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

Vim version 6.0at BETA available

Expand Messages
  • Bram Moolenaar
    This is the fifth BETA for Vim 6.0. Again a number of problems were fixed. All the pending items have been done. It s now really bug fixing only! I expect
    Message 1 of 17 , Aug 27 12:57 PM
    • 0 Attachment
      This is the fifth BETA for Vim 6.0. Again a number of problems were fixed.

      All the pending items have been done. It's now really bug fixing only!
      I expect the 6.0 release in just a few weeks now. Please try to find every
      bug, documentation mistake, omission, etc.

      I noticed that some people didn't report problems, assuming someone else would
      do that. Well, if you still see a problem now, apparently nobody reported it
      yet, so you better do that _now_!


      Major changes
      -------------

      On MS-Windows, when compiled with gettext support, detect the language and set
      $LANG to get the appropriate translated messages (if supported). Removed the
      message box in NSIS that suggests to set $LANG manually. Also use $LANG to
      select a menu language, v:lang is a very different kind of name.

      Made gvimext.dll use translated messages, if possible. (Yasuhiro Matsumoto)

      Added the "[c" and "]c" commands to move to the next/previous change in diff
      mode.


      Other changes
      -------------

      Made the default mappings for <S-Insert> in Insert mode insert the text
      literally, avoids that special characters like BS cause side effects.

      Made CTRL-R CTRL-O at the command line work like CTRL-R CTRL-R, so that it's
      consistent with Insert mode.

      Added Makefile for Borland 5 to compile gvimext.dll. (Yasuhiro Matsumoto)

      Added lispindent() and cindent() functions: Can be used from an indent script.

      PostScript printing supports underlining and background colors. Improved the
      produced output. Added the $VIMRUNTIME/procset.ps file. Made it possible to
      compile with PostScript printing on MS-Windows. (Mike Williams)

      Added "jobsplit" to 'printoptions': Make each copy a new job, for N-up
      printing. For Win32: Add the progress status to the Cancel message box.
      (partly by Vipin Aravind)

      Cleaned up names of variables used for X11 resources. Removed a few unused
      global variables.


      Fixes
      -----

      The Motif version with normal features didn't compile.

      Unix: Installing the icons ignored $DESTDIR.

      Win32 GUI: The tooltip for tear-off items remained when the tear-off item was
      no longer selected.

      Win32 client-server:
      - GUI: The --remote command wasn't handled until the mouse pointer was moved
      over the window. (Paul Moore)
      - Console: The --remote command wasn't handled until after a delay.
      - Evaluating an expression didn't work.
      - Console: Use the input buffer instead of the typeahead buffer to avoid
      problems.

      GUI: When typing ":" at the more prompt, would return to Normal mode and not
      redraw the screen.

      When in diff mode and moving over a large number of filler lines, the scroll
      binding could be wrong. CTRL-U, CTRL-D and CTRL-F could get stuck at one
      position near a large number of filler lines.
      In diff mode, in a block of changed lines, an inserted line was not
      highlighted as being inserted.

      When starting Vim with an argument "-c g/at/p" the printed lines would
      overwrite eachother.

      BeOS: Didn't compile. Configure didn't add the os_beos files, the QNX check
      removed them. Various changes to os_beos.cc. (Joshua Haberman)
      Removed the check for the hardware platform, the BeBox has not been produced
      for a long time now.

      Win32 GUI: don't use a message box when the shell returns an error code,
      display the message in the Vim window.

      Make_mvc.mak always included "/debug" for linking. "GUI=no" argument didn't
      work. Use "DEBUG=yes" instead of "DEBUG=1" to make it consistent. (Dan Sharp)

      The BufWinLeave event was triggered when a buffer was wiped out that was not
      displayed in a window.

      When a line in the tags file ended in ;" (no TAB following) the command would
      not be recognized as a search command.

      Refreshing the Buffers menu caused unlisted buffers to appear.

      Motif and Athena: The menu font did not adjust to the locale. Now it uses a
      default fontset that should work. (David Harrison)

      X11: The inputMethod resource never worked. Don't use the "none" input method
      for SGI, it apparently makes the first character in Input method dropped.

      Calling foreground() when the GUI is not running caused a crash.

      When using ":mkview", would add ":2next", which doesn't always work. Now only
      add that when the view is part of a session.

      When 'list' and 'hlsearch' are on, searching for a character at the end of the
      line caused a space after the "$" to be highlighted.

      Fixed incorrect tests in os_mac.h. (Axel Kielhorn)

      Win32 console: When the console where Vim runs in is closed, Vim could hang in
      trying to restore the window icon. (Yasuhiro Matsumoto)

      The gzip.vim plugin used ":set ma", which changed the global default for the
      'modifiable' option. Use ":setlocal ma" instead.

      When using ":silent ts" would wait for the user to type a number, but the
      prompt wasn't displayed. Fixed a few other places to skip getting a response
      or reset msg_silent to force getting a prompt.

      When swapping screens, starting Vim halfway the xterm, the cursor was moved to
      the bottom line after exiting.

      When using ":3call func()" or ":3,3call func() the line number was ignored.

      When hitting CR in the quickfix window, would display the error message, even
      though it's already visible in the error window.

      Using "vim >& file" made Vim hang, it would not read from stdin if stdout
      isn't connected to a tty.

      Ruby: the if_ruby.o file was not put in the objects directory.

