Improvement during review period

[Ken Helberg]

Good day!

I wanted to have a better understanding on how one of the rules applies to someone else:

"If an author breaks his own record within 30 days, the older submission is regarded as withdrawn."

Let's say someone (Author) submits a solution that meets the improvement requirements. This solution is now open to public comment/scrutiny for the next 30 days. Now let's say someone else (Not-The-Author) wants to make a submission within this 30 day period:

1) Are additional submissions that are not from Author denied until after the 30 day comment period? Or,

2) Does a submission from Not-The-Author during the 30 day period need to beat Author's submission based on the current record that existed prior to Author's submission? Or,

3) Does a submission from Not-The-Author during the 30 day period need to meet the improvement requirements based on Author's solution?

I had written out a number of example situations that could occur that would provide weight for choosing any of the above options, but it mostly boils down to:

a) Would Not-The-Author's submission that is determined to be derivative of Author's work be accepted?

b) Would Not-The-Author's submission that is determined to not be derivative of Author's work, that is also determined to affect an area of compression that is exclusive of Author's submission, be accepted?

c) Would Not-The-Author's submission that is determined to not be derivative of Author's work, that is also determined to not affect an area of compression that is not exclusive to Author's submission be accepted?

d) Given any of the above, if it matters if Not-The-Author's submission makes a >1% improvement of the current record (and beats Author's submission) versus a >1% improvement over Author's submission?

There are two things I'm trying to catch/understand with these questions:

1) Understanding whether Author's improvement (and prize money) could be stolen from them by Not-The-Author making a derivative work of Author's submission (and assuming Not-The-Author doesn't play nice).

2) Understanding whether we could lose/miss Not-The-Author's non-derivative, non-exclusive work that might bring about better compression on the whole, but is insufficient to win any of the prize money relative to Author's current submission.

Example:

* Author has a great new compression concept that beats the current record by >1%. It's coded poorly (larger executables), but it meets the time and RAM requirements.

* Not-The-Author reviews Author's submission and determines a smaller/cleaner code implementation of Author's idea that results in smaller executables and still meets the time and RAM requirements.

* Can Not-The-Author submit it during the 30 day period?

** Are there further requirements it must meet?

** Is it considered exclusive of Author's submission despite it being derivative?

Thanks,

Ken

[Matt Mahoney]

The rules state that if an author submits a revised version within 30 days of submission, then the original submission is invalidated. My interpretation is that the 30 day clock is reset (because we have to retest). Also my interpretation is that if you submit between a competitor's two submissions, then you only have to meet the 1% threshold against the first submission. If you pass, then the other revision has to improve 1% over yours. Otherwise they could lock you out with endless small changes.

--
You received this message because you are subscribed to the Google Groups "Hutter Prize" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hutter-prize...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/hutter-prize/0ba448f5-02bd-4a64-8515-c650b871df1bn%40googlegroups.com.