gisle@ifi.uio.no (Gisle Hannemyr) (02/01/90)
In article <12030@smoke.BRL.MIL> gwyn@brl.arpa (Doug Gwyn) writes: >In article <2518@ifi.uio.no> gisle@ifi.uio.no (Gisle Hannemyr) writes: >>Does anyone know what the relationship is between ISO DIS 9899 and >>ANS X3.159-1989? > >They're supposed to be technically identical, although the British ISO >WG14 representatives are expected to provide a supplement concerning >instances of "undefined behavior" that may become part of the IS. In the january 26, 1990 edition of IDG Communications "ComputerWorld" (Norwegian edition), there is an article: "C -- ISO or ANSI" on page 61. Among other things, it concludes (my translation): "It seems that the efforts to create an unified international standard for C is diverging in two main factions. One faction will let ISO develop the final C standard, based on the draft ANSI standard, the second faction want to publish the ANSI draft standard, and continue work on a new ANSI C standard to replace the present. The new ANSI C standard is expected to be purged of several strange features that several commentators has opinied are present in the draft ANSI standard The British has commented that they believe that ISO will create a more complete specification of C than the present ANSI draft" Is this totally bogus? - gisle hannemyr (Norwegian Computing Center) EAN: gisle@nr.uninett Inet: gisle@ifi.uio.no UUCP: ...!mcvax!ifi!gisle ------------------------------------------------
gwyn@smoke.BRL.MIL (Doug Gwyn) (02/02/90)
In article <2551@ifi.uio.no> gisle@ifi.uio.no (Gisle Hannemyr) writes: > "It seems that the efforts to create an unified international > standard for C is diverging in two main factions. One faction > will let ISO develop the final C standard, based on the draft > ANSI standard, the second faction want to publish the ANSI > draft standard, and continue work on a new ANSI C standard > to replace the present. The new ANSI C standard is expected > to be purged of several strange features that several > commentators has opinied are present in the draft ANSI > standard > The British has commented that they believe that ISO will > create a more complete specification of C than the present > ANSI draft" >Is this totally bogus? Pretty much so. There is no intention among X3J11 members (developers of the ANSI C standard) to begin work on a replacement standard for many years to come. (The standard does need to be reviewed every 10 years.) Some X3J11 members are working toward a set of extensions targeted at numerical programming applications. My feeling is that the only proper way to do that is through extensions of, not conflicts with, the ANSI C standard. (For example, #include <nmath.h> could make errno-free replacement math functions available for those who think they prefer them.) ISO WG14 and X3J11 have agreed to work toward a joint "information bulletin" that would augment the C standard with additional coverage of the issue of "undefined behavior", which the British representatives feel is needed. This would NOT affect the official ANSI C standard, although so long as it is compatible with it it would serve as guidelines for implementors and programmers. I don't know whether or not ISO would formally graft such an appendix onto the IS or not. In any event it would not, as I understand it, affect the intention of the ANSI C standard, merely clarify it. Of course, when it comes to politics, all sorts of silly things might happen. However, I have heard nothing about "purging features from the ANSI C standard". The only one that might have been in question was trigraphs, and ISO WG14 voted the Danes down on that issue, which should have (according to gentlemen's rules of this business) closed the issue.