      Vimdiff: After ":diffput" undo was not synced in other windows, would undo all
      ":diffput" commands at once.
      ":diffput" below a filler line didn't update the following diff blocks
      correctly. Could result in another diff block in the next line to be put as
      well.

      Use "int_u" instead of "uint", not all systems have "uint".

      For each GUI there were three functions to return color values. Reduced that
      to one and added a generic function to compute the grey level from an RGB
      value.

      Multi-byte: When 'cpo' contains "$", "cw" that ends on a multi-byte character
      put the "$" in the next character. And it could leave the right halve of a
      double-wide character, now it's blanked. And when typing a single-width
      character over a double-width character the right halve would remain.

      Unix: When installing the non-GUI version, would still create a link for
      "gvimdiff".

      ":diffpatch" allocated a buffer two bytes too small, could cause a crash.
      (Servatius Brandt) When using a relative file name, would not find the patch
      file.

      Win32 GUI: When executing an external command failed Vim would hang.

      When 'showbreak' and 'linebreak' were both set, Visual highlighting sometimes
      continued until the end of the line.

      In an ever growing file, was getting "file changed" dialogs over and over.

      Folds were not updated correctly when using markers and deleting a line with
      an end marker just above the start of another fold at the same level.

      When 'virtualedit' is "block" or "all", using "I#" when the selected block
      starts on a TAB didn't work correctly.

      Using 'virtualedit' in Insert mode: When typing at the end of the line a
      character replaced the last one in the line.

      When using ":map <silent>" that includes inputdialog(), the prompt would not
      be given until a <BS> is used.

      Using ":startinsert" in a vimrc file didn't work.

      VisVim: When using a BufRead autocommand with a system() call, the connection
      between Visual studio and gvim didn't work, got "server busy" errors. Was
      caused by the wait for the system command dropping messages other than repaint
      messages.



      Epilogue
      --------

      WARNING: This is still an unstable version. Most functionality should work
      fine, but in some cases it might crash or even destroy your work.

      If you run into something that doesn't work, please try to figure out why,
      try to solve it and send me a patch. If you can't do that, at least let me
      know exactly how to reproduce the problem.

      More info about the new 6.0 features with ":help version6". Or look here:
      http://vim.sf.net/whyvim.php.

      If you don't like the syntax of a command, the name of an option or how the
      new features work, let's discuss this in the vim-dev maillist.


      You can find Vim 6.0 here: ftp://ftp.vim.org/pub/vim/unreleased/
      More information about downloading: http://vim.sf.net/download.php

      Unix:
      unix/vim-6.0at.tar.bz2 sources + runtime files, bzip2 compressed
      unix/vim-6.0at-rt1.tar.gz runtime files part 1
      unix/vim-6.0at-rt2.tar.gz runtime files part 2
      unix/vim-6.0at-src1.tar.gz sources part 1
      unix/vim-6.0at-src2.tar.gz sources part 2
      unix/vim-6.0as-6.0at.diff.gz diff between 6.0as and 6.0at

      Various:
      extra/vim-6.0at-extra.tar.gz extra files
      extra/vim-6.0at-lang.tar.gz multi-language files
      extra/vim-6.0as-6.0at-extra.diff.gz extra diff
      extra/vim-6.0as-6.0at-lang.diff.gz multi-lang diff
      doc/vim60athtml.zip help files converted to HTML

      MS-Windows:
      pc/gvim60at.exe self-installing, includes runtime files
      pc/vim60atrt.zip runtime files
      pc/vim60atlang.zip extra files for translated messages and menus
      pc/gvim60at.zip GUI binary for Windows 95/98/NT/2000
      pc/gvim60atole.zip GUI binary with OLE support
      pc/gvim60at_s.zip GUI binary for Windows 3.1
      pc/vim60atd16.zip 16 bits real mode - works on any system
      pc/vim60atd32.zip 32 bits protected mode - needs 386 and DPMI
      pc/vim60atw32.zip console version for Windows NT/2000/XP
      pc/vim60atsrc.zip sources for PC (with CR-LF)

      Amiga:
      amiga/vim60atrt.tgz runtime files
      amiga/vim60atbin.tgz binaries
      amiga/vim60atsrc.tgz sources packed for Amiga

      OS/2:
      os2/vim60atrt.zip runtime files
      os2/vim60atos2.zip binaries


      Happy Vimming!

      --
      hundred-and-one symptoms of being an internet addict:
      180. You maintain more than six e-mail addresses.

      /// Bram Moolenaar -- Bram@... -- http://www.moolenaar.net \\\
      ((( Creator of Vim -- http://vim.sf.net -- ftp://ftp.vim.org/pub/vim )))
      \\\ Help me helping AIDS orphans in Uganda - http://iccf-holland.org ///
    • Preben Guldberg
      Just compiled vim 6.0at with gtk interface on a Solaris 8 box. When started it dumps core with the following message: Vim: Caught deadly signal BUS Vim:
      Message 2 of 17 , Aug 27 2:48 PM
      • 0 Attachment
        Just compiled vim 6.0at with gtk interface on a Solaris 8 box.

        When started it dumps core with the following message:

        Vim: Caught deadly signal BUS
        Vim: Finished.

        Tracked it down to menu.vim, where any of the :amenu (or :am) commands
        would trigger this. This only happens if the right number of arguments
        is provided:

        [put :finish on first line of menu.vim]

        $ gvim60at -u NONE -U NONE --noplugin

        :amenu
        --- Menus ---
        Hit ENTER or type command to continue
        :amenu 9999.10
        E329: No menu of that name
        :amenu 9999.10 &Help.&Overview<Tab><F1>
        E329: No menu of that name
        :amenu 9999.10 &Help.&Overview<Tab><F1> :help<CR>

        And the message above can be seen in the terminal.

