Quote:
Originally Posted by Stellaferox
I have to eyeball the data, hence the CSV format.
|
You can't possibly eyeball all 133Mitems. Why not just generate the data into an array then put some testing code in to write out selected chunks of it to MsgBoxes or to small text files, they don't have to be CSV files to be readable.
Quote:
|
Because computing from code takes a lot longer then reading a table (using binary search) I am able to have any entry I want in a 133 million base within 27 steps maximum
|
Sorry - you missed my point. I meant generate all the items into the array rather than read all those items from a file and then do your lookups and build your smaller tables. Although for 27 steps involving 27 equality checks to save any time it must be a relatively long algorithm to generate the data? Also that implies sorting the table. Unless you can generate the data in sorted form that could take a long time as well. Have you done any timings to estimate how long this mammoth thing will take to run?
Quote:
|
lot easier in the earlier days with assembly languages, but Basic4PPC doesn support this (yet)
|
It probably never will as B4PPC outputs C# code which is compiled to IL (intermediate language) and written to the exe which the .NET Framework CLR (Common Language Runtime) JITs (Just In Time compiles) it at runtime to "safe" machine code. All this is done to stop humans fiddling with Assembly code and making mistakes. There are ways to cross over into "unsafe" code in DLLs but for B4PPC that would need a library and I'll go out on a limb and bet that Erel will never support this natively in B4PPC. If you are enough of an anorak (like me) you can buy books to find out how IL and the CLR works.