Hi Doug, thanks for commenting, also showing that you are not opposed to the general idea.
Doug McIlroy wrote on Sat, Jun 24, 2017 at 07:22:11PM -0400: > Please change "gradient" to a generic name. > It is usually read "del" Using that designation would be unfortunate because "del" is also commonly used for the partial derivative symbol, round d, \(pd. > or "nabla" I'm certainly fine with "nabla"! > and it is often symbolized "grad" or "div" I wouldn't like "div"; in tensor analysis, "div" == \(gr\(pc is not the same as "grad" == \(gr; the latter is a vector operator taking a scalar argument (or increasing the order of the tensor it operates on when using the tensor product), while the former is a linear form (or decreasing the order of the tensor it operates on). > But "gradient", by picking out one meaning of several > (gradient, divergence, laplacian) I have never seen the "nabla" symbol used for "laplacian". The Laplacian is the scalar product of the gradient with itself, or equivalently, the divergence of the gradient. The Laplacian is a scalar, while both "grad" and "div" are first order tensors. So, for the Laplacian, we have \(*D = \(gr \(pc \(gr but the "nabla" symbol itself is never called "laplacian" as far as i can tell. > can obscure, rather than reveal, the meaning. Indeed, arguably, pronouncing grad s = \(gr s div v = \(gr \(pc v curl v = \(gr \(mu v lapl s = \(gr \(pc \(gr s as grad s = nabla s div v = nabla dot v curl v = nabla cross v lapl s = nabla dot nabla s can be considered more natural and less likely to cause misunderstandings than grad s = gradient s div v = gradient dot v curl v = gradient cross v lapl s = gradient dot gradient s though i did certainly hear both in spoken language. Anyway, it would be nice to get the patch in! Individual characters can always be reviewed and tweaked on a case-by-case basis once the general policy is established. Yours, Ingo