        The same command works fine in vim with :amenu followed by :emenu.

        :version
        VIM - Vi IMproved 6.0at BETA (2001 Aug 27, compiled Aug 27 2001
        23:12:00) +statusline -sun_workshop +syntax +tag_binary
        +tag_old_static -tag_any_white -tcl +terminfo +termresponse
        +textobjects +title
        +toolbar +user_commands +vertsplit +virtualedit +visual +visualextra
        +viminfo
        +vreplace +wildignore +wildmenu +windows +writebackup +X11 -xfontset
        +xim
        +xterm_clipboard -xterm_save
        system vimrc file: "$VIM/vimrc"
        user vimrc file: "$HOME/.vimrc"
        user exrc file: "$HOME/.exrc"
        system gvimrc file: "$VIM/gvimrc"
        user gvimrc file: "$HOME/.gvimrc"
        system menu file: "$VIMRUNTIME/menu.vim"
        fall-back for $VIM: "/usr/local/share/vim"
        Compilation: gcc -c -I. -Iproto -DHAVE_CONFIG_H -DFEAT_GUI_GTK
        -I/usr/local/lib/glib/include -I/usr/local/include
        -I/usr/openwin/include -I/usr/local/include -g -O2
        -I/usr/openwin/include
        Linking: gcc -L/usr/openwin/lib -R/usr/openwin/lib -L/usr/local/lib
        -o ggvim60at -L/usr/local/lib -L/usr/openwin/lib -R/usr/openwin/lib
        -lgtk -lgdk -lgmodule -lglib -ldl -lXext -lm -lXt -lncurses


        Hope this is of some help.

        Peppe
        --
        "Before you criticize someone, walk
        Preben "Peppe" Guldberg __/-\__ a mile in his shoes. That way, if
        c928400@... (o o) he gets angry, he'll be a mile away
        ----------------------oOOo (_) oOOo-- - and barefoot." --Sarah Jackson
      • Benoit Cerrina
        Hi Bram, thanks for the new version, I noticed you failed to mention the fact that vim -d file dir now works if the second argument is a directory with a
        Message 3 of 17 , Aug 28 12:18 AM
        • 0 Attachment
          Hi Bram,
          thanks for the new version, I noticed you failed to mention the fact that
          vim -d file dir now works
          if the second argument is a directory with a file of the same name as the
          first.
          Also there is an error in the documentation (or in the code) for
          :help sub-replace-special
          <CR> split line in two at this point
          \r idem
          \<CR> insert a carriage-return (CTRL-M)
          \n insert a <NL> (<NUL> in the file) (does NOT break the
          line)


          the <cr> and \<cr> do not work (they insert "<cr>") although \n and \r do.
          Also it is not clear why
          in a match \n matches the end of line and not the character inserted when
          substituting something for a \n.
          Finally I don't think that \r in a search pattern works correctly, I tried
          and couldn't make it match the carriage
          return (to do this I switched a fileformat to dos and tried to have it
          match, I also tried to match the result of a
          substitution by the \r character).
          Benoit
          PS:
          I don't know if it is best to correct the documentation or the code...

          ----- Original Message -----
          From: "Bram Moolenaar" <Bram@...>
          To: <vim-dev@...>
          Sent: Monday, August 27, 2001 9:57 PM
          Subject: Vim version 6.0at BETA available


          >
          > This is the fifth BETA for Vim 6.0. Again a number of problems were
          fixed.
          >
          > All the pending items have been done. It's now really bug fixing only!
          > I expect the 6.0 release in just a few weeks now. Please try to find
          every
          > bug, documentation mistake, omission, etc.
          >
          > I noticed that some people didn't report problems, assuming someone else
          would
          > do that. Well, if you still see a problem now, apparently nobody reported
          it
          > yet, so you better do that _now_!
          >
        • Bohdan Vlasyuk
          ... I still can t reliably reproduce it, but I have random crashes, even with 60as, the problem still seem to exist. I don t know it it would help, but it
          Message 4 of 17 , Aug 28 1:14 AM
          • 0 Attachment
            On Mon, Aug 27, 2001 at 09:57:51PM +0200, Bram Moolenaar wrote:

            > I noticed that some people didn't report problems, assuming someone
            > else would do that. Well, if you still see a problem now,
            > apparently nobody reported it yet, so you better do that _now_!
            I still can't reliably reproduce it, but I have random crashes, even
            with 60as, the problem still seem to exist. I don't know it it would
            help, but it apparently appears after long time working, when >50
            files are edited at once (with :set hidden), after simple actions,
            like typing first character in '/' search prompt, etc...

            I don't think this would help, but, at least, I've tried to.

            --
            "It is better to have tried and failed than to have failed to try, but
            the result's the same."
            - Mike Dennison
          • Martin Dalecki
            Here are the following fixes for the Xt/Motif based GUI variants: 1. guiupdate. This is fixing some apparently lost exposure areas on in esp. remote
            Message 5 of 17 , Aug 28 1:28 AM
            • 0 Attachment
              Here are the following fixes for the Xt/Motif based GUI variants:

              1. guiupdate. This is fixing some apparently lost exposure areas on in
              esp. remote
              connections.

