How much memory may a HPPL program use on a G2?

[jte]

(09-21-2023 04:12 AM)komame Wrote:  ⋮ (many many good points)
Such a dictionary should already be stored in a 6-bit binary version before the program is executed, but I'm afraid the time required for decoding the result would completely exclude this solution in terms of performance. And yet we haven't even mentioned 21-character words

Thanks for giving some actual timing measurements.

Ideally, in case it isn’t obvious, lists would be shortened as soon as possible for decoding. As I’ve mentioned, it may well be that some changes to PPL (adding a few routines, for example), might substantially improve the efficiency of this type of coding. Actual real-world measurements, like you’ve helpfully done, can help convert idle thought experiments / conjectures into practical code. One possible already-present routine that could be used for trimming is SORT. This does more than we want (we’re wanting to separate the zeroes from the non-zeroes; the “not a match” from the “matches”). (To determine the boundary after sorting, a ΣLIST of a list of zeroes / ones could be used, or a binary search on the sorted list.)

At a more abstract level, one can ask what the goal of the program is. If it is to determine the number of matches, decoding isn’t necessary. If it is to show matches, only a screenful need be shown at once. (This is just a general comment.)