Tony Mechelynck wrote: > On 10/06/08 15:09, Charles E Campbell Jr wrote: > [...] > >> Whereas, requiring a vim to include the interfaces to >> python/ruby/pick-your-poison ... well, now you're talking real bloat (as >> in increasing the vim executable size). >> >> Regards, >> Chip Campbell >> > > Well, like you I think, I believe that vim-script is usually quite > enough for most of the things one would want to do with Vim; but I have > seen people who apparently cannot think of programming Vim in other than > Perl or Python (etc.). As I think you know, for people who think of such > interfaces as "useless bloat", they can be excluded bodily at > compile-time, and compiling Vim is not really difficult. The only bloat > then would be in sources which are left out of the executable, either > because the modules in question are simply not compiled, or because, in > included modules, the instructions in question are in the false branch > of some #ifdef; and I believe that such "source-only bloat" can be > tolerated. > I agree that vim-script is usually "quite enough...", and I understand that there are those who prefer Perl/Python/etc. I probably misstated my point, which I intended to be that worrying about the few extra lines of source to support sin()/cos()/etc in vim is misplaced -- maintenance would be nearly non-existent, the interface code would be tiny, runtime image size would likely be affected only if the functions were invoked, and that the counter-proposal to "just use python" involves considerably more bytes in the runtime image -- without invoking python. Plus a lot of trouble for anyone who doesn't happen to have python compiled for their system or perhaps doesn't know how to compile python support into vim (or whichever external program is proposed: perl, ruby, ...). I myself happen to have both python and perl on my linux boxes plus generally have perl support compiled into my vim executable, so this isn't a difficulty I'd have myself.
Now, the contention that a text editor just shouldn't have sin()/cos()/floating-point support; that's a different matter. Personally, I think it depends on what an individual wants to do. For example, consider someone who wishes to present coordinates in both rectangular and polar forms. Perhaps someone would like to do some fancy textwork and have it justified inside some mathematically defined shape (cirles/ellipses) -- maybe even provide a plugin to do such. I know that in the case of circles/ellipses one could likely do something with the Bresenham integer-only algorithm (I've provided a circle/ellipse drawing tool with DrawIt), but that requires knowledge about that type of algorithm. Perhaps one would like to provide a small spreadsheet capability in vim, to crosscheck or spot-check output to a file... Bram has already mentioned ease of doing something with columns of numbers. However, I'd agree that there shouldn't be any effort devoted to supporting "special" math functions in vim (ie. Bessel functions, parabolic cylinder functions, elliptical functions, ...) -- just the small interfaces to those functions that come with most compilers (trig, log, exponential). And, as an off-subject item: I'd really like to have Vince Negri's conceal/ownsyntax patch incorporated! Working with LaTeX would be so much better... --~--~---------~--~----~------------~-------~--~----~ You received this message from the "vim_dev" maillist. For more information, visit http://www.vim.org/maillist.php -~----------~----~----~----~------~----~------~--~---