[Actually the patch I attached goes over a couple of other patches I also had applied; but it still applies to the current source and runtimes OK, just with some offsets.]
Ben Schmidt wrote: >> I'm not sure how to add the variables. > > I am. :-) > >> I've patched 7.1.386 with the new floating point patch. > > 326 perhaps? > >> The attached patch to that includes log, exp, sin, cos, tan, >> asin, acos, atan and atan2. > > OK, here's a counter-patch, including documentation, which is the bigger > task, I think. It includes all my suggestions, as well as Bill's work, > plus a couple of other things that have been suggested/that I noticed: > > - New variables: v:inf, v:nan, v:e, v:pi. > - New functions: abs, acos, asin, atan, atan2, cos, cosh, deg2rad, exp, > log, rad2deg, sin, sinh, tan, tanh. > - Change division by zero behaviour for Numbers to match what float2nr() > does. > - Fixes get_float_arg so it's inside #ifdef FEAT_FLOAT and you can still > compile without float support! > - Documentation for new variables, new functions, float and altered > Number division by zero behaviour, float precision, another pow() > example, and more conspicuous noting of the use of str2float as > nr2float. Fix in alphabetical order in a list, too. I noticed also > that only using %e can you be sure of getting maximum precision. %g > you can in some situations get more or less shown than you want. > >> Caution - it's DOS format and is applied from the src >> directory. > > This one is in Unix format and applies from the root of the source tree > with -p1. It goes on top of 7.1.326 and Bram's floating point patch. > > So at least this won't take any signficant amount of Bram's time, now. We > just > have to convince him about what are the best parts of it to include! > > I suggest: > - We could easily do without: cosh, sinh, tanh (they are simple to > implement with scripts and rarely used). > - We could probably do without: v:inf, v:nan, v:e, v:pi, exp, log, > tan, deg2rad, rad2deg (simple to implement with scripts, but they > are convenient and common so worth keeping, particularly the last > two, which make sin, cos and atan much more useful). > - We could probably do without: asin, acos, atan2 (but implementing > them with script, although possible, is quite a lot of pain, so I > would suggest keeping them, even though they are more rare). > - The rest of the functional changes have either already been agreed > to or I personally deem to be pretty much essential. > > By the way, Bram, I am getting warnings when compiling on the Mac about > MAXPATHL being redefined. It is defined at os_mac.h:222 and > os_unix.h:438, which are evidently both included. I think simply > removing line os_mac.h:222 will correctly fix it. > > Ben. > > > > > > > --~--~---------~--~----~------------~-------~--~----~ You received this message from the "vim_dev" maillist. For more information, visit http://www.vim.org/maillist.php -~----------~----~----~----~------~----~------~--~---
