J3/01-298r1 Date: 21 Aug 2001 To: J3 From: Richard Maine Subject: Remove bindname Wg5 resolution L4 in N1443 endorsed N1462, which, among other things, suggested removing BINDNAME. Although there is some use for the feature, it is not adequately explained in the draft and would require work to fix. The feature is inherently nonportable anyway and can be done as a processor extension. J3 voted unanimously on the specs of J3/01-298 (to remove bindname). The edits to do it are: {Without bindname, the always consists of exactly one . We can simplify by collapse out the bnf term for . Other edits below easly fix the few uses of that term. The result ends up easier to read.} [70:20] "" -> "NAME = " [70:21-25] Delete R514, C542, C543 [70:26] "R514" -> "R513" {Delete references to bindname} [246:25-26] Delete C1216. [258:26-35] Delete C1240 and note 12.35. [264:15-31] Delete para and note 12.42. [363:35-37] Delete para {Fix references to } [65:23],[79:5] "" -> "NAME= specifier" [258:22] "" -> "NAME= specifier" [70:26] Delete "in a " (The R513 ref covers this). {Improve exposition while fixing references. Combine C1217 and C1218. It was particularly strange that C1217 talked about "any procedure entity" in a context where C1218 says there can be only one.} [246:27-30] replace C1217-C1218 with "C1217 (R1211) If a with a NAME= is specified, then shall contain exactly one , which shall neither have the POINTER attribute nor be a dummy procedure."