Austin Group Minutes of the 21 August Teleconference Austin-185 Page 1 of 1 Submitted by Andrew Josey, The Open Group. August 22, 2003 Attendees Andrew Josey, The Open Group Don Cragun , Sun, PASC OR Joanna Farley, Sun Ulrich Drepper, Red Hat Nick Stoughton, USENIX, WG15 OR Mark Brown, IBM, TOG OR Keld Simonsen, DKUUG, WG15 OR Alternate Apologies Dave Butenhof, HP Andrew reported that he has requested ballot group to be formed for TC2 by IEEE. After the call he received confirmation from IEEE about that and we expect the invite to go out soon. Andrew reported that the ISO 2003 edition was published on the 18 August 2003 (official pub date was 15 August 2003) Latest status of the reports is as follows Volume Total #open XBD 38 10 XSH 167 27 XCU 36 9 XRAT 2 0 TOTAL 243 46 Future rdvk will not now make TC2 and be diverted to a new inbound pool. XSH --- XSH ERN 146 A/M This was reopened with further information after the last call. See the rdvk for details XSH ERN 97 Accept (note the editors may do some editorial fixup) XSH ERN 101 OPEN The feeling was that the standard already requires this, we just need to work on the wording to make this clearer. Leave open for one more week. XSH ERN 113 A/M We decided independent of the Base WG that this request is outside of the scope of what can be done in a TC and should therefore be left on file for the next revision. The A/M will be that the request is being held on file. Action : Andrew store this request in a document in the docreg XSH ERN 117 OPEN Fred T has filed a DR against ISO C, so this will remain open pending response XSH ERN 138 OPEN Discussion on the call that we need to add a new term implementation-specific -- Andrew subsequently thinks that this should be "implementation-specified" . Andrew has action to work on this and circulate proposal. ERN 139 Accept XSH ERN 142 Accept -> OPEN Andrew marked this OPEN since further information needs to be reviewed next time. (see below also added to rdvk) (gwc) I agree with Don that the apparent requirement for freopen() to close and reopen the underlying fd when called with a null pathname looks to be an oversight when the new C99 requirement was folded into the text. However, there is a second issue here which Don has not raised, concerning the use of "as if": The standard says freopen() with a null pathname is supposed to change the stream "as if the name of the file currently associated with the stream had been used". The question is whether this means the access mode of the underlying fd must also change as if the file name had been used. E.g. suppose fp is a stream open read-write (thus the underlying fd has access mode O_RDWR), and a call to freopen(NULL, "w", fp) is made. Should the underlying fd now have an access mode of O_WRONLY, as it would if freopen("thefilename", "w", fp) had been used? ERN 147 Accept ERN 148 Accept ERN 149 Accept ERN 150 Accept ERN 151 Accept ERN 152 Accept ERN 153 OPEN There was much discussion about removing ENOTSUP, a feeling on some parts that removing it could require a gratuitous change for implementations (for next time) XBD ERN 27 OPEN Andrew has action to put words together for pathconf(). We did not get time to check this homework item, its included in the rdvk Next Steps ----------- Andrew will update the aardvark reports on the web site with the latest information . The next call will be August 28