I read about the Loebner Prize Protocol and I have some questions about the monotonic sequence:

* Must the sequence start at 0000 or 0001? Or is any start allowed as long as each character increases the sequence and the number of digits is fixed?

* Assuming the bot follows the sequence, is there any (reasonable) limit to the number of digits in the sequence. Eg must “HI” be 0001.H 0002.I or is 10000001.H 10000002.I equally OK? (I assume whatever the sequence is, it must have the same number of digits)

* As long as the bot follows the increasing sequence, can it jump upwards, so long as the sequence always has the same fixed number of digits. Eg could the characters in the bots first reply be 00010001.letter to 00010010.letter (assuming 10 characters in the first reply), be 00020001.letter to 00020020.letter (assuming 20 characters in the bot’s second reply) etc. Or can there never be a gap?

You have an example here : Loebner protocol

0000000123.H.other
0000000235.e.other
0000000456.l.other
0000000789.l.other
0000000888.o.other
0000001234.comma.other
0000002222.space.other
0000002345.J.other
0000004567.i.other
0000006789.m.other
0000007777.period.other
0000008123.Return.other

So the sequence can begin at any number and have gaps (123, 235, ...)

The time between keystrokes could be significant. I would guess that the example that Denis posted is numbered with the time in milliseconds since the start of the conversation.

Many thanks Denis. I didn’t know about that document.

Does the sequence always require 10 digits? Or can any number be used (as long as the number does not vary)?

The number of digits that your program uses to send LPP data can be any amount of digits, as long as you stick to that amount consistently, and as long as they can accommodate the length of the conversation(s).
I just incremented the number +1 for each letter my program sends, and use a repeating findfirst() method to catch whatever the LPP program sends. Worked fine last year.
It seems the LPP program reads them in similar way to how windows explorer lists files alphanumerically.

Best would be to download the file Denis linked, change its extention to “.pl” and try it out, then you’ll see.
I wish the AISB would put this stuff online though, newcomers would be quite puzzled.

I think the original idea was to use Unix time so that the number would not only be in sequence, but would also give a timestamp that a player program could then use to “play back” the conversation in it’s original timing.

I copied this from Wikipedia: “Unix time (aka POSIX time or Epoch time), is a system for describing instants in time, defined as the number of seconds that have elapsed since 00:00:00 Coordinated Universal Time (UTC), Thursday, 1 January 1970,[1][note 1] not counting leap seconds.[1][2][note 2] It is used widely in Unix-like and many other operating systems and file formats. Due to its handling of leap seconds, it is neither a linear representation of time nor a true representation of UTC.[note 3] Unix time may be checked on most Unix systems by typing date +%s on the command line.”

I think Andrew Smith’s time in milliseconds is probably the most useful. If you use “Date +%s” it is possible to get two identical numbers unless they are *seconds* apart.

I think Robby has it spot on. There used to be a Loebner Player in which you could load these folders and see the contest unfold on your computer in “real time”. Not sure if that is still around though.