Loading ...
Sorry, an error occurred while loading the content.

AutoTrace 0.24a on Tru64: No-go.

Expand Messages
  • Stefan A. Deutscher
    Hi all, I d love to (try to) contribute Tru64 (Alpha) and OS/2 builds, but I can t get the source to begin with. Is it just me, or is the download really slow
    Message 1 of 12 , Sep 20, 2000
    • 0 Attachment
      Hi all,

      I'd love to (try to) contribute Tru64 (Alpha) and OS/2 builds, but I
      can't get the source to begin with. Is it just me, or is the download
      really slow / near impossible. Netscape gives up after about 24kB, wget
      makes it up to about 70kB, and over the course of a few hours of
      retries, finally gets the archive.

      Next: It seems the permissions in the archive are a bit off, the
      directory autotrace-0.24a probably should be executable, too, so that
      one can change to it. Did that by hand for now, but still.


      drwxr--r-- 3 admin admin 8192 Jun 14 18:40 autotrace-0.23/
      drw-r--r-- 2 admin admin 8192 Sep 19 20:26 autotrace-0.24a/


      This is the Tru64 build (failure) story:

      Next: Running configure is an uneventful sucess:

      creating cache ./config.cache
      checking for a BSD compatible install... ./install-sh -c
      checking whether build environment is sane... yes
      checking whether make sets ${MAKE}... yes
      checking for working aclocal... missing
      checking for working autoconf... missing
      checking for working automake... missing
      checking for working autoheader... missing
      checking for working makeinfo... found
      checking for gcc... gcc
      checking whether the C compiler (gcc ) works... yes
      checking whether the C compiler (gcc ) is a cross-compiler... no
      checking whether we are using GNU C... yes
      checking whether gcc accepts -g... yes
      checking how to run the C preprocessor... gcc -E
      checking for ranlib... ranlib
      checking for gtkDPS-config... no
      configure: warning: Cannot find GtkDPS; GUI frontend is omitted
      updating cache ./config.cache
      creating ./config.status
      creating Makefile


      However, make dies early, I tried both the Tru64 make as also gnumake
      3.79:

      gcc -I. -I. -g -O2 -c atou.c
      cpp: .deps/atou.pp: No such file or directory
      *** Exit 1
      Stop.


      This was for gcc 2.95.2. After removing config.cache, setting CC to cc,
      and rerunning configure and make, I get the same with the DECompaq cc
      compiler:

      cc -I. -I. -g -c atou.c
      cc: Severe: No such file or directory
      ... file is '.deps/atou.pp'
      *** Exit 1
      Stop.


      Seems that somehow the 'DEPS_MAGIC := ...' line in the Makefile doesn't
      do the magic it's supposed to for me.
      Also, the makefile going through great length makeing me believe it's
      compiling, yet it calls .dep stuff:

      %.o: %.c
      @echo '$(COMPILE) -c $<'; \
      $(COMPILE) -Wp,-MD,.deps/$(*F).pp -c $<
      @-cp .deps/$(*F).pp .deps/$(*F).P; \
      tr ' ' '\012' < .deps/$(*F).pp \
      | sed -e 's/^\\$$//' -e '/^$$/ d' -e '/:$$/ d' -e 's/$$/ :/' \
      >> .deps/$(*F).P; \
      rm .deps/$(*F).pp



      Changing this to

      %.o: %.c
      $(COMPILE) -c $<

      actually starts the compile, but now I am stuck on:


      cc: Error: fit.c, line 1655: In the initializer for worst_error, "FLT_MIN" is not declared. (undeclared)
      real worst_error = FLT_MIN;
      ---------------------^
      *** Exit 1

      which should be in <float.h>, but that gets included only under the
      condition that FLT_MAX is not define.

