The ongoing saga in Florida has certainly helped illuminate three large and basic problems with how we cast and count votes in this country.

The first problem is that we take so long to count votes that a huge industry of exit polls and voting projections has developed in the news media, as part of a competition to "call" races before, and sometimes long before, the results are actually tabulated. The serial decisions by television networks on election night to call Florida for Al Gore, then un-call it, then call Florida (and the presidency) for George W. Bush, then un-call it still again, have been an important part of the conflict and confusion surrounding this election.

The second problem is that the kind of ballots we use often confuse voters and make it difficult for them to cast a valid vote as they intend. The infamous Palm Beach County "butterfly ballot" is the best example, but virtually all current ballots give voters no chance to correct inadvertent errors.

And the third problem is that today's ballots are often difficult to accurately count or recount, either by machines or by election officials. The tragi-comedy being played out with efforts to count punch-card ballots in parts of Florida, where machines miss thousands of votes and hand-counts must deal with hanging, pregnant or dimpled "chads," are proof enough that we are dealing with a less than ideal system.

Most of all, the Florida dispute has called into question our peculiar reluctance to update the technology we use in voting, even as we update the technologies we use in every other part of our public and private lives. Are any organizations of any type other than election units still using punch cards designed for the very first generation of computers? Not that we know of.

All three of the "Florida problems" -- which actually exist all over the country -- could be addressed at once if we moved to a system of "e-voting": casting ballots electronically on user-friendly screens, with a paper-receipt print-out for election officials as a backup.

A variety of software already exists for letting voters at polling places cast votes on a laptop, with user-friendly, consecutive screens similar to those used by ATMs, providing opportunities to confirm or correct votes before completing the ballot, and counting the votes electronically.

If this kind of system were used everywhere, you would have (1) almost instantaneous voter counts, instead of a laborious process of machine-counting ballots, making the whole exit poll/projection system completely redundant; (2) a sharp drop in voter confusion and mistaken or voided ballots, and (3) an even sharper drop in the need for recounts (though a paper receipt kept by election officials should be provided in case of a system "crash" or some other hardware or software problem).

The main objection to this kind of system (other than the possibility of "crashes," or viruses, or hacker attacks, which can all be taken care of by the paper print-out backup) is cost. That is why the Progressive Policy Institute's Robert D. Atkinson has proposed that Congress and the states enter into a financial and technical partnership to make national e-voting a reality within the next few years.

But Atkinson also thinks on-site e-voting is simply the first step towards online voting from home, which would combine the swiftness and accuracy of e-voting with the convenience and ability to make voting more deliberative associated with the growing use of liberalized absentee voting and other forms of mail balloting. The key issues in remote online voting are network security and potential fraud.

Atkinson proposes that the National Institute of Standards and Technology should head up a cooperative effort with states and the private sector to address the voting security and privacy issues, recognizing that no system can be 100 percent secure (as this year's cases of lost paper ballots in various localities has shown).

The key to dealing with the voter fraud issue in remote online voting is to move forward nationally with a system of digital certificates that citizens can generally use to authenticate their identities online. PPI has proposed that state Departments of Motor Vehicles issue such certificates when citizens obtain or renew drivers' licenses.

But we should not wait for a national solution to every problem with remote online voting before taking the first step to e-voting at polling places. To be sure, situations like those we are experiencing in Florida may only come around every half-century or so. But the least we can do is make sure when those moments do come around, we're not a century behind in the technology used to discern the will of the people.