I warned you at the start: There are many questions. I have moved on, but in a spiral fashion The question will be re-visited but for now Ive got enough working possibilities to get on with. It seems I was not the only person in need of such a routine. As it very well can be done as a project in its own right, if anyone is at a loose end, please fell free to have a go! Ill be happy to put it to good usemartyn_hill wrote:Whilst Per has moved on to to fry bigger fish
Of course Ill have a think on it. But the (temporary) conclusion Ive come to is as I stated earlier: a distributed statistical sample (I dont have the language, but I know what I mean ) seems to be the most efficient approach. For now Ill settle for the brute force "methodology", as I reckon that itll be a one-off cost in most instances, applied to mainly older, unchanging files. A user making new screen dumps would then hopefully learn to use _scr4 and _scr8 extension names in the future..I thought I'd close the Mode4/Mode8 'sample-point' conversation...
Should anyone find it of any value, the address offsets (integer) for a set of 127x test-points forming a spiral, based on the 32k QL display can be found attached as a .txt file and saved as spiral_offsets.zip
The pattern itself looks like:
spiral_testpoints.png
Confucius say: "By three methods we may learn wisdom: First, by reflection, which is noblest; Second, by imitation, which is easiest; and third by experience, which is the bitterest." I trust that these forces will be at work here too..