      #ifndef FLT_MAX
      #include <float.h>
      #endif

      Changing that such that fit.c _always_ includes float.h results in a
      build, with lots of pointer mismatch warnings, maybe because the Alpha
      is 64 bit and the code somehow assumes the world was 32bit?


      cc -I. -I. -g -c input-pnm.c
      cc: Warning: input-pnm.c, line 170: In this statement, the referenced type of the pointer value "buf" is "signed char", which is not compatible with "unsigned char". (ptrmismatch)
      pnmscanner_gettoken (scan, buf, BUFLEN);
      -----------------------------^
      cc: Warning: input-pnm.c, line 188: In this statement, the referenced type of the pointer value "buf" is "signed char", which is not compatible with "unsigned char". (ptrmismatch)
      pnmscanner_gettoken(scan, buf, BUFLEN);
      ----------------------------^
      cc: Warning: input-pnm.c, line 195: In this statement, the referenced type of the pointer value "buf" is "signed char", which is not compatible with "unsigned char". (ptrmismatch)
      pnmscanner_gettoken(scan, buf, BUFLEN);
      ----------------------------^
      cc: Warning: input-pnm.c, line 204: In this statement, the referenced type of the pointer value "buf" is "signed char", which is not compatible with "unsigned char". (ptrmismatch)
      pnmscanner_gettoken(scan, buf, BUFLEN);
      --------------------------------^
      cc: Warning: input-pnm.c, line 265: In this statement, the referenced type of the pointer value "buf" is "signed char", which is not compatible with "unsigned char". (ptrmismatch)
      pnmscanner_gettoken(scan, buf, BUFLEN);
      --------------------------------------------^
      cc: Warning: input-pnm.c, line 267: In this statement, the referenced type of the pointer value "buf" is "signed char", which is not compatible with "unsigned char". (ptrmismatch)
      pnmscanner_getsmalltoken(scan, buf);
      -------------------------------------------------^



      Anyway, the executable shows the help data, but it produces
      faulty postscript output that looks as if some format statement in a
      prinf() call got confused. Here are the first few lines:


      "!PS-Adobe-3.0 EPSF-3.0"
      %%Creator: Adobe Illustrator by AutoTrace version 0.24a
      %%Title:
      %%CreationDate: Wed Sep 20 17:50:17 2000
      %%BoundingBox: 0 0 640 480
      "%DocumentData: Clean7Bit"
      "%EndComments"
      "%BeginProlog"
      "/bd { bind def } bind def"
      "/incompound false def"
      "/m { moveto } bd"
      "/l { lineto } bd"
      "/c { curveto } bd"
      "/F { incompound not {fill} if } bd"
      "/f { closepath F } bd"
      "/*u { /incompound true def } bd"
      "/*U { /incompound false def f} bd"
      "/k { setcmykcolor } bd"
      "%EndProlog"
      "%BeginSetup"
      "%EndSetup"
      0.000000 0.000000 0.000000 1.000000 k
      36 471 "m" "\n"35 464 "l" "\n"38 464 "l" "\n"38 465 "l" "\n"36 471 "l" "\n""f"
      0.000000 0.000000 0.000000 1.000000 k



      So, for autotrace 0.24a isn't working on Alpha Tru64 (4.0F). Any fixes
      are most welcome. Dunno when I'll get around to trying it on my OS/2
      box.

      Cheers,
      Stefan

      --
      ============================================================================
      Stefan A. Deutscher |
      Donostia International Physics Center | office: ++34-943-018174
      Universidad del Pais Vasco, Facultad de Quimica | fax : ++34-943-015600
      Departamento de Fisica de Materiales | home : ++34-943-270647
      Apartado 1072, San Sebastian 20080, Spain | email : sad@...
      ============================================================================
      The early bird gets the worm but the second mouse gets the cheese.
    • Masatake YAMATO
      I ve also used ALPHA: alpha036 % uname -a OSF1 alpha036.aist-nara.ac.jp V4.0 1229 alpha I ve tried to compile. ### With vender s make alpha036 % make gcc -I.
      Message 2 of 12 , Sep 20, 2000
      • 0 Attachment
        I've also used ALPHA:
        alpha036 % uname -a
        OSF1 alpha036.aist-nara.ac.jp V4.0 1229 alpha

        I've tried to compile.

        ### With vender's make
        alpha036 % make
        gcc -I. -I. -g -O2 -c atou.c
        cpp: .deps/atou.pp: No such file or directory
        *** Exit 1
        Stop.

