R From:	CRDGW2::CRDGW2::MRGATE::"SMTP::AI.MIT.EDU::GNULISTS" 23-OCT-1990 23:39:06.77 To:	MRGATE::"ARISIA::EVERHART" CC:	/ Subj:	GNU SOFTWARE C++ COMPILER FOR VMS SYSTEMS   * Received:  by crdgw1.ge.com (5.57/GE 1.76)) 	 id AA19233; Tue, 23 Oct 90 23:33:03 EDT R Received: by life.ai.mit.edu (4.1/AI-4.10) id AA05774; Mon, 22 Oct 90 21:48:22 EDT" Return-Path: <gnulists@ai.mit.edu>z Received: from wheat-chex (wheat-chex.ai.mit.edu) by life.ai.mit.edu (4.1/AI-4.10) id AA05703; Mon, 22 Oct 90 21:47:18 EDTM Received: by wheat-chex (4.1/AI-4.10) id AA05589; Mon, 22 Oct 90 21:47:15 EDT # Resent-Date: 17 Oct 90 05:18:47 GMT   Resent-From: gnulists@ai.mit.edu2 Resent-Message-Id: <9010230147.AA05589@wheat-chex>" To: gnu-g++-bug@cis.ohio-state.eduM From: vkb%BBRVKI51.BITNET@ohstvma.ircc.ohio-state.edu (Patrick VANKEIRSBILCK)  Sender: gnulists@ai.mit.edu 2 Subject: GNU SOFTWARE C++ COMPILER FOR VMS SYSTEMS Date: 17 Oct 90 05:18:47 GMT. X-Unparsable-Date: tue, 16 oct 90 15.15:00 +028 Message-Id: <9010171904.AA07996@giza.cis.ohio-state.edu># Resent-To: help-g++@prep.ai.mit.edu   	 Dear Sir,   1 After some weeks, we finally succeeded in getting G the GnuSoftware C++ compiler running (Distribution FALL89 DECUS) on the = VAX-cluster at the von Karman Institute in Brussels, Belgium.   0 However, still some small problems are existing.  J First of all, the compiler doesn't seem to distinguish between 'float' andJ 'double' variables.  Print statements using the 'cout <<' operation alwaysK give the same number of digits regardless the type of the printed variable.   G Secondly, our linker doesn't find functions like chr(), hex(),... which @ should normally be searched through the 'stream.h' include file.  J Finally, 'inline' functions defined on separatedly compiled files to whichL a 'main' program is linked, are not recognized by our linker.  Nevertheless,H if the program was written and compiled using one single file everythingB went fine.  Perhaps some implementation error was introduced while/ splitting the file, but I couldn't find it out.   7 To give you an idea of how our environment was defined, - some notes that our system manager wrote down = about how he installed the GNU Software C++ Compiler on a VMS $ system version V5.3 are given below:  G =======================================================================   3 Software from FALL 89 DECUS tape saveset LT89B2.BCK   ' The changes from the distribution are :   / 1) Define the logical name GNU_CC:[000000] with   K    $ DEFINE /EXEC /SYSTEM GNU_CC disk_name:[directory.subdir.]/TRANS=CONCEA   < 2) Move the requested files to the GNU_CC:[000000] directory  2 3) Insert the new command into the DCL tables with  K    $ SET COMMAND/TABLES=SYS$LIBRARY:DCLTABLES/OUT=SYS$LIBRARY:DCLTABLES GXX K    $ SET COMMAND/TABLES=SYS$LIBRARY:DCLTABLES/OUT=SYS$LIBRARY:DCLTABLES GCC *    $ INSTALL/REPLACE SYS$LIBRARY:DCLTABLES  > With a cluster use sys$common:[syslib] instead of sys$library:    4) Define some logical name with      $ @GXX_INSTALL   C         -----------------------------------------------------------   - To link the object files use the GXX.OPT file   2    $ LINK prog_name,GNU_CC:[000000]GXX.OPT/OPTIONS    L ============================================================================  G Hoping that this information will be sufficient for you to consider the G problems I mentioned above, I thank you very much for the time you will  be spending on this matter,   6                                       yours sincerely,  ;                                       Patrick VANKEIRSBILCK :                                       von Karman Institute?                                              for Fluid Dynamics ;                                       Waterloosesteenweg 72 =                                       1640 SINT-GENESIUS-RODE -                                       BELGIUM 9                                       vkb@bbrvki51.bitnet   