              2. mnemonic. Adds mnemonic handling to the Motif dialogues.

              3. wid. Use the proper target of mouse wraps and similar. We habe now a
              proper window
              for the whole text area and don't need to calculate menu height and
              friends...

              4. SM. Well the name says it all. Add session callbacks on the Xt level.
              Just to be sure
              nothing end's in the void...

              5. cmk. This is fixing compilation of Solaris 8 with the bonus gnu utils
              installed.

              6. defines. Kill some rotted outdated defines and an empty functions.
              This is
              bringing the whole game a bit into the direction of the GTK code.

              That's all... The patches should all apply independently from each
              other.
            • Bram Moolenaar
              ... Sorry, forgot to add this to the list of changes. Yes, you can now do: vimdiff file dir [dir [dir]] ... The here is a real CR, not four characters.
              Message 6 of 17 , Aug 28 3:48 AM
              • 0 Attachment
                Benoit Cerrina wrote:

                > thanks for the new version, I noticed you failed to mention the fact that
                > vim -d file dir now works if the second argument is a directory with
                > a file of the same name as the first.

                Sorry, forgot to add this to the list of changes. Yes, you can now do:

                vimdiff file dir [dir [dir]]

                > Also there is an error in the documentation (or in the code) for
                > :help sub-replace-special
                > <CR> split line in two at this point
                > \r idem
                > \<CR> insert a carriage-return (CTRL-M)
                > \n insert a <NL> (<NUL> in the file) (does NOT break the
                > line)
                >
                >
                > the <cr> and \<cr> do not work (they insert "<cr>") although \n and \r do.

                The <CR> here is a real CR, not four characters. It's typed as CTRL-V
                CR.

                > Also it is not clear why in a match \n matches the end of line and not
                > the character inserted when substituting something for a \n.

                I don't understand this one. Don't forget that using a \n in a search
                pattern is different from using \n in the substitute part.

                > Finally I don't think that \r in a search pattern works correctly, I tried
                > and couldn't make it match the carriage return (to do this I switched
                > a fileformat to dos and tried to have it match, I also tried to match
                > the result of a substitution by the \r character).

                For me "/\r" finds a <CR> in the text. I would not know why this does
                not work for you.

                --
                A computer program does what you tell it to do, not what you want it to do.

                /// Bram Moolenaar -- Bram@... -- http://www.moolenaar.net \\\
                ((( Creator of Vim -- http://vim.sf.net -- ftp://ftp.vim.org/pub/vim )))
                \\\ Help me helping AIDS orphans in Uganda - http://iccf-holland.org ///
              • Bram Moolenaar
                ... I m using the GTK version sometimes, but never noticed this problem. The menu code wasn t changed recently. Are you sure your GTK libraries are installed
                Message 7 of 17 , Aug 28 3:48 AM
                • 0 Attachment
                  Preben Peppe Guldberg wrote:

                  > Just compiled vim 6.0at with gtk interface on a Solaris 8 box.
                  >
                  > When started it dumps core with the following message:
                  >
                  > Vim: Caught deadly signal BUS
                  > Vim: Finished.

                  I'm using the GTK version sometimes, but never noticed this problem.
                  The menu code wasn't changed recently. Are you sure your GTK libraries
                  are installed correctly?

                  > Tracked it down to menu.vim, where any of the :amenu (or :am) commands
                  > would trigger this. This only happens if the right number of arguments
                  > is provided:
                  >
                  > [put :finish on first line of menu.vim]
                  >
                  > $ gvim60at -u NONE -U NONE --noplugin
                  >
                  > :amenu
                  > --- Menus ---
                  > Hit ENTER or type command to continue
                  > :amenu 9999.10
                  > E329: No menu of that name
                  > :amenu 9999.10 &Help.&Overview<Tab><F1>
                  > E329: No menu of that name
                  > :amenu 9999.10 &Help.&Overview<Tab><F1> :help<CR>
                  >
                  > And the message above can be seen in the terminal.

                  Perhaps you can run Vim in a debugger and find out where it crashes?

                  --
                  hundred-and-one symptoms of being an internet addict:
                  192. Your boss asks you to "go fer" coffee and you come up with 235 FTP sites.

                  /// Bram Moolenaar -- Bram@... -- http://www.moolenaar.net \\\
                  ((( Creator of Vim -- http://vim.sf.net -- ftp://ftp.vim.org/pub/vim )))
                  \\\ Help me helping AIDS orphans in Uganda - http://iccf-holland.org ///
                • Bram Moolenaar
                  ... This report doesn t help, because I can t guess what goes wrong. What you can do is compile Vim for debugging and run it from a debugger. This normally
                  Message 8 of 17 , Aug 28 3:48 AM
                  • 0 Attachment
                    Bohdan Vlasyuk wrote:

                    > On Mon, Aug 27, 2001 at 09:57:51PM +0200, Bram Moolenaar wrote:
                    >
                    > > I noticed that some people didn't report problems, assuming someone
                    > > else would do that. Well, if you still see a problem now,
                    > > apparently nobody reported it yet, so you better do that _now_!
                    > I still can't reliably reproduce it, but I have random crashes, even
                    > with 60as, the problem still seem to exist. I don't know it it would
                    > help, but it apparently appears after long time working, when >50
                    > files are edited at once (with :set hidden), after simple actions,
                    > like typing first character in '/' search prompt, etc...
                    >
                    > I don't think this would help, but, at least, I've tried to.