        ### With GNU make
        alpha036 % gmake
        gcc -I. -I. -g -O2 -c atou.c
        gcc -I. -I. -g -O2 -c fit.c
        fit.c: In function `fit_with_least_squares':
        fit.c:1191: warning: cast from pointer to integer of different size
        fit.c:1204: warning: cast from pointer to integer of different size
        fit.c: In function `find_error':
        fit.c:1655: `FLT_MIN' undeclared (first use in this function)
        fit.c:1655: (Each undeclared identifier is reported only once
        fit.c:1655: for each function it appears in.)
        fit.c: In function `append_index':
        fit.c:1919: warning: cast from pointer to integer of different size
        fit.c:1919: warning: cast from pointer to integer of different size
        gmake: *** [fit.o] Error 1

        ### Gcc version
        alpha036 % gcc -v
        Reading specs from /usr/local/lib/gcc-lib/alpha-dec-osf4.0f/2.95/specs
        gcc version 2.95 19990728 (release)

        > Seems that somehow the 'DEPS_MAGIC := ...' line in the Makefile doesn't
        > do the magic it's supposed to for me.
        Does gnu make not do the magic?
        See above the result of mine.

        > which should be in <float.h>, but that gets included only under the
        > condition that FLT_MAX is not define.
        >
        > #ifndef FLT_MAX
        > #include <float.h>
        > #endif
        >
        > Changing that such that fit.c _always_ includes float.h results in a
        > build, with lots of pointer mismatch warnings, maybe because the Alpha
        > is 64 bit and the code somehow assumes the world was 32bit?

        FLT_MAX is also defined in limits.h.
        I've added following #ifdef-#endif after FLT_MAX related block.

        #ifndef FLT_MIN
        #include <float.h>
        #endif

        Then I'll invoke gnumake again.
        alpha036 % gmake
        gcc -I. -I. -g -O2 -c atou.c
        gcc -I. -I. -g -O2 -c fit.c
        fit.c: In function `fit_with_least_squares':
        fit.c:1194: warning: cast from pointer to integer of different size
        fit.c:1207: warning: cast from pointer to integer of different size
        fit.c: In function `append_index':
        fit.c:1922: warning: cast from pointer to integer of different size
        fit.c:1922: warning: cast from pointer to integer of different size
        gcc -I. -I. -g -O2 -c main.c
        gcc -I. -I. -g -O2 -c remove-suffx.c
        gcc -I. -I. -g -O2 -c bitmap.c
        gcc -I. -I. -g -O2 -c make-suffix.c
        gcc -I. -I. -g -O2 -c spline.c
        spline.c: In function `append_spline':
        spline.c:110: warning: cast from pointer to integer of different size
        spline.c:110: warning: cast from pointer to integer of different size
        spline.c: In function `concat_spline_lists':
        spline.c:128: warning: cast from pointer to integer of different size
        spline.c:128: warning: cast from pointer to integer of different size
        spline.c: In function `append_spline_list':
        spline.c:180: warning: cast from pointer to integer of different size
        spline.c:180: warning: cast from pointer to integer of different size
        gcc -I. -I. -g -O2 -c concat3.c
        gcc -I. -I. -g -O2 -c getopt.c
        gcc -I. -I. -g -O2 -c output-eps.c
        gcc -I. -I. -g -O2 -c substring.c
        gcc -I. -I. -g -O2 -c curve.c
        curve.c: In function `append_point':
        curve.c:69: warning: cast from pointer to integer of different size
        curve.c:69: warning: cast from pointer to integer of different size
        curve.c: In function `log_curve':
        curve.c:98: warning: cast from pointer to integer of different size
        curve.c: In function `log_entire_curve':
        curve.c:159: warning: cast from pointer to integer of different size
        curve.c: In function `append_curve':
        curve.c:223: warning: cast from pointer to integer of different size
        curve.c:223: warning: cast from pointer to integer of different size
        curve.c: In function `append_curve_list':
        curve.c:265: warning: cast from pointer to integer of different size
        curve.c:265: warning: cast from pointer to integer of different size
        gcc -I. -I. -g -O2 -c getopt1.c
        gcc -I. -I. -g -O2 -c output-fig.c
        gcc -I. -I. -g -O2 -c edge.c
        gcc -I. -I. -g -O2 -c input-pnm.c
        gcc -I. -I. -g -O2 -c output-sk.c
        gcc -I. -I. -g -O2 -c epsilon-equal.c
        gcc -I. -I. -g -O2 -c output-svg.c
        gcc -I. -I. -g -O2 -c vector.c
        gcc -I. -I. -g -O2 -c extend-fname.c
        gcc -I. -I. -g -O2 -c version.c
        gcc -I. -I. -g -O2 -c output.c
        gcc -I. -I. -g -O2 -c input-bmp.c
        gcc -I. -I. -g -O2 -c xfopen.c
        gcc -I. -I. -g -O2 -c xfseek.c
        gcc -I. -I. -g -O2 -c output-p2e.c
        gcc -I. -I. -g -O2 -c color.c
        gcc -I. -I. -g -O2 -c find-suffix.c
        gcc -I. -I. -g -O2 -c logreport.c
        gcc -I. -I. -g -O2 -c pxl-outline.c
        pxl-outline.c: In function `append_pixel_outline':
        pxl-outline.c:130: warning: cast from pointer to integer of different size
        pxl-outline.c:130: warning: cast from pointer to integer of different size
        pxl-outline.c: In function `append_outline_pixel':
        pxl-outline.c:182: warning: cast from pointer to integer of different size
        pxl-outline.c:182: warning: cast from pointer to integer of different size
        gcc -I. -I. -g -O2 -c output-dxf.c
        gcc -I. -I. -g -O2 -c output-dxf12.c
        gcc -I. -I. -g -O2 -c median.c
        gcc -g -O2 -o autotrace atou.o fit.o main.o remove-suffx.o bitmap.o make-suffix.o spline.o concat3.o getopt.o output-eps.o substring.o curve.o getopt1.o output-fig.o edge.o input-pnm.o output-sk.o epsilon-equal.o output-svg.o vector.o extend-fname.o version.o output.o input-bmp.o xfopen.o xfseek.o output-p2e.o color.o find-suffix.o logreport.o pxl-outline.o output-dxf.o output-dxf12.o median.o -lm
        alpha036 %
      • Masatake YAMATO
        ... With gcc (and gnumake), following patch kills pointer mismatch warnings . Masatake p.s. ./a.out sizeof long == 8 sizeof void * == 8 sizeof int == 4 cd
        Message 3 of 12 , Sep 20, 2000
        • 0 Attachment
          > Changing that such that fit.c _always_ includes float.h results in a
          > build, with lots of pointer mismatch warnings, maybe because the Alpha
          > is 64 bit and the code somehow assumes the world was 32bit?
          With gcc (and gnumake), following patch kills "pointer mismatch warnings".

