He did not copy numbers, which were _way_different_, he did not copy code (there was not any, just the numbers). There was just the hypothesis that his numbers were compatible with the same loop mechanics, using the same ramps -4, -2, -1, 0 as template or whatever that was. And ths was for 6 or 7 out of 11 tables (the others did not fit or were in a couple of cases just zeros). Stockfish (glaurung actually) may have copy fruit numbers much more likely than rybka. But this was used all of over the place as an incredibly incriminating evidence. Now... mehh. You admit there was no smoking code gun. Let's all shrug our shoulders. The fact is that at the time of the investigation nobody payed enough close attention (you even missed a match between a fruit table and Crafty) and things were taken at face value.bob wrote:Rebel wrote:Note what you said in the past -bob wrote:I do not believe so. IIRC, copyright infringement specifically excludes tables of numbers.hgm wrote:But that does not really answer the question. Which was:bob wrote:Your PST point addresses an issue I have raised many times. One PST is just "interesting". MANY PSTs is damning. The latter shows that little original research was done to choose the values.
If some public copyrighted code XXX reads
and in my own program I would writeCode: Select all
int KingPST[5][5]; init_pst() { int i, j; for(i=1; i<5; i++) for(j=1; j<5; j++) KingPST[i][j] = (i-2)*(i-2) + (j-2)*(j-2); }
(plus similar stuff for all other piece types, to make sure it could be damningI), would it be a copyright infringement? Would there be striking similarity, substantial similarity or no similarity at all in the sense of copyright law between these two code fragments?Code: Select all
int KingPST[25] = { // use same values as XXX 8, 5, 4, 5, 8, 5, 2, 1, 2, 5, 4, 1, 0, 1, 4, 5, 2, 1, 2, 5, 8, 5, 4, 5, 8, }
Bob - Sorry, but YOU are the idiot that says one can copy a table of numbers without violating copyright.
What changed your mind ?
Simple. I went back, when this discussion started, and carefully re-read existing copyright law, and case law examples. Copyright law specifically excludes a group of numbers, things like lists of primes, lists of Mersenne primes, trig tables, etc. Things that can be computed by any sort of formula or algorithm. Our believe was that vas did not just copy a set of tables, he copied the code that produced the tables and used that to produce them. Which would be a copyright violation. But without the actual "smoking gun" of the actual purloined code, we simply used it as one of many pieces of data showing that rule 2 had clearly been violated...
Oh, well... I am not going to go through this all over again...
Miguel
PS: Let's not even mention that this was reported for R1 (because it was Zach's analysis), but for R2 (which is the one that actually played in ICGA) the PSTs changed dramatically. That was not properly acknowledged in the report. It was only hand-waved.