Reponse to Lucid Lost

Editor, Unix Review

Dear Sir:

I read with interest your editorial in the December 1994 issue entitled “Lucid Lost.” I supposed that because my name was mentioned I would be allowed to make some remarks on it.

Lucid had many faults, but none of the ones you mentioned had much to do with why Lucid went out of business—especially not the dramatic way it did. Whether Lucid would have fared well against its competition will never be known because Lucid failed for silly business reasons. At the time Lucid shut down it had healthy enough sales to sustain a company of about 50 people and, in fact, had just posted a record quarter in Energize sales. The sales were clearly on the increase, and the tide seemed turning in its favor. At OOPSLA last year, for example, many (former) customers approached me and lamented the demise Energize and its technology—telling me that the remaining environments and tools just aren’t up to the task.

No, the real reason Lucid went under was poor business decision making. After the CEO was fired in April by the Board, the Board was somewhat unpleasantly surprised to learn that several of the otherwise ordinary pre-paid royalty Lisp porting contracts had pay-back clauses. Normally such contracts have a walk-away clause that says that if revenues don’t materialize, the risking parties just walk away; but these contracts required Lucid to pay back the pre-paid royalties. Most people call such contracts “loans”. These liabilities along with the bank loan amounted to enough debt that the venture-capital investors decided to give up and merely shut down Lucid.
You see, companies like Lucid—companies without a hit product—always have a tough time making progress. Some CEO’s try to husband resources to stay in business as long as possible—others swing for the fences.

Nine months before Lucid folded, the CEO called for a layoff to save enough money over the remainder of the fiscal year to stay afloat no matter what. When you lay people off your expenses spike that month, and the month after you start enjoying the savings. Lucid hired enough people the third month to get expenses back up to where they were before the layoff. When you averaged it out, it was as if no layoff had ever occurred.

When you lay people off, you had better make it count because it’s tough all around—on the people laid off and their families, on the people who stay, on the fabric of the company itself.

I think it was this failed layoff that caused the Board to fire the CEO—though I was Chairman of the Board, the investors tended to act like the real Board and people like the CEO and I (insiders) were not taken seriously.

Lucid would have had a tough road even had these mistakes not been made, and it isn’t clear that Lucid would have survived as long as it did without the “loans” the CEO secured. Energize was an overly ambitious undertaking: we were fighting the hardware vendors, we were fighting Lucid’s business persona, we were fighting our own history.

We designed an architecture that could work with a simple protocol that any C/C++ compiler could implement (in fact, it could have been a compiler for any language from FORTRAN to Lisp to Smalltalk), but we didn’t have the influence or clout to get any hardware vendor signed up with their compilers, so we had to buy a compiler company. It’s hard to assimilate another company’s culture, but we did it. The compiler that company produced is possibly still unsurpassed in compiled code size and speed.

Lucid was originally an OEM company—providing Common Lisp to hardware companies who marketed and sold it. We did this quite well, but because we could not OEM Energize we had to go direct. This is a very hard transition to make, and we did a poor job at it.

The technical design we came up with for Energize was really too ambitious, and it wasn’t until the second release that most of the specifications were met. We were unable to convince the venture capitalists that they needed to fund the full development effort so we tried to get away with doing part of it and funding the remainder with Energize sales and by cash-cowing the Lisp business.

You said that one of Lucid’s failings was its “inability to move out of the shadow of its founder, Dick Gabriel.” I couldn’t figure out whether the shadow you thought I cast on Lucid was good or bad, and I guess I don’t really care which. I made plenty of mistakes—two large ones were the CEO I hired who ran the company into the ground and a development manager who drove excellent developers away—but I believe we built something people will remember. Last September we had a party—sort of a wake—and nearly 70 ex-employees showed up to celebrate the company we made. The company spawned friendships that will persist for years, a number of marriages between employees, 3 or 4 rock bands, births, a death.

Once we did a layoff and several of the laid-off people continued to work for free. The CEO told me he couldn’t understand how people could be so stupid as to work for free for a company that essentially fired them—perhaps he wouldn’t have understood, either, how my managers and I sat with each laid-off employee and literally cried over the terrible news that we could not work together anymore.

The technology we developed will have a lasting effect: Lucid Emacs and ILD, the incremental loader, are being used by various companies in future products—perhaps even by Sun.

I know that we put together a development team unsurpassed in the Valley and built an institution that outlives the foibles of its CEO.


Dick Gabriel