          Masatake
          p.s.
          ./a.out
          sizeof long == 8
          sizeof void * == 8
          sizeof int == 4

          cd ~/tmp/
          diff -c /home/is/masata-y/tmp/autotrace-0.24a /home/is/masata-y/tmp/autotrace-0.24a.new
          Only in /home/is/masata-y/tmp/autotrace-0.24a.new: LOG
          diff -c /home/is/masata-y/tmp/autotrace-0.24a/curve.c /home/is/masata-y/tmp/autotrace-0.24a.new/curve.c
          *** /home/is/masata-y/tmp/autotrace-0.24a/curve.c Tue May 2 04:08:58 2000
          --- /home/is/masata-y/tmp/autotrace-0.24a.new/curve.c Thu Sep 21 05:10:56 2000
          ***************
          *** 95,101 ****

          if (!log_file) return;

          ! LOG1 ("curve id = %x:\n", (unsigned) curve);
          LOG1 (" length = %u.\n", CURVE_LENGTH (curve));
          if (CURVE_CYCLIC (curve))
          LOG (" cyclic.\n");
          --- 95,101 ----

          if (!log_file) return;

          ! LOG1 ("curve id = %lx:\n", (unsigned long) curve);
          LOG1 (" length = %u.\n", CURVE_LENGTH (curve));
          if (CURVE_CYCLIC (curve))
          LOG (" cyclic.\n");
          ***************
          *** 156,162 ****

          if (!log_file) return;

          ! LOG1 ("curve id = %x:\n", (unsigned) curve);
          LOG1 (" length = %u.\n", CURVE_LENGTH (curve));
          if (CURVE_CYCLIC (curve))
          LOG (" cyclic.\n");
          --- 156,162 ----

          if (!log_file) return;

