was many pages of useless paper filling the recycle boxes.  So, I wrote a little primitive GUI using extended ASCII graphics characters that would put a map of the building on the screen showing where all of the printers were.  The user would pick a location, and the map would change to a picture of the various printers at that location.  Next they would pick the printer, and then my program would not only setup the proper settings with Netware, it also patched all of their installed software such that the proper printer driver was in place.  Over night, the paper waste problem was eliminated, along with the time people had been losing, and the frustrations they had been feeling.

Then I wrote a program displayed the information tree for the whole P-3 program based on the org chart.  Each function had its own data storage area, and could design its own shared file system.  My program leveraged a commercial package called "X-Tree" to set up all of the DOS path commands so that from within their applications they could save and retrieve files from any shared location on the program where they had permission.  This turned our seven Netware servers into what was effectively a cloud service.  This was 1988, a full year before Tim Berners-Lee setup the first website and put out a primitive browser.  But our really big accomplishment came the next year with our Higgins email system.

Boeing Computer Services had been running two mainframe based systems.  One was an IBM Profs system (later renamed Office VIsion).  Profs was being used by a couple thousand people in the commercial airplane part of the company.  They also had a DEC All-In-One system for the exclusive use of BCS managers.  The two were joined by a Boeing-written mail switch called BEMS, for the Boeing Electronic Mail System.  The engineering communities within Boeing were using simpler systems that were native to the VAX and Sun computers they had.  The VAXes were mostly on the defense side of Boeing, and were also connected to two external networks: DECnet and the Internet.  Every defense contractor and military R&D site were also on both of those.   Profs was the worst of the worst of these early messaging systems because it was tied to a user name and addressing system from the 1960s that was limited to an 8 character user name and an 8 character host name.  There was no concept of routing beyond your own private collection of machines.  Routing was done with manual entries in a directory database, and transport protocol translation inside the mail switch or gateway.  Given the process and budget approval process one had to go through to engage the necessary support labor required to send a message to a recipient not yet listed in the directory, BEMS was arguably slower than the Pony Express.  Calling email was being generous.


For our PC-LAN system, Ralph's favorite technology scout, Verle Helsel, had gotten us a Higgins system, which was modeled after Profs, but run on Novell severs for PC clients.  Ed Owens, the genius developer behind Higgins had provided it with a totally flexible user name and addressing system.  Except for the needed gateways, it was network ready out of the box.

A little bit of a caution is needed here.   While Higgins equipped with gateways could in theory send email anyplace, in those days of the leased line Internet with no carrier support of involvement, one could not simply specify the short form of an SMTP (internet) address.  The entire route the packets had to travel had to also be specified.  Sometimes the 80 character address length limitation that was common in those days would not be enough.  So we supplemented Higgins with directory support too.  But this was done locally at the department level, and could be implemented in a few minutes instead of the weeks required by the BEMS process.

In February 1989 I was at the Network World conference at the Dallas Infomart.  During the Higgins user group meeting I met Lih-Tah Wong, an MIT trained engineer, who was writing Internet (SMTP/MIME) gateways for all of the early Lan mail systems.  There was also a company at the show who had a gateway to the mainframe systems using IBM's SNADS protocols.  Two months later I had both of them up an running on our system.  I might as well have fired a cannon in the halls of one of Boeing's executive office buildings.  Let's just say it got noticed - big time.

While a cc:Mail system on the Commercial Airplane side of Boeing had beaten me to connecting to Boeing's central mail switch, the connection to Navy Labs was historic.  It was an application on the everyday user's desktop that was leveraging the Internet.  Software engineers with access to Sun and Vax equipment had been doing that for a couple years, but this was something new.  It was the very first time a Boeing executive had email connectivity from his desk to a customer executive.  And, that customer was the admiral responsible for the procurement of the P-3 Update IV system, making him the first flag officer in the US Military to have that sort of connectivity with an executive with one of his suppliers.  So we started getting visitors again, including AR&T.

