Hi Glen, Yes I've already raised the (potential) issue of the real-time computation cost of the length tuning rules - it's essentially an O(MxN) algorithm (M = number of rules, N = number of PCB items), which does worry me from a scaling point of view as people hopefully develop more complex boards as the high speed tools develop. As you say there are various mitigations to this: worker threads, intelligent caching, pre-computation, etc etc. All stuff to consider.
Yours, James. On Fri, 1 Mar 2024 at 21:58, Glen English <[email protected]> wrote: > Hi James. Yeah I agree text based rule core is a fine way of representing > it. > What we will find is that as the rule base grows, the online DRC > functions will start chewing considerable processor resources- it will > likely need to be spun out into another low priority thread. > Maybe the plain english rule base language will need to be tokenized in > the background to reduce the text parsing load of rule threads. (maybe it > already is I dont know) > There is a desire by many users to bring more of the rule base into real > time "online" space, IE many things I beleive should be examined as the > user places tracks, via etc - so that it does not need to be cleaned up > later (big productivity issue ) and so how this is done and how > efficienctly under the hood will begin to matter. > -glen > > On Saturday, March 2, 2024 at 8: > > -- You received this message because you are subscribed to the Google Groups "KiCad Developers" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/a/kicad.org/d/msgid/devlist/CAMVX%3DtY3rej0%3DvOhc82vemNpnQun36DRWHYHXOp6w%2B6iYN1jzw%40mail.gmail.com.
