(part 2)

USPS's serialized Intelligent Mail barcodes can provide election officials near-real-time visibility into uniquely identified ballot mail as they move through postal processing.

That is especially interesting in terms of the TCP analogy. USPS says uniquely serialized Intelligent Mail barcodes can facilitate tracking individual ballots to and from individual voters. Unlike TCP, however, the postal tracking system itself does not automatically replace a ballot when something goes wrong. What it does provide is much greater visibility into the transportation chain once the ballot enters the mailstream for USPS and election officials. In many states, the voter can also see tracking information. If this type of individualized tracking continues to expand, I would expect that eventually voters in every state will be able to see that kind of information.

The important distinction is that the voter requested absentee system adds information before USPS tracking even begins. USPS can track either type of properly barcoded ballot. The advantage created by the absentee request is the additional voter initiated and verified step at the beginning of the chain.

Put another way, USPS tracking can tell us more about what happened after a ballot entered the mail. The absentee request can also tell us why that particular ballot was sent in the first place. We know the voter requested it, the request was verified, and the ballot was issued in response to that request. The voter is expecting that ballot, and the election office can potentially be expecting a ballot back from that particular voter.

That does not mean the election office knows the voter will return it. The voter may decide not to vote or may use another voting method. But the requested ballot gives the system a documented, voter-initiated starting point before the ballot even enters the mail system.

With a physical ballot, determining that something is actually missing or lost is more complicated.

Improved postal tracking can provide considerably more visibility into a ballot once it enters and moves through the postal system. When that information is combined with the original voter request, verification of that request, ballot issuance records, individualized postal tracking, return tracking, verification and confirmation of receipt, it creates a more complete record of that ballot's movement through the entire process.

That is why I used the computer file analogy in the first place. Moving information electronically may involve an enormous number of individual packets, yet TCP can monitor the transmission, detect when expected data has not been acknowledged and retransmit that data. Absentee ballots cannot do that automatically, but a voter-requested absentee system can create a much more traceable chain of events around each individual ballot through the original request, verification, issuance records, individualized postal tracking, return tracking and confirmation of receipt.

Obviously, the analogy isn't exact. But it provides a way to visualize the difference between simply transporting something and building a transportation system that provides better real-time feedback when something has gone wrong.

This may sound like overkill, but remember that we're talking about creating a standardized reliable voting system that meets everyone's needs today and can continue doing so for many elections to come.

Read #45 first.

Referring back to my post #45, there is one part of the analogy I left out that I think makes the whole thing much easier to visualize.

I was comparing the physical movement of paper ballots from one place to another with moving a file stored on one computer to another computer.

Think of a computer file as a very long train made up of a great many cars. When that file is transmitted across a network, the "train" doesn't simply travel down one track from beginning to end. The data is broken into many smaller pieces called packets. Those packets can travel independently through physical network equipment and routers, sometimes taking different routes, until they reach the destination computer and are put back into the correct order so the original file can be reassembled.

Now consider the difference between TCP and UDP.

With TCP, there is a great deal of control automatically built into the transportation process. The receiving computer keeps track of what arrived and in what order. If part of the data doesn't arrive, that missing data will be retransmitted. TCP provides sequencing, acknowledgments and integrity checking so the computers can determine whether the transmission was successfully completed.

If the destination computer receives thousands of pieces but determines one piece is missing, it might say, "I received everything except packet 327. Send packet 327 again."

UDP works differently. The packets are sent, but UDP itself does not provide TCP style acknowledgments, ordering or retransmission. If packet 327 never arrives, UDP will not say, "Packet 327 is missing. Send it again." The acknowledgments, sequencing and retransmission provided by TCP are not provided by UDP itself.

That is the part of the computer analogy that I think helps visualize the transportation problem with physical ballots. The comparison isn't that requested absentee ballots are exactly like TCP while universal mail-in ballots are exactly like UDP. Both types of mailed ballots can use verification and tracking. The useful comparison is how much information, tracking and feedback can be built around each individual ballot as it moves through the system.

With a requested absentee type of ballot, there is an additional recorded step in the process. A particular voter requests a ballot, the request is verified according to that state's procedures, and the election office issues a ballot to that voter.

Other systems send ballots automatically to active registered voters without that individual request step. Keep in mind that voter rolls are not necessarily accurate and not all people listed may be technically eligible to vote. It's up to each state to address that. We'll save that problem for another time.

That request step can provide several important pieces of information before the ballot even enters the postal system.

1. The voter initiated the process by asking for a ballot.

2. Election officials have a record that this particular voter requested one.

3. The request is verified before the ballot is issued according to the procedures of that state.

4. The election office has a record showing that a ballot was issued in response to that specific request.

5. Most importantly for this analogy, the request creates the potential expectation of a return ballot before the outgoing ballot is even mailed. The voter asked for a ballot and the election office sent one. There is no guarantee that the voter will actually use or return it.

6. That gives us a clearer chain of recorded events from the very beginning. The voter requested the ballot, the request was verified, the ballot was issued, the ballot entered the mail system, the ballot may be returned, and the election office can record its receipt.

(end part 1)

please don't close this thread.

I'm working on a follow up to #45