                    This report doesn't help, because I can't guess what goes wrong.

                    What you can do is compile Vim for debugging and run it from a debugger.
                    This normally doesn't slow down Vim, only the startup is a bit clumsy.
                    Then when it crashes you should be able to see where it happens.

                    Also, when you compiled Vim for debugging, the core file should be
                    helpful in finding out what happened.

                    --
                    hundred-and-one symptoms of being an internet addict:
                    199. You read this entire list of symptoms, looking for something
                    that doesn't describe you.

                    /// Bram Moolenaar -- Bram@... -- http://www.moolenaar.net \\\
                    ((( Creator of Vim -- http://vim.sf.net -- ftp://ftp.vim.org/pub/vim )))
                    \\\ Help me helping AIDS orphans in Uganda - http://iccf-holland.org ///
                  • Benoit Cerrina
                    ... Thanks, it works but I believe it should be made more explicit. As for r in the search string, it does find the inserted in the middle of the line
                    Message 9 of 17 , Aug 28 4:25 AM
                    • 0 Attachment
                      > > :help sub-replace-special
                      > > <CR> split line in two at this point
                      > > \r idem
                      > > \<CR> insert a carriage-return (CTRL-M)
                      > > \n insert a <NL> (<NUL> in the file) (does NOT break the
                      > > line)
                      > >
                      > >
                      > > the <cr> and \<cr> do not work (they insert "<cr>") although \n and \r
                      >do.
                      >
                      >The <CR> here is a real CR, not four characters. It's typed as CTRL-V
                      >CR.
                      >
                      Thanks, it works but I believe it should be made more explicit.
                      As for \r in the search string, it does find the <cr> inserted in the middle
                      of the line but not those at the end of the line when in dos mode.
                      Also I found the fact that \r and \n do not have the same meaning in the
                      search and substitute part of a substitute troubling enough that it may be
                      worth enphasizing it in the help.
                      Of course now I know how it works so I don't need it that much anymore but
                      I'm sure others are troubled by it.
                      Benoit

                      _________________________________________________________________
                      Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp
                    • Bram Moolenaar
                      ... Since we are very close to a release, I don t include changes without understanding what they do. ... I have not seen a problem with this. Why would it be
                      Message 10 of 17 , Aug 28 8:09 AM
                      • 0 Attachment
                        Martin Dalecki wrote:

                        > Here are the following fixes for the Xt/Motif based GUI variants:

                        Since we are very close to a release, I don't include changes without
                        understanding what they do.

                        > 1. guiupdate. This is fixing some apparently lost exposure areas on in
                        > esp. remote connections.

                        I have not seen a problem with this. Why would it be needed?

                        > 2. mnemonic. Adds mnemonic handling to the Motif dialogues.

                        I tried including it. I can see the letter is now underlined, but it
                        doesn't work. This is with Lesstif, if that matters. I think I tried
                        this before and it didn't work. Any additional code required? If it
                        works for you, perhaps this should be in an #ifdef for Lesstif?

                        > 3. wid. Use the proper target of mouse wraps and similar. We habe now a
                        > proper window for the whole text area and don't need to calculate menu
                        > height and friends...

                        Apparently this fixes the problem that the toolbar height wasn't taken
                        care of. That's a bug fix, I'll include this.

                        > 4. SM. Well the name says it all. Add session callbacks on the Xt level.
                        > Just to be sure nothing end's in the void...

                        This changes a few things without comments why this is needed. For
                        example, what is the effect of this:

                        > - applicationShellWidgetClass, gui.dpy, NULL);
                        > + sessionShellWidgetClass, gui.dpy, NULL);

                        Can you explain what the effect of adding this patch is?

                        > 5. cmk. This is fixing compilation of Solaris 8 with the bonus gnu utils
                        > installed.

                        This removes @SET_MAKE@ from config.mk.in. I'm quite sure it was there
                        for a reason. This is a standard autoconf thing. Perhaps the problem
                        is with your Solaris configuration? Many people struggle with GNU tools
                        on Solaris, thus that would be my first guess.

                        > 6. defines. Kill some rotted outdated defines and an empty functions.
                        > This is bringing the whole game a bit into the direction of the GTK
                        > code.

                        I didn't see this one attached.

                        --
                        hundred-and-one symptoms of being an internet addict:
                        205. You're constantly yelling at your spouse, family, roommate, whatever,
                        for using the phone for stupid things...like talking.

                        /// Bram Moolenaar -- Bram@... -- http://www.moolenaar.net \\\
                        ((( Creator of Vim -- http://vim.sf.net -- ftp://ftp.vim.org/pub/vim )))
                        \\\ Help me helping AIDS orphans in Uganda - http://iccf-holland.org ///
                      • Bram Moolenaar
                        ... When the fileformat is dos , then the CR-LF is the line separator and no CR appears at the end of the line. When you edit the same file in unix format
                        Message 11 of 17 , Aug 28 10:40 AM
                        • 0 Attachment
                          Benoit Cerrina wrote:

                          > > > :help sub-replace-special
                          > > > <CR> split line in two at this point
                          > > > \r idem
                          > > > \<CR> insert a carriage-return (CTRL-M)
                          > > > \n insert a <NL> (<NUL> in the file) (does NOT break the
                          > > > line)
                          > > >
                          > > >
                          > > > the <cr> and \<cr> do not work (they insert "<cr>") although \n
                          > > > and \r do.
                          > >
                          > >The <CR> here is a real CR, not four characters. It's typed as CTRL-V
                          > >CR.
                          > >
                          > Thanks, it works but I believe it should be made more explicit.
                          > As for \r in the search string, it does find the <cr> inserted in the middle
                          > of the line but not those at the end of the line when in dos mode.