          ! LOG1 ("curve id = %lx:\n", (unsigned long) curve);
          LOG1 (" length = %u.\n", CURVE_LENGTH (curve));
          if (CURVE_CYCLIC (curve))
          LOG (" cyclic.\n");
          diff -c /home/is/masata-y/tmp/autotrace-0.24a/fit.c /home/is/masata-y/tmp/autotrace-0.24a.new/fit.c
          *** /home/is/masata-y/tmp/autotrace-0.24a/fit.c Tue May 2 04:04:19 2000
          --- /home/is/masata-y/tmp/autotrace-0.24a.new/fit.c Thu Sep 21 05:15:29 2000
          ***************
          *** 16,21 ****
          --- 16,24 ----
          #ifndef FLT_MAX
          #include <float.h>
          #endif
          + #ifndef FLT_MIN
          + #include <float.h>
          + #endif
          #include <string.h>


          ***************
          *** 1187,1194 ****
          if (left_spline_list == NULL)
          {
          WARNING ("could not fit left spline list");
          ! LOG1 ("Could not fit spline to left curve (%x).\n",
          ! (unsigned) left_curve);
          }
          else
          {
          --- 1190,1197 ----
          if (left_spline_list == NULL)
          {
          WARNING ("could not fit left spline list");
          ! LOG1 ("Could not fit spline to left curve (%lx).\n",
          ! (unsigned long) left_curve);
          }
          else
          {
          ***************
          *** 1200,1207 ****
          if (right_spline_list == NULL)
          {
          WARNING ("could not fit right spline list");
          ! LOG1 ("Could not fit spline to right curve (%x).\n",
          ! (unsigned) right_curve);
          }
          else
          {
          --- 1203,1210 ----
          if (right_spline_list == NULL)
          {
          WARNING ("could not fit right spline list");
          ! LOG1 ("Could not fit spline to right curve (%lx).\n",
          ! (unsigned long) right_curve);
          }
          else
          {
          diff -c /home/is/masata-y/tmp/autotrace-0.24a/xmem.h /home/is/masata-y/tmp/autotrace-0.24a.new/xmem.h
          *** /home/is/masata-y/tmp/autotrace-0.24a/xmem.h Sun Feb 20 20:48:26 2000
          --- /home/is/masata-y/tmp/autotrace-0.24a.new/xmem.h Thu Sep 21 05:09:34 2000
          ***************
          *** 41,48 ****
          { \
          new_mem = (address) realloc (old_ptr, size); \
          if (new_mem == NULL) \
          ! FATAL4 ("realloc: request for %x to be %u bytes failed in %s line %d", \
          ! (unsigned) (old_ptr), (size), __FILE__, __LINE__); \
          } \
          \
          old_ptr = new_mem; \
          --- 41,48 ----
          { \
          new_mem = (address) realloc (old_ptr, size); \
          if (new_mem == NULL) \
          ! FATAL4 ("realloc: request for %lx to be %u bytes failed in %s line %d", \
          ! (unsigned long) (old_ptr), (size), __FILE__, __LINE__); \
          } \
          \
          old_ptr = new_mem; \
        • Stefan A. Deutscher
          Hi, ... [ snip cc ] ... [snip] ... No, maybe because I don t have any of that automake stuff installed that seems to be there by default on Linux, but not so
          Message 4 of 12 , Sep 21, 2000
          • 0 Attachment
            Hi,


            > I've also used ALPHA:
            > alpha036 % uname -a
            > OSF1 alpha036.aist-nara.ac.jp V4.0 1229 alpha

            [ snip cc ]

            > ### With GNU make
            > alpha036 % gmake
            > gcc -I. -I. -g -O2 -c atou.c
            > gcc -I. -I. -g -O2 -c fit.c
            > fit.c: In function `fit_with_least_squares':
            > fit.c:1191: warning: cast from pointer to integer of different size
            > fit.c:1204: warning: cast from pointer to integer of different size
            [snip]

            >
            > ### Gcc version
            > alpha036 % gcc -v
            > Reading specs from /usr/local/lib/gcc-lib/alpha-dec-osf4.0f/2.95/specs
            > gcc version 2.95 19990728 (release)
            >
            > > Seems that somehow the 'DEPS_MAGIC := ...' line in the Makefile doesn't
            > > do the magic it's supposed to for me.
            > Does gnu make not do the magic?
            > See above the result of mine.

