From rybin@possum.srcc.msu.su Mon Jan 19 10:05:27 1998 Return-Path: Received: from cs.ida.org by cronus.csed.ida.org (SMI-8.6/SMI-SVR4) id KAA18386; Mon, 19 Jan 1998 10:05:26 -0500 Received: from mail.acm.org (mail.acm.org [199.222.69.4]) by cs.ida.org (8.8.7/8.8.7) with ESMTP id KAA24512 for ; Mon, 19 Jan 1998 10:06:42 -0500 (EST) Received: from sw-eng.falls-church.va.us (ns1.sw-eng.falls-church.va.us [199.75.54.2]) by mail.acm.org (8.8.5/8.7.5) with ESMTP id KAA21642; Mon, 19 Jan 1998 10:06:32 -0500 Received: from crocus.gamma.ru by sw-eng.falls-church.va.us (8.8.8/) id PAA06482; Mon, 19 Jan 1998 15:02:03 GMT Received: from srcc.UUCP (uucp@localhost) by crocus.gamma.ru (8.8.7/8.7.3) with UUCP id SAA28416 for asis-comment@sw-eng.falls-church.va.us; Mon, 19 Jan 1998 18:05:21 +0300 (MSK) Received: by gamma.srcc.msu.su; Mon, 19 Jan 1998 18:04:29 +0300 Received: by possum.srcc.msu.su (UUPC/@ v5.09gamma, 14Mar93); Mon, 19 Jan 1998 18:05:27 +0300 To: asis-comment@sw-eng.falls-church.va.us Message-Id: Organization: Information Systems, SRCC, MSU From: "Sergey I. Rybin" Date: Mon, 19 Jan 98 18:05:27 +0300 X-Mailer: BML [MS/DOS Beauty Mail v.1.36] Subject: 10.18: what is the same physical CU? Lines: 31 Content-Length: 1288 Status: OR 10.18 (function Is_Identical for Compilation Units) says: -- Returns True if Left and Right represent the same physical compilation -- unit, from the same open ASIS Context variable, or, if both are -- Nil_Compilation_Unit values. ("The same physical compilation -- unit" have the same name and the same compilation time stamp.) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ "The same _compilation_ time stamp" does not make sense for the source-based GNAT compilation model. Even though ASIS-for-GNAT is not (completely) source-based any more, "the same physical compilation unit" may be compiled several times to create the persistent data structures used in ASIS-for-GNAT as the basis for ASIS Context implementation. Therefor in the implementation of Asis.Compilation_Units.Is_Identical I've replaced "the same compilation time stamp" with "the same source file time stamp", which exactly reflects the situation with Compilation Units in GNAT environment and in ASIS-for-GNAT. In the same time, I'm not sure, that this replacement would work for some "heavy-library-based" Ada implementation, where every unit is compiled exactly one time. So, I think, that we have a problem with 10.18. Unfortunately, at the moment I have no idea how to fix it :( Sergey From dewar@gnat.com Mon Jan 19 10:24:24 1998 Return-Path: Received: from cs.ida.org by cronus.csed.ida.org (SMI-8.6/SMI-SVR4) id KAA18397; Mon, 19 Jan 1998 10:24:23 -0500 Received: from mail.acm.org (mail.acm.org [199.222.69.4]) by cs.ida.org (8.8.7/8.8.7) with ESMTP id KAA24750 for ; Mon, 19 Jan 1998 10:25:39 -0500 (EST) Received: from sw-eng.falls-church.va.us (ns1.sw-eng.falls-church.va.us [199.75.54.2]) by mail.acm.org (8.8.5/8.7.5) with ESMTP id KAA12824; Mon, 19 Jan 1998 10:25:15 -0500 Received: from nile.gnat.com by sw-eng.falls-church.va.us (8.8.8/) id PAA06928; Mon, 19 Jan 1998 15:21:32 GMT Received: by nile.gnat.com (5.0/1.20) id AA20151; Mon, 19 Jan 98 10:24:47 EST Date: Mon, 19 Jan 98 10:24:47 EST From: dewar@gnat.com (Robert Dewar) Message-Id: <9801191524.AA20151@nile.gnat.com> To: asis-comment@sw-eng.falls-church.va.us, rybin@possum.srcc.msu.su Subject: Re: 10.18: what is the same physical CU? Content-Length: 934 Status: OR <<-- Returns True if Left and Right represent the same physical compilation -- unit, from the same open ASIS Context variable, or, if both are -- Nil_Compilation_Unit values. ("The same physical compilation -- unit" have the same name and the same compilation time stamp.) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ >> Sergey is right, this is nonsense semantically. The proper notion is that the versions of the compilation unit are the same. We already went through getting this write for the version attributes in the RM, ASIS should simply follow the path set out here. The whole notion of time of compilation is a flawed one which cannot be given any sensible meaning with respect to the RM. THis has nothing to do with source based approaches or anything else, it has to do with fundamental semantic levels of abstraction, and the notion of "time of compilation" is completely meaningless semantically. From rybin@possum.srcc.msu.su Mon Jan 19 11:04:31 1998 Return-Path: Received: from cs.ida.org by cronus.csed.ida.org (SMI-8.6/SMI-SVR4) id LAA18466; Mon, 19 Jan 1998 11:04:31 -0500 Received: from mail.acm.org (mail.acm.org [199.222.69.4]) by cs.ida.org (8.8.7/8.8.7) with ESMTP id LAA25195 for ; Mon, 19 Jan 1998 11:05:43 -0500 (EST) Received: from sw-eng.falls-church.va.us (ns1.sw-eng.falls-church.va.us [199.75.54.2]) by mail.acm.org (8.8.5/8.7.5) with ESMTP id LAA39650; Mon, 19 Jan 1998 11:05:37 -0500 Received: from crocus.gamma.ru by sw-eng.falls-church.va.us (8.8.8/) id QAA08070; Mon, 19 Jan 1998 16:01:52 GMT Received: from srcc.UUCP (uucp@localhost) by crocus.gamma.ru (8.8.7/8.7.3) with UUCP id TAA29839; Mon, 19 Jan 1998 19:05:19 +0300 (MSK) Received: by gamma.srcc.msu.su; Mon, 19 Jan 1998 19:03:58 +0300 Received: by possum.srcc.msu.su (UUPC/@ v5.09gamma, 14Mar93); Mon, 19 Jan 1998 19:04:20 +0300 To: asis-comment@sw-eng.falls-church.va.us, dewar@gnat.com References: <9801191524.AA20151@nile.gnat.com> Message-Id: Organization: Information Systems, SRCC, MSU From: "Sergey I. Rybin" Date: Mon, 19 Jan 98 19:04:20 +0300 X-Mailer: BML [MS/DOS Beauty Mail v.1.36] Subject: Re: 10.18: what is the same physical CU? Lines: 18 Content-Length: 814 Status: OR > The proper notion is that the versions of the compilation unit are the > same. We already went through getting this write for the version > attributes in the RM, ASIS should simply follow the path set out > here. The whole notion of time of compilation is a flawed one which > cannot be given any sensible meaning with respect to the RM. Robert, is you suggestion to use the definition from RM 95, E.3(5): The version of a compilation unit changes whenever the version changes for any compilation unit on which it depends semantically. The version also changes whenever the compilation unit itself changes in a semantically significant way. It is implementation defined whether there are other events (such as recompilation) that result in the version of a compilation unit changing. ? From dewar@gnat.com Mon Jan 19 11:06:14 1998 Return-Path: Received: from cs.ida.org by cronus.csed.ida.org (SMI-8.6/SMI-SVR4) id LAA18470; Mon, 19 Jan 1998 11:06:13 -0500 Received: from mail.acm.org (mail.acm.org [199.222.69.4]) by cs.ida.org (8.8.7/8.8.7) with ESMTP id LAA25206 for ; Mon, 19 Jan 1998 11:07:28 -0500 (EST) Received: from sw-eng.falls-church.va.us (ns1.sw-eng.falls-church.va.us [199.75.54.2]) by mail.acm.org (8.8.5/8.7.5) with ESMTP id LAA27550; Mon, 19 Jan 1998 11:07:24 -0500 Received: from nile.gnat.com by sw-eng.falls-church.va.us (8.8.8/) id QAA08101; Mon, 19 Jan 1998 16:03:33 GMT Received: by nile.gnat.com (5.0/1.20) id AA21453; Mon, 19 Jan 98 11:06:47 EST Date: Mon, 19 Jan 98 11:06:47 EST From: dewar@gnat.com (Robert Dewar) Message-Id: <9801191606.AA21453@nile.gnat.com> To: asis-comment@sw-eng.falls-church.va.us, dewar@gnat.com, rybin@possum.srcc.msu.su Subject: Re: 10.18: what is the same physical CU? Content-Length: 336 Status: OR exactly that is the right definition to use, nothing else makes sense. Since the RM has gone to the trouble of defining the notion of Version, and this is really what ASIS is getting at here, we should simply share the notion of version. For example, in GNAT, a trivial change in the spelling of a comment does NOT change the version. From rybin@possum.srcc.msu.su Mon Jan 19 15:21:03 1998 Return-Path: Received: from cs.ida.org by cronus.csed.ida.org (SMI-8.6/SMI-SVR4) id PAA18671; Mon, 19 Jan 1998 15:20:46 -0500 Received: from mail.acm.org (mail.acm.org [199.222.69.4]) by cs.ida.org (8.8.7/8.8.7) with ESMTP id PAA27728 for ; Mon, 19 Jan 1998 15:21:59 -0500 (EST) Received: from sw-eng.falls-church.va.us (ns1.sw-eng.falls-church.va.us [199.75.54.2]) by mail.acm.org (8.8.5/8.7.5) with ESMTP id PAA28674; Mon, 19 Jan 1998 15:21:53 -0500 Received: from crocus.gamma.ru by sw-eng.falls-church.va.us (8.8.8/) id UAA20078; Mon, 19 Jan 1998 20:18:05 GMT Received: from srcc.UUCP (uucp@localhost) by crocus.gamma.ru (8.8.7/8.7.3) with UUCP id XAA05144; Mon, 19 Jan 1998 23:21:23 +0300 (MSK) Received: by gamma.srcc.msu.su; Mon, 19 Jan 1998 23:20:49 +0300 Received: by possum.srcc.msu.su (UUPC/@ v5.09gamma, 14Mar93); Mon, 19 Jan 1998 23:02:56 +0300 To: asis-comment@sw-eng.falls-church.va.us, dewar@gnat.com References: <9801191606.AA21453@nile.gnat.com> Message-Id: Organization: Information Systems, SRCC, MSU From: "Sergey I. Rybin" Date: Mon, 19 Jan 98 23:02:56 +0300 X-Mailer: BML [MS/DOS Beauty Mail v.1.36] Subject: Re: 10.18: what is the same physical CU? Lines: 25 Content-Length: 1241 Status: OR > exactly that is the right definition to use, nothing else makes sense. > Since the RM has gone to the trouble of defining the notion of Version, > and this is really what ASIS is getting at here, we should simply share > the notion of version. > > For example, in GNAT, a trivial change in the spelling of a comment does > NOT change the version. But this example shows, that for Is_Identical and Is_Equial tests for Compilation Units in ASIS, the RM-defined version of a version is not enough! If for two ASIS Compilation Units, U1 and U2, Is_Equial (U1, U2) (or Is_Identical (U1, U2) ), this means, that all the ASIS queries applied to these units should give the same results, does not i? But what about queries from Asis.Text? If they should return the same results for U1 and U2, then U1 and U2 should have the same source text, and any change in any comment is important! I think, we should take into account, that ASIS Compilation Unit is in some sense "superset" of the notion of compilation unit as it is refined in RM. I would suggest to define in ASIS "the same physical Compilation Units" as "units having the same version, as defined in RM, E.3(5) AND the same source text. including line layout and comments" From dewar@gnat.com Mon Jan 19 15:38:48 1998 Return-Path: Received: from cs.ida.org by cronus.csed.ida.org (SMI-8.6/SMI-SVR4) id PAA18730; Mon, 19 Jan 1998 15:38:48 -0500 Received: from mail.acm.org (mail.acm.org [199.222.69.4]) by cs.ida.org (8.8.7/8.8.7) with ESMTP id PAA27895 for ; Mon, 19 Jan 1998 15:40:01 -0500 (EST) Received: from sw-eng.falls-church.va.us (ns1.sw-eng.falls-church.va.us [199.75.54.2]) by mail.acm.org (8.8.5/8.7.5) with ESMTP id PAA45028; Mon, 19 Jan 1998 15:39:57 -0500 Received: from nile.gnat.com by sw-eng.falls-church.va.us (8.8.8/) id UAA20576; Mon, 19 Jan 1998 20:36:17 GMT Received: by nile.gnat.com (5.0/1.20) id AA06142; Mon, 