                          When the fileformat is "dos", then the CR-LF is the line separator and
                          no CR appears at the end of the line. When you edit the same file in
                          "unix" format then you will see the CR and a search pattern will match it.

                          > Also I found the fact that \r and \n do not have the same meaning in the
                          > search and substitute part of a substitute troubling enough that it may be
                          > worth enphasizing it in the help.
                          > Of course now I know how it works so I don't need it that much anymore but
                          > I'm sure others are troubled by it.

                          It is explained at ":help sub-replace-special". Unfortunately, it can't
                          be changed, it would not be backwards compatible.

                          --
                          "You mean there really is an answer?"
                          "Yes! But you're not going to like it!"
                          "Oh do please tell us!"
                          "You're really not going to like it!"
                          "but we MUST know - tell us"
                          "Alright, the answer is...."
                          "yes..."
                          "... is ..."
                          "yes... come on!"
                          "is 42!"
                          (Douglas Adams - The Hitchhiker's Guide to the Galaxy)

                          /// Bram Moolenaar -- Bram@... -- http://www.moolenaar.net \\\
                          ((( Creator of Vim -- http://vim.sf.net -- ftp://ftp.vim.org/pub/vim )))
                          \\\ Help me helping AIDS orphans in Uganda - http://iccf-holland.org ///
                        • Martin Dalecki
                          ... After checking this again I see that the actual acceleartor buttons don t work for me properly as well. I will have to dig into the the popup menu handling
                          Message 12 of 17 , Aug 29 12:50 AM
                          • 0 Attachment
                            Bram Moolenaar wrote:
                            >
                            > Martin Dalecki wrote:
                            >
                            > > Here are the following fixes for the Xt/Motif based GUI variants:
                            >
                            > Since we are very close to a release, I don't include changes without
                            > understanding what they do.
                            >
                            > > 1. guiupdate. This is fixing some apparently lost exposure areas on in
                            > > esp. remote connections.
                            >
                            > I have not seen a problem with this. Why would it be needed?
                            >
                            > > 2. mnemonic. Adds mnemonic handling to the Motif dialogues.
                            >
                            > I tried including it. I can see the letter is now underlined, but it
                            > doesn't work. This is with Lesstif, if that matters. I think I tried
                            > this before and it didn't work. Any additional code required? If it
                            > works for you, perhaps this should be in an #ifdef for Lesstif?

                            After checking this again I see that the actual acceleartor buttons
                            don't work
                            for me properly as well. I will have to dig into the the popup menu
                            handling
                            code of Motif to see why this is the case.

                            > > 3. wid. Use the proper target of mouse wraps and similar. We habe now a
                            > > proper window for the whole text area and don't need to calculate menu
                            > > height and friends...
                            >
                            > Apparently this fixes the problem that the toolbar height wasn't taken
                            > care of. That's a bug fix, I'll include this.

                            grinn... Now you can see why I have introduced the separate window for
                            the
                            text in first place ... grinn.... I see as well that the basesize for
                            the window manager isn't calculated properly as well. You can perhaps
                            rememeber how long we have struggled with this in GTK - I think I know
                            now
                            what's going on.

                            > > 4. SM. Well the name says it all. Add session callbacks on the Xt level.
                            > > Just to be sure nothing end's in the void...
                            >
                            > This changes a few things without comments why this is needed. For
                            > example, what is the effect of this:
                            >
                            > > - applicationShellWidgetClass, gui.dpy, NULL);
                            > > + sessionShellWidgetClass, gui.dpy, NULL);
                            >
                            > Can you explain what the effect of adding this patch is?

                            It's supposed to provide proper interaction with xsm, by doing it on the
                            Xt level
                            instead of one level deeper. However agreed this is a bit experimental.

                            > > 5. cmk. This is fixing compilation of Solaris 8 with the bonus gnu utils
                            > > installed.
                            >
                            > This removes @SET_MAKE@ from config.mk.in. I'm quite sure it was there
                            > for a reason. This is a standard autoconf thing. Perhaps the problem
                            > is with your Solaris configuration? Many people struggle with GNU tools
                            > on Solaris, thus that would be my first guess.

                            For sure not. And for sure this isn't proper autoconf tactics. Without
                            this
                            the problem will be that SET_MAKE will pick make instead of gmake on
                            Solaris.
                            config.mk should just inherit the make from the includer.

                            > > 6. defines. Kill some rotted outdated defines and an empty functions.
                            > > This is bringing the whole game a bit into the direction of the GTK
                            > > code.
                            >
                            > I didn't see this one attached.

                            Oops.. there it is.
                          • Bram Moolenaar
                            ... No remark for this one? ... Please do. I would like the accelerators to work. But be careful: When there is a textfield typed text might have to go
                            Message 13 of 17 , Aug 29 5:20 AM
                            • 0 Attachment
                              Martin Dalecki wrote:

                              > > > 1. guiupdate. This is fixing some apparently lost exposure areas on in
                              > > > esp. remote connections.
                              > >
                              > > I have not seen a problem with this. Why would it be needed?

                              No remark for this one?