The management culture shock that was generated by making this sort of connectivity visible and easy to use cannot be overstated.  My in-laws from St. Louis were visiting shortly after this connectivity was accomplished.  My father-in-law was a quality manager with Monsanto and I told him that if he would give me the name of his VAX system administrator, I could work with him to set things up so we could exchange mail directly.  He was horrified and explained that before any two companies could exchange data, that the lawyers for both had to negotiate a formal data exchange agreement.  The idea that two people could just initiate a correspondence on their own was actually frightening to him.  Of course, that's the way the world works now, and all one needs is the short form of someone's email address, and viola, exchanging notes can happen.  By sometime around 1993 or 94, it started to be common for people to add their email address to their business cards, and for some of us, we did that and kept our phone numbers to ourselves.  LAN based email changed the world.


One day my friend Doug Selix brought Joe Muldoon, an AR&T manager, over for a tour.  At lunch Joe offered me a very interesting job.  His exact quote was: "How would you like to do it again for the whole company?"  I was startled and asked him who he thought he was, T. Wilson?  T. was the CEO of the company.  Joe said no, just that he was a guy with budget who could provide me political cover.  And besides, he pointed out that following the company's rules about computing had not gotten in my way here.  I had to agree that he had a point, so I accepted.  It took about a month to make the transfer, most of which I stayed at home for the birth of my daughter and helping reorganize our home life around that change. When I went back to work it was directly to AR&T.  So after nine years that was then end of my stint on the defense side of Boeing.

There is one other curious thing that happened at that day when Doug first introduced Joe and me.  Doug was clearly trying to sell Joe on the idea of offering me a job.  During the conversation while I was showing Joe around Doug described me as a bulldog, and said that once I sunk my teeth into something I didn't let go.  I've thought about that description a lot over the years.  There is more truth in it than I sometimes like to admit.

I need to mention two more things that happened while I was on the P-3 Update IV program.  The first item deals with our relations with Boeing Computer Services.  BCS had the purchasing function for most things related to computing.  We could buy some things directly from the vendors and have program finance pay the vendor directly.  But for items from vendors with an existing vendor relationship with BCS, we had to go through BCS Purchasing, much to their dismay.  The BCS manager caught in the middle, who had to process our orders was Kay Gagnan.  Kay would meet with us, at least every couple of weeks. 


I liked Kay but Ralph couldn't get past his dislike of the whole organization.  In one of those weekly meetings when we were discussing why it was simply impossible for us to work through BCS I happened to ask Kay why it was that BCS never measured itself by the impact their services had or lacked on the internal Boeing customers.  My question put her in an untenable position, and I was convinced she was sympathetic to our plight.  Her response was interesting, and neatly summed up the whole problem with the top BCS leadership.  Measuring their real effectiveness was too hard, so they left it to their customers.  Shortly after that, Kay and Ralph both retired, leaving me in charge on the program side.  Kay is now my neighbor on Lopez Island.  After she left, BCS hired Joe Madrano.  Joe later told me that his explicitly stated assignment was to keep an eye on me. 

The other thing that happened on the P-3 program that I think worth mentioning occurred toward the end of my first year there.  One day Ralph called me into his office and asked me to close the door.  I figured he had dreamed up some new project.  But he turned off his computer and looked at me across his desk.  He said that another one of my jobs from that point on was to help him be a better person.  That was astounding.  Imagine a company with a culture in which your boss would say something like that.  But, that actually was not all that strange in the Boeing of the totem.  Things were different in those days.

The Boeing totem logo shown on the patch below was the symbol of the company from its first appearance painted on a Boeing Model 221 Monomail sometime around 1930.  We stopped putting it on our planes in the late 1950s, but it continued to be used in all other situations requiring a simple logo until the summer of 1997.  The Boeing of the totem was a distinctly different company, culturally, compared to what came afterward, when a stylized derivative of the Douglas World Cruiser logo, which had gone through many revisions, was adopted.