            No, maybe because I don't have any of that automake stuff installed
            that seems to be there by default on Linux, but not so on Tru64?

            > > which should be in <float.h>, but that gets included only under the
            > > condition that FLT_MAX is not define.
            > >
            > > #ifndef FLT_MAX
            > > #include <float.h>
            > > #endif
            > >
            > > Changing that such that fit.c _always_ includes float.h results in a
            > > build, with lots of pointer mismatch warnings, maybe because the Alpha
            > > is 64 bit and the code somehow assumes the world was 32bit?
            >
            > FLT_MAX is also defined in limits.h.
            > I've added following #ifdef-#endif after FLT_MAX related block.
            >
            > #ifndef FLT_MIN
            > #include <float.h>
            > #endif
            >
            > Then I'll invoke gnumake again.
            > alpha036 % gmake
            > gcc -I. -I. -g -O2 -c atou.c
            > gcc -I. -I. -g -O2 -c fit.c
            > fit.c: In function `fit_with_least_squares':
            > fit.c:1194: warning: cast from pointer to integer of different size
            > fit.c:1207: warning: cast from pointer to integer of different size
            > fit.c: In function `append_index':


            [cut]

            After similar modifications, I got that far, too, just the executable
            produces faulty ps output.

            --
            ============================================================================
            Stefan A. Deutscher |
            Donostia International Physics Center | office: ++34-943-018174
            Universidad del Pais Vasco, Facultad de Quimica | fax : ++34-943-015600
            Departamento de Fisica de Materiales | home : ++34-943-270647
            Apartado 1072, San Sebastian 20080, Spain | email : sad@...
            ============================================================================
            The early bird gets the worm but the second mouse gets the cheese.
          • Stefan A. Deutscher
            Thanks. By the way, does the executable also work afterwards, that is, produce valid PS output? I built an OS/2 version last night (details to follow) that
            Message 5 of 12 , Sep 21, 2000
            • 0 Attachment
              Thanks.

              By the way, does the executable also work afterwards, that is, produce
              valid PS output? I built an OS/2 version last night (details to follow)
              that generates valid PS output, while the Tru64 one I built (w/o your
              patches yet) does not. If your patched version fixes things, that would
              be a great patch to include in 0.24b, together with the FLT_MIN one.


              Cheers,
              Stefan

              > > Changing that such that fit.c _always_ includes float.h results in a
              > > build, with lots of pointer mismatch warnings, maybe because the Alpha
              > > is 64 bit and the code somehow assumes the world was 32bit?
              > With gcc (and gnumake), following patch kills "pointer mismatch warnings".
              >
              > Masatake
              > p.s.
              > ./a.out
              > sizeof long == 8
              > sizeof void * == 8
              > sizeof int == 4
              >
              > cd ~/tmp/
              > diff -c /home/is/masata-y/tmp/autotrace-0.24a /home/is/masata-y/tmp/autotrace-0.24a.new
              > Only in /home/is/masata-y/tmp/autotrace-0.24a.new: LOG
              > diff -c /home/is/masata-y/tmp/autotrace-0.24a/curve.c /home/is/masata-y/tmp/autotrace-0.24a.new/curve.c
              > *** /home/is/masata-y/tmp/autotrace-0.24a/curve.c Tue May 2 04:08:58 2000
              > --- /home/is/masata-y/tmp/autotrace-0.24a.new/curve.c Thu Sep 21 05:10:56 2000
              > ***************
              > *** 95,101 ****
              >
              > if (!log_file) return;
              >

              --
              ============================================================================
              Stefan A. Deutscher |
              Donostia International Physics Center | office: ++34-943-018174
              Universidad del Pais Vasco, Facultad de Quimica | fax : ++34-943-015600
              Departamento de Fisica de Materiales | home : ++34-943-270647
              Apartado 1072, San Sebastian 20080, Spain | email : sad@...
              ============================================================================
              The early bird gets the worm but the second mouse gets the cheese.
            • Masatake YAMATO
              ... I think my patch is nothing to do with PS output but warning output only. I d like to work for this 64bit issue when the project introduce CVS. If you are
              Message 6 of 12 , Sep 21, 2000
              • 0 Attachment
                > By the way, does the executable also work afterwards, that is, produce
                > valid PS output? I built an OS/2 version last night (details to follow)
                > that generates valid PS output, while the Tru64 one I built (w/o your
                > patches yet) does not. If your patched version fixes things, that would
                > be a great patch to include in 0.24b, together with the FLT_MIN one.

