Path: csiph.com!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail From: rbowman Newsgroups: comp.os.linux.misc,alt.folklore.computers Subject: Re: The joy of FORTRAN Date: 24 Sep 2024 23:31:52 GMT Lines: 9 Message-ID: References: <5mqdnZuGq4lgwm_7nZ2dnZfqnPSdnZ2d@earthlink.com> <1r0e6u9.1tubjrt1kapeluN%snipeco.2@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-Trace: individual.net hpPOnB5uugMrYqdA8q75fQ5WVpLh2CXVhdEmtg8RM99D30V8Wn Cancel-Lock: sha1:jEFXi33CJD3obLubRbw43n80ae4= sha256:hAzuABRHg+GH4Nl8AO+vSRYLBeWuK9+kW25o9277aXM= User-Agent: Pan/0.149 (Bellevue; 4c157ba) Xref: csiph.com comp.os.linux.misc:58408 alt.folklore.computers:226966 On Tue, 24 Sep 2024 20:01:31 +0200, R Daneel Olivaw wrote: > ALGOL60 was the language where a test of equality between two floating > point numbers was actually a test of "close enough for ALGOL". If I > want to test for "approximately equal" then I want a different operator. Agreed, but that would be a handy operator. I do some GIS programming where equality close enough for horseshoes or hand grenades is what you really want so you wind up testing with some arbitrary epsilon value.