Loudcloud is loud simply because the company was founded by Marc Andreessen, the Internet legend who led the team responsible for the first graphical Web browser. After Mosaic, Andreessen went on to co-found Netscape; the rest is Wall Street legend. Andreessen's presence guarantees volume.

So where does the "cloud" part come in? Well, at the start there was a certain amount of mist or fuzz surrounding exactly what it is that Loudcloud does. Andreessen's early statements about the company were deliberately vague. Today Loudcloud calls itself an "Internet infrastructure company" -- and that term is a kind of Rorschach buzzword, meaning you're free to project onto it anything you like.

Remember the passage in "Hamlet" where the prince tweaks Polonius by insisting that a cloud looks like a camel, then a weasel, then a whale? In its short history (the firm is about a year old) Loudcloud has at times been equally hard to pin down. Today it offers clients "solutions" in the form of a "suite of services" that "addresses the challenges of deploying, maintaining and scaling mission-critical Internet operations."

Now that the company has filed to go public, however, it has been required as part of its IPO to issue a disclosure statement known as an S-1 filing. The S-1 typically includes a description of a company, full financial statements, stock breakdowns and a section on "Risk Factors" outlining everything that could go wrong for the firm -- every potential mishap is laid on the table so that investors aren't, as it were, blindsided by any downsides.

It's this odd juxtaposition of hopeful business plans and disastrous doom scenarios that make S-1 filings the signature documents of digital-age finance. S-1s are capsules for all the new economy's dreams and nightmares -- the promise of instant fortunes based on technological innovations, and the threat that they might all vanish in a flash.

Loudcloud's S-1 makes for fascinating reading. For one thing, at this stage of the digital economy's new sobriety, the company's IPO plans sound like a nostalgic throwback to the heady times of late 1998 and 1999. Loudcloud has already raised about $188 million in private rounds and venture capital, and it hopes to raise as much as $150 million more in the IPO. In the six months ending July 31, it reported revenue of $1.941 million and a loss of $49 million.

All that deploying, maintaining and scaling of mission-critical operations doesn't come cheap.

Loudcloud promises its customers that it will take all the messy work of configuring and maintaining arrays of servers off their hands and guarantee 100 percent "uptime" -- or they'll get their money back. Andreessen's company doesn't itself build and maintain massive "data centers," the secure, bandwidth-fortified facilities operated by firms like Exodus and Global Crossing where industrial-strength sites "co-locate" their servers; instead, Loudcloud rents space at such centers and builds a whole back-end setup for its clients there -- software, hardware, the works. It offers its own technology, with names like "Opsware" and "Smart Cloud," to streamline and automate the work involved, and touts its ability to get companies on the Web quickly, flexibly and with room to grow.

For corporate executives whose heads swim at the expense and difficulty of attracting engineering talent and building Web operations from scratch, Loudcloud's pitch is doubtless attractive. How well Loudcloud will deliver on these promises is something no S-1 can tell. But how well the market receives Loudcloud's stock launch will provide a noteworthy benchmark for the mood of the manic-depressive tech-investment world.

Today, while the stocks of e-commerce and so-called B-to-C (business to consumer) companies have been widely forsaken, the infrastructure companies among whose ranks Loudcloud pitches its tents have fared better. Trouble is, who are the customers of these infrastructure companies? Other Internet firms -- mostly those same forsaken e-commerce and B-to-C companies.

That's why, among all the other risk factors in Loudcloud's S-1 -- like earthquake threats, fleeing executives and the possibility that people might just get tired of the Net -- this point really stands out: "To the extent we continue to rely significantly on Internet-based customers, we will be subject to an increased risk that these customers encounter financial difficulties and fail to pay for our services or delay payment substantially." The same paragraph notes that in the most recent quarter, 51 percent of Loudcloud revenue came from Internet-based companies.

For investors who have decided to shun last year's darlings in e-commerce, infrastructure companies may look like a safer haven. But the Net economy is truly interdependent -- a big web of financial symbiosis. There's no great opportunity in a company like Loudcloud unless there is a growing market for it to serve. So far, the Net has been one big boom; but if that changes, the demand for infrastructure will wither, too.

Today's Internet market has supposedly matured to the point where it is more interested in the "path to profitability" than in the coolest technology or the biggest market share. Loudcloud has been able to make lots of noise, thanks to Andreessen; the question investors presumably will ask is, Can it make lots of money?

If they don't ask that loudly or clearly -- if they simply shower Andreessen and company with dollars -- then brace yourself: We may be in for another round of Internet stock mania. Who'd have thought that the path to profitability could circle right back to the amusement park of speculation?

The Free Software Project
Read Andrew Leonard's book-in-progress on Linux and open source -- and post your comments.