                              > > > 2. mnemonic. Adds mnemonic handling to the Motif dialogues.
                              > >
                              > > I tried including it. I can see the letter is now underlined, but it
                              > > doesn't work. This is with Lesstif, if that matters. I think I tried
                              > > this before and it didn't work. Any additional code required? If it
                              > > works for you, perhaps this should be in an #ifdef for Lesstif?
                              >
                              > After checking this again I see that the actual acceleartor buttons
                              > don't work for me properly as well. I will have to dig into the the
                              > popup menu handling code of Motif to see why this is the case.

                              Please do. I would like the accelerators to work. But be careful: When
                              there is a textfield typed text might have to go there, don't want the
                              "o" triggering the "Ok" button then!

                              > > > 4. SM. Well the name says it all. Add session callbacks on the Xt level.
                              > > > Just to be sure nothing end's in the void...
                              > >
                              > > This changes a few things without comments why this is needed. For
                              > > example, what is the effect of this:
                              > >
                              > > > - applicationShellWidgetClass, gui.dpy, NULL);
                              > > > + sessionShellWidgetClass, gui.dpy, NULL);
                              > >
                              > > Can you explain what the effect of adding this patch is?
                              >
                              > It's supposed to provide proper interaction with xsm, by doing it on the
                              > Xt level instead of one level deeper. However agreed this is a bit
                              > experimental.

                              Since I have not noticed problems in this area, I rather not change
                              this right now.

                              > > > 5. cmk. This is fixing compilation of Solaris 8 with the bonus gnu utils
                              > > > installed.
                              > >
                              > > This removes @SET_MAKE@ from config.mk.in. I'm quite sure it was there
                              > > for a reason. This is a standard autoconf thing. Perhaps the problem
                              > > is with your Solaris configuration? Many people struggle with GNU tools
                              > > on Solaris, thus that would be my first guess.
                              >
                              > For sure not. And for sure this isn't proper autoconf tactics. Without
                              > this the problem will be that SET_MAKE will pick make instead of gmake
                              > on Solaris. config.mk should just inherit the make from the includer.

                              Vim uses $(MAKE), thus it must be set. Some versions of make don't set
                              $(MAKE), that's what this autoconf check is needed for. When I remove
                              it, it will break on systems with ordinary make.

                              Your problem is probably that you have an old "make" on the system,
                              which configure finds, but then you run "gmake" for the Makefile.
                              Configure can't know you will use "gmake". The solution is not to run
                              "./configure", but use "gmake" to run configure. Then $MAKE will be set
                              and configure will do the right thing. Or set $MAKE before you run
                              configure.

                              > > > 6. defines. Kill some rotted outdated defines and an empty functions.
                              > > > This is bringing the whole game a bit into the direction of the GTK
                              > > > code.
                              > >
                              > > I didn't see this one attached.
                              >
                              > Oops.. there it is.

                              OK, just some cleaning up, I'll include it.

                              --
                              The term "free software" is defined by Richard M. Stallman as
                              being software that isn't necessarily for free. Confusing?
                              Let's call it "Stallman software" then!
                              -- Bram Moolenaar

                              /// Bram Moolenaar -- Bram@... -- http://www.moolenaar.net \\\
                              ((( Creator of Vim -- http://vim.sf.net -- ftp://ftp.vim.org/pub/vim )))
                              \\\ Help me helping AIDS orphans in Uganda - http://iccf-holland.org ///
                            • Martin Dalecki
                              ... Well OK. The problem over a slow link is that Xt will get more chances to compress or time out X11 events. However the problem is just there no matter
                              Message 14 of 17 , Aug 29 5:52 AM
                              • 0 Attachment
                                Bram Moolenaar wrote:
                                >
                                > Martin Dalecki wrote:
                                >
                                > > > > 1. guiupdate. This is fixing some apparently lost exposure areas on in
                                > > > > esp. remote connections.
                                > > >
                                > > > I have not seen a problem with this. Why would it be needed?
                                >
                                > No remark for this one?

                                Well OK. The problem over a slow link is that Xt will get more chances
                                to compress or time out X11 events. However the problem is just there no
                                matter
                                whatever the connection is remote or not. It manifests as follows, quite
                                sponatneously:
                                Start vim -f -g.
                                Open a file with *complex* (read: slow) syntax highlighting.
                                User for example the find dialogue. Progressive search will lead
                                to scrolling text under the dialogue. And *then* refresh problems
                                will occure in the scrolling text. The gui_mch_update stuff, is just
                                making sure, that the exposure events will be handled immediately
                                before the next search can happen.
                                Hmm... I know this is a bit difficult to capture, but I never
                                noticed this happen again after I did the change above.

                                I how this explains it a bit?

                                > > > > 5. cmk. This is fixing compilation of Solaris 8 with the bonus gnu utils
                                > > > > installed.
                                > > >
                                > > > This removes @SET_MAKE@ from config.mk.in. I'm quite sure it was there
                                > > > for a reason. This is a standard autoconf thing. Perhaps the problem
                                > > > is with your Solaris configuration? Many people struggle with GNU tools
                                > > > on Solaris, thus that would be my first guess.
                                > >
                                > > For sure not. And for sure this isn't proper autoconf tactics. Without
                                > > this the problem will be that SET_MAKE will pick make instead of gmake
                                > > on Solaris. config.mk should just inherit the make from the includer.
                                >
                                > Vim uses $(MAKE), thus it must be set. Some versions of make don't set
                                > $(MAKE), that's what this autoconf check is needed for. When I remove
                                > it, it will break on systems with ordinary make.
                                >
                                > Your problem is probably that you have an old "make" on the system,
                                > which configure finds, but then you run "gmake" for the Makefile.
                                > Configure can't know you will use "gmake". The solution is not to run
                                > "./configure", but use "gmake" to run configure. Then $MAKE will be set
                                > and configure will do the right thing. Or set $MAKE before you run
                                > configure.