19 Jan 98 15:39:33 EST Date: Mon, 19 Jan 98 15:39:33 EST From: dewar@gnat.com (Robert Dewar) Message-Id: <9801192039.AA06142@nile.gnat.com> To: asis-comment@sw-eng.falls-church.va.us, dewar@gnat.com, rybin@possum.srcc.msu.su Subject: Re: 10.18: what is the same physical CU? Content-Length: 986 Status: OR <> Shrug, the original wording about compilation time stamps certainly does not meet this requirement either. You cannot talk about source text, some implementations have no such context in the sense you are using the term. I would not require identical source text, too much trouble, if you want to strengthen the test, just test time stamps of sources as well as versions. But note that version includes all dependent units, time stamp does not From rybin@possum.srcc.msu.su Mon Jan 19 16:14:43 1998 Return-Path: Received: from cs.ida.org by cronus.csed.ida.org (SMI-8.6/SMI-SVR4) id QAA18785; Mon, 19 Jan 1998 16:14:43 -0500 Received: from mail.acm.org (mail.acm.org [199.222.69.4]) by cs.ida.org (8.8.7/8.8.7) with ESMTP id QAA28292 for ; Mon, 19 Jan 1998 16:15:58 -0500 (EST) Received: from sw-eng.falls-church.va.us (ns1.sw-eng.falls-church.va.us [199.75.54.2]) by mail.acm.org (8.8.5/8.7.5) with ESMTP id QAA44140; Mon, 19 Jan 1998 16:15:37 -0500 Received: from crocus.gamma.ru by sw-eng.falls-church.va.us (8.8.8/) id VAA21456; Mon, 19 Jan 1998 21:11:47 GMT Received: from srcc.UUCP (uucp@localhost) by crocus.gamma.ru (8.8.7/8.7.3) with UUCP id AAA06572; Tue, 20 Jan 1998 00:15:16 +0300 (MSK) Received: by gamma.srcc.msu.su; Tue, 20 Jan 1998 00:14:10 +0300 Received: by possum.srcc.msu.su (UUPC/@ v5.09gamma, 14Mar93); Tue, 20 Jan 1998 00:13:19 +0300 To: asis-comment@sw-eng.falls-church.va.us, dewar@gnat.com References: <9801192039.AA06142@nile.gnat.com> Message-Id: Organization: Information Systems, SRCC, MSU From: "Sergey I. Rybin" Date: Tue, 20 Jan 98 00:13:19 +0300 X-Mailer: BML [MS/DOS Beauty Mail v.1.36] Subject: Re: 10.18: what is the same physical CU? Lines: 28 Content-Length: 1259 Status: OR > Shrug, the original wording about compilation time stamps certainly > does not meet this requirement either. Agree completely, "compilation time stamp" should go away. > You cannot talk about source text, some implementations have no such > context in the sense you are using the term. My intent was to use "source text" as some abstraction, RM also uses the notion of "program text". I agree, that it is really hard (may be, even impossible) to operate with this notion in the ASIS definition. > I would not require identical source text, too much trouble, if you > want to strengthen the test, just test time stamps of sources as > well as versions. Fine for ASIS-for-GNAT. Probably, this even works in general > But note that version includes all dependent units, time stamp does not Well, my idea was to define equiality and identity for ASIS CUs as "to have the same sources (here the source time stamp works) and the same semantic (here the save versions work for all the units upon which these two units depend semantically). In other words, to be equial in ASIS, two unit should have the same sources, and all their supporters (units upon which they depend) should have the same semantic, but they may have different sources) From dewar@gnat.com Mon Jan 19 16:52:08 1998 Return-Path: Received: from cs.ida.org by cronus.csed.ida.org (SMI-8.6/SMI-SVR4) id QAA18849; Mon, 19 Jan 1998 16:52:07 -0500 Received: from mail.acm.org (mail.acm.org [199.222.69.4]) by cs.ida.org (8.8.7/8.8.7) with ESMTP id QAA28691 for ; Mon, 19 Jan 1998 16:53:23 -0500 (EST) Received: from sw-eng.falls-church.va.us (ns1.sw-eng.falls-church.va.us [199.75.54.2]) by mail.acm.org (8.8.5/8.7.5) with ESMTP id QAA40720; Mon, 19 Jan 1998 16:53:09 -0500 Received: from nile.gnat.com by sw-eng.falls-church.va.us (8.8.8/) id VAA22218; Mon, 19 Jan 1998 21:49:27 GMT Received: by nile.gnat.com (5.0/1.20) id AA21493; Mon, 19 Jan 98 16:52:44 EST Date: Mon, 19 Jan 98 16:52:44 EST From: dewar@gnat.com (Robert Dewar) Message-Id: <9801192152.AA21493@nile.gnat.com> To: asis-comment@sw-eng.falls-church.va.us, dewar@gnat.com, rybin@possum.srcc.msu.su Subject: Re: 10.18: what is the same physical CU? Content-Length: 141 Status: OR <> no, it does not, source time stamps have no meaning in many environments. From sblake@sd.aonix.com Mon Jan 26 11:45:08 1998 Return-Path: Received: from cs.ida.org by cronus.csed.ida.org (SMI-8.6/SMI-SVR4) id LAA05767; Mon, 26 Jan 1998 11:45:05 -0500 Received: from mail.acm.org (mail.acm.org [199.222.69.4]) by cs.ida.org (8.8.7/8.8.7) with ESMTP id LAA04966 for ; Mon, 26 Jan 1998 11:46:16 -0500 (EST) Received: from sw-eng.falls-church.va.us (ns1.sw-eng.falls-church.va.us [199.75.54.2]) by mail.acm.org (8.8.5/8.7.5) with ESMTP id LAA43850; Mon, 26 Jan 1998 11:45:22 -0500 Received: from gw.alsys.com by sw-eng.falls-church.va.us (8.8.8/) id QAA14234; Mon, 26 Jan 1998 16:41:04 GMT Received: from rasht.sd.aonix.com (mailhub.alsys.com) by gw.alsys.com (4.1/SMI-4.1.1) id AA00932; Mon, 26 Jan 98 08:43:23 PST Received: from puumba.telesoft by rasht.sd.aonix.com (4.1/TS-1.2c) id AA12868; Mon, 26 Jan 98 08:41:57 PST Received: by puumba.telesoft (SMI-8.6/SMI-SVR4) id IAA13302; Mon, 26 Jan 1998 08:41:56 -0800 Date: Mon, 26 Jan 1998 08:41:56 -0800 From: sblake@sd.aonix.com (Steve Blake @pulsar) Message-Id: <199801261641.IAA13302@puumba.telesoft> To: asis-comment@sw-eng.falls-church.va.us, dewar@gnat.com, rybin@possum.srcc.msu.su Subject: Re: 10.18: what is the same physical CU? Content-Length: 272 Status: OR The use of the term "program text" (in place of time stamp) should be fine to describe the Is_Equal and Is_Identical operations. It is sufficiently abstract but gets the point across. We can avoid the use of details like time stamps, file names, sources, etc. Steve From rybin@possum.srcc.msu.su Mon Jan 26 16:37:52 1998 Return-Path: Received: from cs.ida.org by cronus.csed.ida.org (SMI-8.6/SMI-SVR4) id QAA06894; Mon, 26 Jan 1998 16:37:51 -0500 Received: from mail.acm.org (mail.acm.org [199.222.69.4]) by cs.ida.org (8.8.7/8.8.7) with ESMTP id QAA14206 for ; Mon, 26 Jan 1998 16:39:03 -0500 (EST) Received: from sw-eng.falls-church.va.us (ns1.sw-eng.falls-church.va.us [199.75.54.2]) by mail.acm.org (8.8.5/8.7.5) with ESMTP id QAA20856; Mon, 26 Jan 1998 16:37:00 -0500 Received: from crocus.gamma.ru by sw-eng.falls-church.va.us (8.8.8/) id VAA22177; Mon, 26 Jan 1998 21:32:43 GMT Received: from srcc.UUCP (uucp@localhost) by crocus.gamma.ru (8.8.7/8.7.3) with UUCP id AAA25582; Tue, 27 Jan 1998 00:36:07 +0300 (MSK) Received: by gamma.srcc.msu.su; Tue, 27 Jan 1998 00:35:45 +0300 Received: by possum.srcc.msu.su (UUPC/@ v5.09gamma, 14Mar93); Tue, 27 Jan 1998 00:04:28 +0300 To: asis-comment@sw-eng.falls-church.va.us, dewar@gnat.com, sblake@sd.aonix.com References: <199801261641.IAA13302@puumba.telesoft> Message-Id: Organization: Information Systems, SRCC, MSU From: "Sergey I. Rybin" Date: Tue, 27 Jan 98 00:04:28 +0300 X-Mailer: BML [MS/DOS Beauty Mail v.1.36] Subject: Re: 10.18: what is the same physical CU? Lines: 15 Content-Length: 613 Status: OR > The use of the term "program text" (in place of time stamp) should be fine > to describe the Is_Equal and Is_Identical operations. It is sufficiently > abstract