From: eplan@kapsch.net Sent: Sunday, January 09, 2000 2:47 PM To: Info-VAX@Mvb.Saic.Com Subject: Re: XPDF 0.90 build problems on Alpha In article <858k81$mjl@gap.cco.caltech.edu>, mathog@seqaxp.bio.caltech.edu writes: >In article <387677ac$1@news.kapsch.co.at>, eplan@kapsch.net (Peter LANGSTOEGER) writes: >>I just tried to build XPDF 0.90 on VAX and Alpha. No problem on VAX but a >>lot of messages on Alpha (approx 25000 lines !!) during VMS_MAKE.COM >> >>Most of the messages are like this >> >>extern XActivateScreenSaver( >>........^ >>%CXX-W-NOSIMPINT, omission of explicit type is nonstandard ("int" assumed) >>at line number 2011 in file SYS$COMMON:[DECW$INCLUDE]XLIB.H;1 > >I just went through the same thing this afternon. Here are enough fixes to >get you running: > >1. Add /warn=disable=NOSIMPINT to the CXXCOMP definition in the top > vms_make.com file. Yes. Fighting the symptom is a lot easier than fighting the cause ;-) But preventing 25000 lines in logfile is worth to fight the symptom. And I gave up the idea of finding out the real cause (the CXX V6.2 vs. earlier versions differences are enough reasons for me now). >2. Add a definition in the same file: > > $ mylink :== cxxlink > >3. replace all "$ LINK" with "$ MYLINK" in all vms_make.com files. > >(or you could just replace link in those files directly with CXXLINK.) It's a shame. I should have known this. Now you told me, I remember it. I went already through this with my last XPDF build (0.70, about 2 years ago). Now I wonder why my docu of this got lost... >It should build ok after that, with just a few warnings. (There's one >variable which is used before it is initialized.) On VAX it builds without any message (just like before). On Alpha it builds now ok with the message obj5.free(); ..^ %CXX-W-CODEUNREACHABLE, statement is unreachable at line number 615 in file DISK$SOFT:[SOURCE.XPDF.XPDF-0_90.XPDF]GFXFONT.CC;1 if (dashLength > 0 && cap == CapProjecting) { ......^ %CXX-W-USEDBEFORESET, variable "dashLength" is used before its value is set at line number 1416 in file DISK$SOFT:[SOURCE.XPDF.XPDF-0_90.XPDF]XOUTPUTDEV.CC;1 It seems, that the code has some shortcomings, which C++ V5.6A/V6.0 didn't note and C++ V6.2 (currently Alpha only) now complains at. This makes sense, since C++ V6.2 arrived weeks after XPDF 0.90 went public. >XPDF ran ok once I did the above, except that the "print" command is hosed, >it tries to write to a file '[USRDISK:[USERS.MATHOG]]file.pxdf;1.ps'! >That's a code problem and it looks kind of familiar, so maybe it was around >all the way back in version 0.7. Omit the version number and try again. Beyond that, can't comment. I did this now: $ dif vms_make.com ************ File DISK$SOFT:[SOURCE.XPDF.XPDF-0_90]VMS_MAKE.COM;2 69 $ IF F$GETSYI ("ARCH_NAME") .EQS. "VAX" 70 $ THEN !C++ V5.6A 71 $ cxxcomp :== "CXX /PREFIX=ALL ''defs' ''incs'" 72 $ ELSE !C++ V6.2 73 $ cxxcomp :== "CXX /PREFIX=ALL /WARN=DIS=NOSIMPINT ''defs' ''incs'" 74 $ ENDIF 75 $ link :== "CXXLINK" 76 $ endif ****** File DISK$SOFT:[SOURCE.XPDF.XPDF-0_90]VMS_MAKE.COM;1 69 $ cxxcomp :== "CXX /PREFIX=ALL ''defs' ''incs'" 70 $ endif ************ Number of difference sections found: 1 Number of difference records found: 7 DIFFERENCES /IGNORE=()/MERGED=1- DISK$SOFT:[SOURCE.XPDF.XPDF-0_90]VMS_MAKE.COM;2- DISK$SOFT:[SOURCE.XPDF.XPDF-0_90]VMS_MAKE.COM;1 Many thanks. -- Peter "EPLAN" LANGSTOEGER Tel. +43 1 81111-2651 Network and OpenVMS system manager Fax. +43 1 81111-888 FBFV/Information Services E-mail eplan@kapsch.net <<< KAPSCH AG Wagenseilgasse 1 PSImail PSI%(0232)281001141::EPLAN A-1121 VIENNA AUSTRIA "I'm not a pessimist, I'm a realist" "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998