How do you separate the players? Anyone can give you connectivity or host a Web site. The standout providers will give you more reliability, enhanced helpdesk support and customized software. The truly exceptional ones will pass any protocol securely, using your choice of private virtual circuits to sites served within the network or an encrypted channel to outside locations. They'll help you develop a comprehensive security strategy, install firewalls when needed and monitor your sites against intrusion. These providers will have l ocal ISDN points of presence (POPs) to connect both standalone clients or network-based workgroups. Finally, they'll back it all up with service-level guarantees.

Granted, it's a tall order. Nobody yet has everything, everywhere, at the right price. And for most of us with substantially more internal than external traffic, the Internet is not a panacea.

There are many approaches that are inherently more reliable and less costly to secure than the Internet. These include frame relay, Switched Multimegabit Data Service (SMDS), private line or private IP. These networks can be easily supplemented with Internet connections when needed.

ISP Review Redux In this follow-up to our June 1 review of ISPs ("Domestic Internet Service Providers: Adapt Or Perish," page 107), we introduce offerings from AT&T, MCI Communications Corp. and Sprint, the three major interexchange carriers. We've also updated information on the traditional ISPs reviewed in that issue and created a new report card with all 10 ISPs. (For The Interactive Report Card ®, where you customize the weigh t of the features reviewed, go online to www.NetworkComputing.com.)

Plainly stated, the major interexchange carriers' legacy of market domination is threatened by the Internet. Cheap, flat-rate pricing combined with packet switching spells trouble for traditional leased-line and long-distance dial services. Ubiquitous, low-cost messaging threatens value-added networks and proprietary electronic mail hosts.

Why does a carrier embrace technology that stands to bring the cash cow to slaughter? Because it has no choice. I'm sure a few railroads wished they had built airlines in the mid-1900s. I'd ask them, but most are now gone, either gobbled up or forced out. The rules of communication and competition have changed, and continuing to apply the old models will mean diminished market share.

Of the three major IXCs, only Sprint has articulated a clear, cohesive strategy for servicing corporate intranet and Internet needs.

MCI's attempts to redefine its approach to the Internet have left network man agers largely confused. AT&T essentially is reselling BBN Planet Corp. services and offering a very basic Web server. Both carriers have introduced cheap dial-up services for their long-distance customers that have been met with crushing demand.

It doesn't have to be this way. The major IXCs can build on their strengths as backbone providers. They can capitalize on their successful data networks. Instead, however, MCI and AT&T distract themselves with the support-intensive consumer dial business, where rock-bottom pricing and substantial competition prevail. And for what? In the end, consumer Internet dial may be dominated by the local exchange carriers, bundled in our home telephone bills as call waiting services are listed.

So how do the IXCs compare as ISPs? The good news is they understand how to support corporate networks, have a wide coverage area and supply a nice menu of available options. The bad news is they're still charging a premium for dedicated connectivity, are slow to deploy IS DN and are thrashing resources between corporate and consumer demands.