                I think my patch is nothing to do with PS output but warning output
                only. I''d like to work for this 64bit issue when the project introduce
                CVS. If you are in hurry, runn gcc with -Wall option. gcc will tell you
                where you should have to inspect.

                Masatake
              • Masatake YAMATO
                Do you have system in that automake is installed? With following steps, I could compile autotrace with vender s make(DEC). 1. Copy autotrace.tar.gz to the
                Message 7 of 12 , Sep 21, 2000
                • 0 Attachment
                  Do you have system in that automake is installed?

                  With following steps, I could compile autotrace with vender's make(DEC).

                  1. Copy autotrace.tar.gz to the system in that automake is installed
                  from your alpha.

                  2. Added following line to the top of Makefile.am in autotrace/.

                  AUTOMAKE_OPTIONS = foreign no-dependencies

                  3. Invoke automake.

                  4. Invoke ./configure

                  5. Invoke make dist
                  You'll get new autotrace.tar.gz.
                  6. Copy new autotrace.tar.gz.to your alpha from the system in
                  that automake is installed.

                  7. Invoke tar zxvf new autotrace.tar.gz on your alpha.

                  8. cd autotrace; ./concfigure; make

                  ---
                  The statement
                  AUTOMAKE_OPTIONS = foreign no-dependencies
                  generates gnumake independent Makefile.in.


                  Masatake
                • Stefan A. Deutscher
                  ... [ step 1 - 8 deleted ] No, I don t, and even if I had: Just changing the compile target in the makefile by hand is faster and works :-) Maybe I don t know
                  Message 8 of 12 , Sep 21, 2000
                  • 0 Attachment
                    On Thu, Sep 21, 2000 at 08:37:13PM +0900, Masatake YAMATO wrote:
                    >
                    > Do you have system in that automake is installed?
                    >
                    > With following steps, I could compile autotrace with vender's make(DEC).

                    [ step 1 - 8 deleted ]

                    No, I don't, and even if I had: Just changing the compile target in the
                    makefile by hand is faster and works :-) Maybe I don't know enough about
                    it, but the makefile seems overly complicated, ooops, ingenious. What
                    happened to the good old days of readable makefiles? ;-)

                    Also, it might be a good idea to have configure generate a little
                    warning in case it can't find all the automake stuff at the end, as in

                    "Some stuff missing. You may need to edit the Makefile by hand."

                    Cheers,
                    Stefan

                    --
                    ============================================================================
                    Stefan A. Deutscher |
                    Donostia International Physics Center | office: ++34-943-018174
                    Universidad del Pais Vasco, Facultad de Quimica | fax : ++34-943-015600
                    Departamento de Fisica de Materiales | home : ++34-943-270647
                    Apartado 1072, San Sebastian 20080, Spain | email : sad@...
                    ============================================================================
                    The early bird gets the worm but the second mouse gets the cheese.
                  • Masatake YAMATO
                    ... Please, see Makefile.am, it is so simple. Anyway, I ll modify Makefile.am as I wrote after CVS repository is opened. Please, wait. ... No. The reason why
                    Message 9 of 12 , Sep 21, 2000
                    • 0 Attachment
                      > No, I don't, and even if I had: Just changing the compile target in the
                      > makefile by hand is faster and works :-) Maybe I don't know enough about
                      > it, but the makefile seems overly complicated, ooops, ingenious. What
                      > happened to the good old days of readable makefiles? ;-)

                      Please, see Makefile.am, it is so simple.
                      Anyway, I'll modify Makefile.am as I wrote after CVS repository
                      is opened. Please, wait.

                      > Also, it might be a good idea to have configure generate a little
                      > warning in case it can't find all the automake stuff at the end, as in
                      >
                      > "Some stuff missing. You may need to edit the Makefile by hand."

                      No. The reason why the make invocation is failed is my mistake.
                      I, the author of Makefile.am should add
                      AUTOMAKE_OPTIONS = foreign no-dependencies
                      to Makefile.am.

                      Masatake
                    Your message has been successfully submitted and would be delivered to recipients shortly.