but gets the point across. We can avoid the use of details like > time stamps, file names, sources, etc. I agree with Steve, I also like "program text" more, then "time stamp", if we use it just in the same way, as RM 95. But there is another question - should we use the notion of a version of a compilation unit for defining Is_Equal and Is_Identical? Now two of us - Robert and me - think, that we should. Any other opinions? Sergey From dewar@gnat.com Sun Feb 1 08:35:43 1998 Return-Path: Received: from cs.ida.org by cronus.csed.ida.org (SMI-8.6/SMI-SVR4) id IAA21348; Sun, 1 Feb 1998 08:35:42 -0500 Received: from mail.acm.org (mail.acm.org [199.222.69.4]) by cs.ida.org (8.8.7/8.8.7) with ESMTP id IAA06001 for ; Sun, 1 Feb 1998 08:36:58 -0500 (EST) Received: from sw-eng.falls-church.va.us (ns1.sw-eng.falls-church.va.us [199.75.54.2]) by mail.acm.org (8.8.5/8.7.5) with ESMTP id IAA43220; Sun, 1 Feb 1998 08:36:17 -0500 Received: from nile.gnat.com by sw-eng.falls-church.va.us (8.8.8/) id NAA00035; Sun, 1 Feb 1998 13:32:46 GMT Received: by nile.gnat.com (5.0/1.20) id AA28605; Sun, 1 Feb 98 08:35:39 EST Date: Sun, 1 Feb 98 08:35:39 EST From: dewar@gnat.com (Robert Dewar) Message-Id: <9802011335.AA28605@nile.gnat.com> To: asis-comment@sw-eng.falls-church.va.us, dewar@gnat.com, rybin@possum.srcc.msu.su, sblake@sd.aonix.com Subject: Re: 10.18: what is the same physical CU? Content-Length: 434 Status: OR <> I agree, this sounds exactly at the right level. One might add a note that the details of what this means are implementation defined (i.e. require ID documentatoin). From roby Thu Feb 12 16:31:57 1998 Return-Path: Received: by cronus.csed.ida.org (SMI-8.6/SMI-SVR4) id QAA22085; Thu, 12 Feb 1998 16:31:52 -0500 From: roby (Clyde Roby) Message-Id: <199802122131.QAA22085@cronus.csed.ida.org> Subject: Re: 10.18: what is the same physical CU? To: rybin@gnat.com (Sergey Rybin), Dewar@GNAT.Com (Robert Dewar), SBlake@Aonix.Com (Steve Blake) Date: Thu, 12 Feb 1998 16:31:51 -0500 (EST) Cc: roby (Clyde Roby), Colket@ACM.Org (Currie Colket), ASIS-Comment@SW-Eng.Falls-Church.Va.US (ASIS-Comment) X-Mailer: ELM [version 2.4 PL25] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Content-Length: 1140 Status: OR Sergey, Robert, Steve, The discussion on 10.17 (Is_Equal) and 10.18 (Is_Identical) has been converted to Issue #089, "What is the same physical CU (compilation unit)". Please look over the problem statement and proposed solution below and provide any further comments. Thank you. Clyde and Currie ------- !ASIS Issue #089 !topic What is the same physical CU (compilation unit) !reference ASIS 95-10.18 !from Sergey Rybin 98-01-19 !keywords compilation unit time stamp Is_Identical !discussion Clause 10.17 (function Is_Equal) and Clause 10.18 (function Is_Identical) should not discriminate based on compilation time stamp. Instead, they should discriminate on using program text. !resolution Open !date 98-01-19 !Notes ------- PROPOSED SOLUTION: Change the parenthetical phrase on both 10.17 (Is_Equal) and 10.18 (Is_Identical) from: ("The same physical compilation unit" have the same name and the same compilation time stamp.) to: ("The same physical compilation unit" have the same name and the same program text. The details of what this means are implementation-defined, i.e., require ID documentation.) ------- From rybin@gnat.com Fri Feb 13 09:14:58 1998 Return-Path: Received: from cs.ida.org by cronus.csed.ida.org (SMI-8.6/SMI-SVR4) id JAA23278; Fri, 13 Feb 1998 09:13:42 -0500 Received: from lglsun11.epfl.ch (lglsun11.epfl.ch [128.178.76.29]) by cs.ida.org (8.8.7/8.8.7) with ESMTP id