tag:blogger.com,1999:blog-5894839.post3340316154144989424..comments2023-10-03T18:59:20.162+02:00Comments on Kazimir Majorinc's Blog (currently not in use): Don't Fear Dynamic Scope (2.)Kazimir Majorinchttp://www.blogger.com/profile/03407339997157446200noreply@blogger.comBlogger2125tag:blogger.com,1999:blog-5894839.post-27009785434275101582013-09-04T18:51:21.585+02:002013-09-04T18:51:21.585+02:00But, if you don't use free variables, you cann...But, if you don't use free variables, you cannot say that you get the "benefits" of dynamic scope, because you aren't using dynamic scope at all. If you don't use free variables, you don't use dynamic scope, with independence from the "implementation" of your language.Anonymoushttps://www.blogger.com/profile/00378641453100687666noreply@blogger.comtag:blogger.com,1999:blog-5894839.post-29203937436491207522008-12-14T19:30:00.000+01:002008-12-14T19:30:00.000+01:00Although it is very interesting to explore the nat...Although it is very interesting to explore the nature of dynamic scoping in newLISP, we must not forget the following points:<BR/><BR/>(1) Many of the disadvantages associated with dynamic scoping can be avoided by avoiding free variables. If globals must be used, they should be distinguishable by some sort of naming convention.<BR/><BR/>(2) newLISP has a context namespace mechanism to totally shield against the disadvantages of dynamic scoping. The namespace overhead memory- and speed- wise is minimal. This way we can reap the benefits of both, dynamic and lexical scoping without the disadvantages of either.<BR/><BR/>I find Kazimir's experiments enlightening and appreciate all of his posts very much. But I am also afraid that they feed the prejudice of less informed readers, who are unable to look at newLISP from any other than a Common LIsp or Scheme perspective.don Luciohttps://www.blogger.com/profile/08731006426725414929noreply@blogger.com