By

Julian Dibbell

11/27/95

Time Magazine

Page 81

© 1995 Time Inc.

TO THOSE FEW SOULS ON the planet who haven't yet wrestled with giant computer-operating systems like Microsoft's Windows 95, memory compression probably sounds like something that starts to set in around the fourth margarita. To the rest of the computer-using public, though, it sounds like a dream come true. Imagine: instead of shelling out hundreds of dollars to feed your system's insatiable appetite for memory, you simply run a $30 program that squeezes the necessary storage space out of your existing chips, instantly increasing their effective capacity by a factor of two or more.

It was on the promise of such memory-compression technology--and a single product, SoftRAM 95--that a small Southern California company called Syncronys Softcorp rose this summer from start-up obscurity to the top of the software best-seller lists. Its glittery success in software sales was matched in the financial markets, where in over-the-counter trading the company's shares skyrocketed from 5 to 32.

But Syncronys may be more than just another high-tech success story. According to a report published last week in the German computer magazine c't--a report vehemently denied by the company--Syncronys' promise of memory compression may have been just that, a promise. The magazine's analysis of the program's code suggests that the reason SoftRAM 95's technology does not work as advertised is that its technology does not exist.

That isn't to say memory compression itself is a pipe dream. On the contrary, the technology is real, and Syncronys is not the only company to have cashed in on its appeal. But neither of SoftRAM 95's top competitors--Connectix's RAM Doubler and Quarterdeck's MagnaRAM2--can boast sales approaching SoftRAM 95's 650,000 copies. And while some have complained about the speed of the other products, neither has been targeted for the kind of virulent attacks mounted against SoftRAM 95.

For if the harshest of these attacks are true, SoftRAM 95 isn't just one more computer product that fails to live up to its hype; it's a hollow piece of Potemkin programming, devoid of the advanced, patent-pending compression technology touted in its packaging. In short, says Mark Russinovich, a University of Oregon computer scientist, "the thing is a fraud."

Suggestions that the program had problems began to circulate in late September, when a study funded by one of its competitors showed that some applications ran more slowly with SoftRAM 95 than without. But it wasn't until last week, when c't broke the story, that the most damaging accusations came to light.

This was the second article in c't on the program. The first, in October, called it "placebo software"--a catchy appellation that triggered a libel suit and a temporary restraining order. In response to the court challenge, c't's editor, Christian Persson, and one of its writers, Ingo Storm, engaged the services of a software engineer, and together they went through the program line by line to try to plumb its inner workings. Their findings: the one patch of SoftRAM 95 code remotely resembling a compression algorithm never gets used by the program. Moreover, the two subprograms actually called on to manage memory usage appear to be copies of programs that Microsoft hands out free. Both modules increase a computer's capacity using "virtual memory," a memory-expansion technique considerably more time consuming than true memory compression.

"It is complete nonsense," says Syncronys president Rainer Poertner of c't's charge. "The compression algorithms are at the heart of our technology." Does SoftRAM 95 actually compress data held in memory chips? "Absolutely!" he insists. Poertner points out that a recent customer survey conducted by Dataquest, a market- research firm, showed that most SoftRAM 95 owners are quite satisfied.

But at least one other researcher has repeated Persson and Storm's software dissection and confirmed their conclusions. "I found exactly what everyone else found," says Oregon's Russinovich. SoftRAM 95's main component, he says, is "little more than copies of the sample programs provided in Microsoft's development kit. The differences are cosmetic."

Meanwhile, the company has raised eyebrows for other reasons. According to the New York Times, Syncronys was not subject to lengthy Securities and Exchange Commission scrutiny or the high cost of an initial public offering because it merged with a corporation that had already gone public. Yet on SoftRAM 95's box, the company lists its ticker name as nasdaq:sycr, even though its shares are still not approved for trading on the national over-the-counter exchange. (Syncronys' stock trades on the OTC bulletin board, which is supervised by NASDAQ.)

Against all odds, Syncronys still has one group of loyal supporters: those legions of customers who continue to report their satisfaction with SoftRAM 95. And the indomitable Poertner remains unbowed. He can be seen regularly on the Internet fiercely debating his critics. He also turned up last week at the giant Comdex computer trade show in Las Vegas to announce new versions of SoftRAM, including one for the Macintosh.