JAA14604 for ; Fri, 13 Feb 1998 09:14:57 -0500 (EST) Received: from lglsun4.epfl.ch (lglsun4 [128.178.76.9]) by lglsun11.epfl.ch (8.8.X/EPFL-8.1a) with ESMTP id PAA11935; Fri, 13 Feb 1998 15:12:51 +0100 (MET) Received: from lglsun4 (localhost [127.0.0.1]) by lglsun4.epfl.ch (8.8.X/EPFL-8.1a) with SMTP id PAA04082; Fri, 13 Feb 1998 15:12:20 +0100 (MET) Sender: rybin@lglsun11.epfl.ch Message-ID: <34E454A5.42AF@gnat.com> Date: Fri, 13 Feb 1998 15:11:49 +0100 From: Sergey Rybin Organization: Ada Core Technologies X-Mailer: Mozilla 3.04Gold (X11; I; SunOS 5.5.1 sun4u) MIME-Version: 1.0 To: Clyde Roby CC: Robert Dewar , Steve Blake , Currie Colket , ASIS-Comment Subject: Re: 10.18: what is the same physical CU? References: <199802122131.QAA22085@cronus.csed.ida.org> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Length: 1328 Status: OR Clyde Roby wrote: > > Sergey, Robert, Steve, > > The discussion on 10.17 (Is_Equal) and 10.18 (Is_Identical) > has been converted to Issue #089, "What is the same physical CU > (compilation unit)". Please look over the problem statement and > proposed solution below and provide any further comments. Thank you. > > Clyde and Currie > > PROPOSED SOLUTION: > > Change the parenthetical phrase on both 10.17 (Is_Equal) and 10.18 > (Is_Identical) from: > > ("The same physical compilation unit" have the same name > and the same compilation time stamp.) > > to: > > ("The same physical compilation unit" have the same name > and the same program text. The details of what this means > are implementation-defined, i.e., require ID documentation.) > ------- I would suggest: ("The same physical compilation unit" have the same version, as defined by RM 95, E.3 (5)) and the same program text) Justification: 1. Having the same program text implies having the same name. 2. Having the same version does not imply having the same program text. 3. If we would like to have Is_Equial as a test, that two units have the same semantic, we need the notion of a unit version. 4. RM 95 E.3 (5) also implies, that "details are implementation-defined". Sergey From sblake@sd.aonix.com Fri Feb 13 14:08:47 1998 Return-Path: Received: from cs.ida.org by cronus.csed.ida.org (SMI-8.6/SMI-SVR4) id OAA25450; Fri, 13 Feb 1998 14:08:46 -0500 Received: from gw.sd.aonix.com (gw.alsys.com [136.175.17.2]) by cs.ida.org (8.8.7/8.8.7) with SMTP id OAA23138 for ; Fri, 13 Feb 1998 14:10:02 -0500 (EST) Received: from rasht.sd.aonix.com (mailhub.alsys.com) by gw.sd.aonix.com (4.1/SMI-4.1.1) id AA01411; Fri, 13 Feb 98 11:06:49 PST Received: from puumba.telesoft by rasht.sd.aonix.com (4.1/TS-1.2c) id AA04791; Fri, 13 Feb 98 11:05:11 PST Received: by puumba.telesoft (SMI-8.6/SMI-SVR4) id LAA00644; Fri, 13 Feb 1998 11:05:10 -0800 Date: Fri, 13 Feb 1998 11:05:10 -0800 From: sblake@sd.aonix.com (Steve Blake @pulsar) Message-Id: <199802131905.LAA00644@puumba.telesoft> To: roby@ida.org, rybin@gnat.com Subject: Re: 10.18: what is the same physical CU? Cc: ASIS-Comment@sw-eng.falls-church.va.us, Colket@ACM.Org, Dewar@gnat.com, SBlake@aonix.com Content-Length: 562 Status: OR >I would suggest: > > ("The same physical compilation unit" have the same version, > as defined by RM 95, E.3 (5)) and the same program text) > >Justification: > >1. Having the same program text implies having the same name. >2. Having the same version does not imply having the same program text. >3. If we would like to have Is_Equial as a test, that two units have the > same semantic, we need the notion of a unit version. >4. RM 95 E.3 (5) also implies, that "details are >implementation-defined". > >Sergey This looks good to me. Steve