                                Thak you for explaining it to me. The problem is in first place
                                the fact that configure doesn't recognize, that the default
                                /usr/ucs/bin/make
                                is just a stiupid shell stub nagging you to buy something which should
                                be already there in first place if you buy a UNIX operating system...
                                They called this idiotism "unbundling" at Sun.

                                However the workaround you mention is the most elegant way to deal with
                                this
                                problem.
                              • Bram Moolenaar
                                ... OK, I now understand why this would be needed. I ll include it and add a remark what the call to gui_mch_update() is for. ... I suppose autoconf should
                                Message 15 of 17 , Aug 29 8:23 AM
                                • 0 Attachment
                                  Martin Dalecki wrote:

                                  > > > > > 1. guiupdate. This is fixing some apparently lost exposure areas on in
                                  > > > > > esp. remote connections.
                                  > > > >
                                  > > > > I have not seen a problem with this. Why would it be needed?
                                  > >
                                  > > No remark for this one?
                                  >
                                  > Well OK. The problem over a slow link is that Xt will get more chances
                                  > to compress or time out X11 events. However the problem is just there no
                                  > matter whatever the connection is remote or not. It manifests as
                                  > follows, quite sponatneously:
                                  > Start vim -f -g.
                                  > Open a file with *complex* (read: slow) syntax highlighting.
                                  > User for example the find dialogue. Progressive search will lead
                                  > to scrolling text under the dialogue. And *then* refresh problems
                                  > will occure in the scrolling text. The gui_mch_update stuff, is just
                                  > making sure, that the exposure events will be handled immediately
                                  > before the next search can happen.
                                  > Hmm... I know this is a bit difficult to capture, but I never
                                  > noticed this happen again after I did the change above.
                                  >
                                  > I how this explains it a bit?

                                  OK, I now understand why this would be needed. I'll include it and add
                                  a remark what the call to gui_mch_update() is for.

                                  > Thak you for explaining it to me. The problem is in first place
                                  > the fact that configure doesn't recognize, that the default
                                  > /usr/ucs/bin/make
                                  > is just a stiupid shell stub nagging you to buy something which should
                                  > be already there in first place if you buy a UNIX operating system...
                                  > They called this idiotism "unbundling" at Sun.
                                  >
                                  > However the workaround you mention is the most elegant way to deal
                                  > with this problem.

                                  I suppose autoconf should check for the stupid Sun make program. But it
                                  doesn't

                                  I now notice that autoconf 2.52 is out. But I'll stick to 2.13 now to
                                  avoid risking problems. Don't know if it fixes this $(MAKE) problem
                                  anyway...

                                  --
                                  hundred-and-one symptoms of being an internet addict:
                                  252. You vote for foreign officials.

                                  /// Bram Moolenaar -- Bram@... -- http://www.moolenaar.net \\\
                                  ((( Creator of Vim -- http://vim.sf.net -- ftp://ftp.vim.org/pub/vim )))
                                  \\\ Help me helping AIDS orphans in Uganda - http://iccf-holland.org ///
                                • Martin Dalecki
                                  ... Just a note. Please don t run after autoconf. It would make it even more difficult for others to have a complete vim developement tool chain ;-). This
                                  Message 16 of 17 , Aug 29 11:47 PM
                                  • 0 Attachment
                                    Bram Moolenaar wrote:

                                    > I suppose autoconf should check for the stupid Sun make program. But it
                                    > doesn't
                                    >
                                    > I now notice that autoconf 2.52 is out. But I'll stick to 2.13 now to
                                    > avoid risking problems. Don't know if it fixes this $(MAKE) problem
                                    > anyway...

                                    Just a note. Please don't run after autoconf. It would make it even
                                    more difficult for others to have a complete "vim developement tool
                                    chain" ;-).
                                    This case isn't that important...
                                  • Walter Briscoe
                                    In article of Thu, 30 Aug 2001 08:47:54 in !vim-dev, Martin Dalecki writes ... Bram
                                    Message 17 of 17 , Aug 30 1:30 AM
                                    • 0 Attachment
                                      In article <3B8DE19A.598B9AFF@...> of Thu, 30 Aug 2001
                                      08:47:54 in !vim-dev, Martin Dalecki <dalecki@...>
                                      writes
                                      >Bram Moolenaar wrote:
                                      >
                                      >> I suppose autoconf should check for the stupid Sun make program. But it
                                      >> doesn't
                                      >>
                                      >> I now notice that autoconf 2.52 is out. But I'll stick to 2.13 now to
                                      >> avoid risking problems. Don't know if it fixes this $(MAKE) problem
                                      >> anyway...
                                      >
                                      >Just a note. Please don't run after autoconf. It would make it even
                                      >more difficult for others to have a complete "vim developement tool
                                      >chain" ;-).
                                      >This case isn't that important...

                                      Bram said he would not change and I can imagine no argument.
                                      Can I suggest that this should be done as part of the next Vim ALPHA
                                      project.
                                      --
                                      Walter Briscoe
                                    Your message has been successfully submitted and would be delivered to recipients shortly.