![]() ![]() gets back all the names of cells with data, then retrieves them one by one in parallel with the search. So while a merge-sort is a cool trick given that you're working with a tree, I doubt it's actually going to be the most efficient solution vs. GitHub - tuxitop/exapunks-solutions: My solutions to the levels and battles of the EXAPUNKS game by Zachtronics. In other words, search / sort isn't the most time-consuming thing here, and you can do it in parallel with actually getting back the stuff you need. My solutions to the levels and battles of the EXAPUNKS game by Zachtronics. There are two standard registers: X is the universal general purpose register. Also, HALT is not needed at the end of a program. ![]() LINK 800 GRAB 200 SEEK 9999 COPY M F COPY M F COPY M F COPY M F COPY M F. GRAB 300 COPY F M COPY F M COPY F M COPY F M COPY F M. 1) While the histograms displayed at the end of a solution run correctly display the 3 metrics you achieved with this solution, those displayed in the missions list screen display your best achievement for each metric INDIVIDUALLY. These must be loaded from files, host name or hardware registers. You could replicate too pretty easily since it is so short but it would end up longer. Hey, Here are a few thoughts about how the score is currently calculated. My first solution was to do all the math in the main loop, but I ran into an issue, which is that counting UP to a loop target, means you cant hold another value except in a file. This one stumped me for a while, and the solutions I came up with I was only half happy with. EXA programs cannot contain string or keyword constants. Heart Level Solution - Not happy about it. R/N means the name of a register, or a number. I feel like you would be better off focusing more on parallelizing "find the next-lowest item to add to the output" and "retrieve it." The reality is that when you account for dealing with race conditions and the like, actually retrieving the next output is going to take longer than determining which one you want - so you can have one EXA determining it, and another one fetching it.Īt that point it doesn't matter that the determine-next-item EXA is searching slowly, because it's parallelized with a task that's taking even longer. EXA instruction parameters are as follows: R means the name of a register. On top of that, you have to yield to your architecture to an extent - inserts are extremely hard to do efficiently in Exapunks, which substantially increases the time needed for a merge sort. Kamikaze Cube 2 VR / TSUN-TSUN VR EXAPUNKS Need to Know Lucah: Born of a. ![]() Like azure said, I don't think the advantage in complexity helps - the lists are so small that the constant increases necessary to set it up will kill your advantage. Earth Muncher ZeroRanger Super Inefficient Golf Tyr: Chains of Valhalla. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |