mlterm-dev-en — mlterm mailing list(English)

Re: [Mlterm-dev-en] commit log [2002/02/11] (mlterm server)

[Araki K. <j00...@ip...>]

	Hi,

	Subject: [Mlterm-dev-en] commit log [2002/02/11] (mlterm server)
	From: Araki Ken <j00...@ip...>
	Message-ID: <200...@pd...>
	Date: Mon, 11 Feb 2002 06:44:17 +0900

>        For example , after you start mlterm -t=false -j blend , execute
>        mlclient with -t=true option , and a new terminal screen , whose background
>        is certainly transparent , will appear.

        Just to be sure , these two terminals work in one process.
	
--
kiken
j00...@ip...

[Mlterm-dev-en] commit log [2002/02/11] (mlterm server)

[Araki K. <j00...@ip...>]

       Hi,

       I committed changes below.
       
* daemon mode is supported and -j/--daemon/daemon_mode option is added.
* mlclient program (tool/mlclient) is added.
* characters in the beginning of lines cannot be selected. fixed. (Thanks to nekoie san)
* 2 dot margin area in both right and left sides are cleared since garbages
  can be left there.

       Now mlterm can become a daemon process.

       "-j genuine" or "-j blend" option gets mlterm to become a daemon process.
       If blend is specified , mlterm process exits as soon as the very last visible
       terminal is destroyed , while genuine option lets it last forever.
       
       mlclient (tool/mlclient) , which accepts almost the same options as mlterm ,
       invokes mlterm server to open a new terminal.

       For example , after you start mlterm -t=false -j blend , execute
       mlclient with -t=true option , and a new terminal screen , whose background
       is certainly transparent , will appear.

       Please test this new feature.
       
       Best Regards.
--
kiken
j00...@ip...

[Mlterm-dev-en] commit log [2002/02/09] (utmp)

[Araki K. <j00...@ip...>]

       Hi ,

-- 2002/02/09 commit log
* README.ja is rewritten.
* BSD type utmp/wtmp/lastlog is supported. (--enable-utmp configure option)

       BSD type utmp is now supported.
       If you want to use it , do ./configure with --enable-utmp option.

       I test it on NetBSD 1.5ZA alone.
       So on other BSD systems , even compilation can fail , I think.
       Please test it and give me reports about whether it works or not.

       SYSV type utmp is not supported , that is , mlterm/kiklib/src/kik_utmp_sysv.c
       (and kik_utmp_new() , kik_utmp_delete() functions in it) is still empty.
       Because I have no SYSV environments at hand , I would like someone to
       implement it ....
--
kiken
j00...@ip...

Re: [Mlterm-dev-en] Scrollback etc in cjk-mode

[Araki K. <j00...@ip...>]

	Hi,

	Subject: [Mlterm-dev-en] Scrollback etc in cjk-mode
	From: Karl Koehler <ko...@or...>
	Message-ID: <Pine.A32.3.91.1020207190552.84649A-100000@r53aix01-tr0.or.uni=
-bonn.de>
	Date: Thu, 7 Feb 2002 20:42:33 +0100 (MEZ)

> This leads to some wishes for the wishlist:
> a) scrollback should work left-to-right, not bottom-to-top.
> b) arabic from bottom-to-top looks strange, especially as it is=20
>    not joined. One could think of doing bidi the following way:
>    + right-to-left: rotate 90deg counterclockwise
>    + left-to-right:=A0maybe rotate 90deg clockwise

        Wow !

        I'm sorry , but it is too difficult to implement them for now.

        For example , rotation makes height of characters half , but mlterm
        assumes all characters has the same height.
        To solve such a problem , I have to modify many codes.
       =20
        After all , I adopted an easy design where such a full-spec vertical
        view (left-to-right scrollbar , arabic shaping , character rotation=
 ...)
        is not considered , since it is not difficult to maintenance and is
        nevertheless practical for CJK people.

        Of cource , I may challenge in the future (and will modify over ten
        thousand steps in the end) , but I have no plan for now.

> In attempting b), I have a question:
> In ml_font.c, the font-name is created with:
> =20
>  sprintf( fontname , "-*-%s-%s-%s-%s--%d-*-%s-*-%s" ,
>           family , weight , slant , width , fontsize , spacing , encoding=
) ;
>
> This specifies pointsize to be '*', and leaves out xres and yres values.
> Is there a reason to not do a full specification, that is:
>
>  sprintf( fontname , "-*-%s-%s-%s-%s--%d-*-*-*-%s-*-%s" ,
>           family , weight , slant , width , fontsize , spacing , encoding=
) ;
>
> ( the three stars being pointsize, xres,yres ).

        This is my mistake.
        I'll fix it , thanks:)

--
kiken
j00...@ip...

[Mlterm-dev-en] Scrollback etc in cjk-mode

[Karl K. <ko...@or...>]

Hi ,

One of the new options added is -G cjk, which adds right-to-left,
to-to bottom text.
This leads to some wishes for the wishlist:
a) scrollback should work left-to-right, not bottom-to-top.
b) arabic from bottom-to-top looks strange, especially as it is=20
   not joined. One could think of doing bidi the following way:
   + right-to-left: rotate 90deg counterclockwise
   + left-to-right:=A0maybe rotate 90deg clockwise

In attempting b), I have a question:
In ml_font.c, the font-name is created with:
=20
 sprintf( fontname , "-*-%s-%s-%s-%s--%d-*-%s-*-%s" ,
          family , weight , slant , width , fontsize , spacing , encoding) =
;

This specifies pointsize to be '*', and leaves out xres and yres values.
Is there a reason to not do a full specification, that is:

 sprintf( fontname , "-*-%s-%s-%s-%s--%d-*-*-*-%s-*-%s" ,
          family , weight , slant , width , fontsize , spacing , encoding) =
;

( the three stars being pointsize, xres,yres ).
Only then, rotation seems to work in a reliable way: 90deg becomes

 sprintf( fontname , "-*-%s-%s-%s-%s--[0 %d ~%d 0]-*-*-*-%s-*-%s" ,
=09   family , weight , slant , width , fontsize ,=20
=09   fontsize , spacing , encoding) ;

Of course, one has to alloca more memory and _not_ rotate asian glyphs .. =
=20
As a side-note, those X11R5-servers need an own solution: best they
die out ( but there still are some here ).=20


=09Karl

[Mlterm-dev-en] commit log [2002/02/06]

[Araki K. <j00...@ip...>]

        Hi,
        
* -O/--sbpos/scrollbar_position option is removed and
  -O/--sbmod/scrollbar_mode option is added.
* scrollbar can be dynamically switched among off , right and left.
* default value of use_scrollbar option is true , not false.

        Now mlterm can switch scrollbar on/off dynamically.
        As a result , some options related to scrollbar are changed a little.
        
        1. use_scrollbar
           This option is left as it is for backward compatibility ,
           but its default value is true , not false now.
           
           False of this option means scrollbar is *never* shown any more.

        2. scrollbar_mode
           This is a new option.
           Its value is "none" , "left"(default)  or "right".

           "none" = scrollbar is not shown(but can be shown later from mlconfig).
           "left" = scrollbar is shown on left side.
           "right" = scrollbar is shown on right side.

           This option is valid only when use_scrollbar option is true.
           
        3. scrollbar_position
           This option is removed.

        Those who are using scrollbar now don't have to change your current
        configurations.

        But if you don't want to use scrollbar , please remember to add explicitly

use_scrollbar = false

        or

scrollbar_mode = none

        to ~/.mlterm/main.

        Best Regards.

--
kiken
j00...@ip...

[Mlterm-dev-en] commit log [2002/02/05-2]

[Araki K. <j00...@ip...>]

        Hi,

* -O/--sbpos/scrollbar_position option is added.
* processing brightness adjustment under transparent mode is tuned.

        Now mlterm supports scrollbar on right side.
        Please specify -O=right option if you want to use it.
        Its value is right or left and the latter is the default.
        
        Don't forget to recompile at scrollbar/sample and install them , since
        scrollbar objects of current cvs repository are ABI/API compatible with
        before but behave a little differently.

--
kiken
j00...@ip...

[Mlterm-dev-en] mlterm-2.2.0pl1 patchkit released

[Araki K. <j00...@ip...>]

       Hi,

       I released 2.2.0 patchkit.

       http://prdownloads.sourceforge.net/mlterm/mlterm-2.2.0pl1.tar.gz

       It includes cygwin port patch by Tetsurou Okazaki san.

ChangeLog
* [patches #512538] cygwin port is merged.(thanks to Tetsurou Okazaki san)
* XRegisterIMInstantiateCallback is never called on sunos.
* baurates are explicitly set with cfsetispeed,cfsetospeed.(thanks to t-matsuo san)
* When CapsLock or NumLock is on , control mask or shift mask etc is not recogized.
  fixed. (Thanks to K.Kawabata san)
  
--
kiken
j00...@ip...

[Mlterm-dev-en] commit log [2002/02/05] and cygwin port (was: 2.2.0 segfault)

[Araki K. <j00...@ip...>]

	Hi,

        I committed changes below.
        
* cywin port patch is merged.(thanks to Tetsurou Okazaki san)
* XRegisterIMInstantiateCallback is never called on sunos.

        Cygwin/XFree86 port by Tetsurou Okazaki san is merged to CVS current.
        Please test it.
        
	Subject: Re: [Mlterm-dev-en] 2.2.0 segfault
	From: Noah Levitt <nl...@co...>
	Message-ID: <200...@co...>
	Date: Mon, 4 Feb 2002 12:47:24 -0500

> I've tried checking out some old versions, but all of them
> seem to dump core in the same place. I'm puzzled... I guess
> something on my system has changed.

        I'm puzzled , too:)

> I'll let you know if I figure out what it is.

       Thanks.
       
       For now I avoid this problem by the previous patch.
	
--
kiken
j00...@ip...

Re: [Mlterm-dev-en] 2.2.0 segfault

[Noah L. <nl...@co...>]

On Mon, Feb 04, 2002 at 02:27:21AM +0900, Araki Ken wrote:
>         Well , when you are free , would you please cvs checkout -D [date] and check
>         when mlterm gets insane ?

I've tried checking out some old versions, but all of them
seem to dump core in the same place. I'm puzzled... I guess
something on my system has changed. I'll let you know if I
figure out what it is.

Anyhow, it seems like you've figured out an appropriate
solution.

Noah

Re: [Mlterm-dev-en] 2.2.0 segfault

[Araki K. <j00...@ip...>]

	Hi,

	Subject: Re: [Mlterm-dev-en] 2.2.0 segfault
	From: MINAMI Hirokazu <mi...@ch...>
	Message-ID: <200...@ch...>
	Date: Mon, 4 Feb 2002 14:18:19 +0900

> FYI, gtk+1.2 seems to be using the same workaround.
>
> from config.in of gtk+ 1.2.10:
>>  # On Solaris, calling XRegisterIMInstantiateCallback seems to
>>  # cause an immediate segfault, so we disable it, unless
>>  # the user specifically forces it to be on.
>>  if test x$enable_xim_inst = xmaybe ; then
>>    case host in
>>          *-*-solaris*)
>>            enable_xim_inst="no"
>>            ;;
>>          *)
>>            enable_xim_inst="yes"
>>            ;;
>>    esac
>>  fi
>>
>>  if test "x$enable_xim" = "xyes"; then
>>    GTK_XIM_FLAGS="-DUSE_XIM"
>>    if test "x$enable_xim_inst" = "xyes"; then
>>      AC_DEFINE(USE_X11R6_XIM)
>>    fi
>>  fi
>
> # Does it mean openwin have a non-X11R6 version of XRegisterIMInstantiateCallback?

         Thanks:)
         I should have searched google with Solaris + XRegisterIMInstantiateCallback
         keywords. I could find similar problems.
         
         I wrote a patch for avoiding calling XRegisterIMIntantiateCallback on sunos.
         Please test an attached patch(for CVS current).

         Of cource , even if XRegisterIMInstatiateCallback is not called , you can
         use XIM.

--
kiken
j00...@ip...

Index: ml_xim.c
===================================================================
RCS file: /home/ken/cvsroot/mlterm/src/ml_xim.c,v
retrieving revision 1.67
diff -u -r1.67 ml_xim.c
--- ml_xim.c	2002/02/03 07:53:06	1.67
+++ ml_xim.c	2002/02/04 16:14:42
@@ -331,9 +331,15 @@
 		}
 	}
 
+	/*
+	 * XXX
+	 * XRegisterIMInstantiateCallback of sunos/openwin seems buggy.
+	 */
+#if  ! defined(sun) && ! defined(__sun__) && ! defined(__sun)
 	XRegisterIMInstantiateCallback( xim_display , NULL , NULL , NULL ,
 		xim_server_instantiated , NULL) ;
-		
+#endif
+	
 	return  1 ;
 }
 

Re: [Mlterm-dev-en] 2.2.0 segfault

[MINAMI H. <mi...@ch...>]

Hi,

On Sun, 03 Feb 2002 17:44:31 +0900
Araki Ken <j00...@ip...> wrote:

>         I did adhoc changes in CVS current version.
>         
>         If --xim=false option is specified , XRegisterIMInstantiateCallback is never
>         called.
> 
>         It is not a right solution , but I think you can start mlterm with --xim=false
>         (or use_xim=false in ~/.mlterm/main) anyway.

FYI, gtk+1.2 seems to be using the same workaround.

from config.in of gtk+ 1.2.10:
>  # On Solaris, calling XRegisterIMInstantiateCallback seems to
>  # cause an immediate segfault, so we disable it, unless
>  # the user specifically forces it to be on.
>  if test x$enable_xim_inst = xmaybe ; then
>    case host in
>          *-*-solaris*)
>            enable_xim_inst="no"
>            ;;
>          *)
>            enable_xim_inst="yes"
>            ;;
>    esac
>  fi
>
>  if test "x$enable_xim" = "xyes"; then
>    GTK_XIM_FLAGS="-DUSE_XIM"
>    if test "x$enable_xim_inst" = "xyes"; then
>      AC_DEFINE(USE_X11R6_XIM)
>    fi
>  fi

# Does it mean openwin have a non-X11R6 version of XRegisterIMInstantiateCallback?
--
MINAMI Hirokazu <mi...@ch...>

Re: [Mlterm-dev-en] 2.2.0 segfault

[Araki K. <j00...@ip...>]

	Hi,

	Subject: Re: [Mlterm-dev-en] 2.2.0 segfault
	From: Noah Levitt <nl...@co...>
	Message-ID: <200...@co...>
	Date: Sun, 3 Feb 2002 11:03:38 -0500

> extern Bool XRegisterIMInstantiateCallback(
> #if NeedFunctionPrototypes
>     Display*                    /* dpy */,
>     struct _XrmHashBucketRec*   /* rdb */,
>     char*                       /* res_name */,
>     char*                       /* res_class */,
>     XIMProc                     /* callback */,
>     XPointer*                   /* client_data */
> #endif
> );
>
> Here's XIMProc:
>
> typedef void (*XIMProc)(
> #ifdef NeedFunctionPrototypes
>     XIC, XPointer, XPointer
> #endif
> );

        The type of the first argument of callback function should not be XIC ,
        since XRegisterIMInstantiateCallback is called before the connection with XIM
        server is established , that is , before XIC is not created.

        But since XIC typedef is as follows in my Xlib.h , the size of XIC and that of
        Display * must be the same.

typedef struct _XIC *XIC;

        I think these warning messages are not the cause of this core dump.

        But I really have no idea of this problem , sorry.
        Well , when you are free , would you please cvs checkout -D [date] and check
        when mlterm gets insane ?
--
kiken
j00...@ip...

Re: [Mlterm-dev-en] 2.2.0 segfault

[Noah L. <nl...@co...>]

Hello,

On Sun, Feb 03, 2002 at 04:32:37PM +0900, Araki Ken wrote:
> 
> ---- Xlib.h
> typedef void (*XIDProc)(
> #if NeedFunctionPrototypes
>     Display*,
>     XPointer,
>     XPointer
> #endif
> );

My system's Xlib.h has no XIDProc.

> 
> ...
> 
> extern Bool XRegisterIMInstantiateCallback(
> #if NeedFunctionPrototypes
>     Display*                    /* dpy */,
>     struct _XrmHashBucketRec*   /* rdb */,
>     char*                       /* res_name */,
>     char*                       /* res_class */,
>     XIDProc                     /* callback */,
>     XPointer                    /* client_data */
> #endif
> );
> ---- Xlib.h

extern Bool XRegisterIMInstantiateCallback(
#if NeedFunctionPrototypes
    Display*                    /* dpy */,
    struct _XrmHashBucketRec*   /* rdb */,
    char*                       /* res_name */,
    char*                       /* res_class */,
    XIMProc                     /* callback */,
    XPointer*                   /* client_data */
#endif
);

Here's XIMProc:

typedef void (*XIMProc)(
#ifdef NeedFunctionPrototypes
    XIC, XPointer, XPointer
#endif
);

This is X11R6.1 I believe. (Sorry I have such an old
version!)

Noah

Re: [Mlterm-dev-en] 2.2.0 segfault

[Araki K. <j00...@ip...>]

	Hi,

	Subject: Re: [Mlterm-dev-en] 2.2.0 segfault
	From: Noah Levitt <nl...@co...>
	Message-ID: <200...@co...>
	Date: Sat, 2 Feb 2002 19:28:09 -0500

>>         And if you use an attached patch , mlterm starts normally ?
>
> Yes, it does.

        I did adhoc changes in CVS current version.
        
        If --xim=false option is specified , XRegisterIMInstantiateCallback is never
        called.

        It is not a right solution , but I think you can start mlterm with --xim=false
        (or use_xim=false in ~/.mlterm/main) anyway.
	
--
kiken
j00...@ip...

Re: [Mlterm-dev-en] 2.2.0 segfault

[Araki K. <j00...@ip...>]

	Hi,

	Subject: Re: [Mlterm-dev-en] 2.2.0 segfault
	From: Noah Levitt <nl...@co...>
	Message-ID: <200...@co...>
	Date: Sat, 2 Feb 2002 19:28:09 -0500

>>         Would you show me your XMODIFIERS and LC_CTYPE environmental variables ?
>
> XMODIFIERS is unset. LC_CTYPE is normally "en_US.UTF-8", but I
> get the same result even when it's "C".

        I tried with LC_CTYPE=en_US.UTF-8 and XMODIFIERS not set
        on NetBSD 1.5ZA / XFree86 4.1.0 , but I couldn't see such a
        problem , though I think these variables are all that
        are related to XIM.

> Until recently I was using a version I had checked out of
> cvs (maybe 3 weeks ago), and it worked.

        Hmmm , I hardly touched ml_xim.c ml_xic.c from 2001/12/25 , especially
        around XRegisterIMInstantiateCallback.

        I wonder where I did something wrong....

> I noticed that I get these warnings when I compile (checked
> out from cvs Sat Feb  2 19:17:25 EST 2002, unpatched):
>
> ml_xim.c: In function `xim_server_destroyed':
> ml_xim.c:96: warning: passing arg 5 of `XRegisterIMInstantiateCallback' from inc
> ompatible pointer type
> ml_xim.c: In function `open_xim':
> ml_xim.c:142: warning: initialization from incompatible pointer type
> ml_xim.c: In function `ml_xim_init':
> ml_xim.c:328: warning: passing arg 5 of `XRegisterIMInstantiateCallback' from in
> compatible pointer type
> ml_xim.c: In function `ml_xim_final': 
> ml_xim.c:349: warning: passing arg 5 of `XUnregisterIMInstantiateCallback' from incompatible pointer type

        Thanks.
        In my /usr/X11R6/include/Xlib.h XRegisterIMInstantiateCallback is defined
        as follows and I get no warning messages.

---- Xlib.h
typedef void (*XIDProc)(
#if NeedFunctionPrototypes
    Display*,
    XPointer,
    XPointer
#endif
);

...

extern Bool XRegisterIMInstantiateCallback(
#if NeedFunctionPrototypes
    Display*                    /* dpy */,
    struct _XrmHashBucketRec*   /* rdb */,
    char*                       /* res_name */,
    char*                       /* res_class */,
    XIDProc                     /* callback */,
    XPointer                    /* client_data */
#endif
);
---- Xlib.h

        Prototype of mlterm's callback function(XIDProc) is like this.

---- ml_xim.c
static void
xim_server_instantiated(
	Display *  display ,
	XPointer  client_data ,
	XPointer  call_data
	)
---- ml_xim.c

        XRegisterIMInstantiateCallback and XIDProc prototypes of your system cannot
        be different from mine , but I have no idea why such warning messages
        were output.

        To make sure , would you show me their prototypes in your Xlib.h ?
        
--
kiken
j00...@ip...

Re: [Mlterm-dev-en] 2.2.0 segfault

[Noah L. <nl...@co...>]

Hello,

On Sat, Feb 02, 2002 at 09:03:37PM +0900, Araki Ken wrote:
>         
>         Would you show me your XMODIFIERS and LC_CTYPE environmental variables ?

XMODIFIERS is unset. LC_CTYPE is normally "en_US.UTF-8", but I
get the same result even when it's "C".

>         And if you use an attached patch , mlterm starts normally ?

Yes, it does.

>         
>         # Perhaps all the versions of mlterm have the same problems ?

Until recently I was using a version I had checked out of
cvs (maybe 3 weeks ago), and it worked.

I noticed that I get these warnings when I compile (checked
out from cvs Sat Feb  2 19:17:25 EST 2002, unpatched):

ml_xim.c: In function `xim_server_destroyed':
ml_xim.c:96: warning: passing arg 5 of `XRegisterIMInstantiateCallback' from inc
ompatible pointer type
ml_xim.c: In function `open_xim':
ml_xim.c:142: warning: initialization from incompatible pointer type
ml_xim.c: In function `ml_xim_init':
ml_xim.c:328: warning: passing arg 5 of `XRegisterIMInstantiateCallback' from in
compatible pointer type
ml_xim.c: In function `ml_xim_final': 
ml_xim.c:349: warning: passing arg 5 of `XUnregisterIMInstantiateCallback' from incompatible pointer type

Noah

[Mlterm-dev-en] commit log [2002/02/03]

[Araki K. <j00...@ip...>]

	Hi,

        There are mainly three new features.
        
        o -H/--bright/brightness option enables you to adjust brightness of background image
          or transparent background.

        o -o/--lsp/line_space option spcifies the number of extra dots between lines.

        o -K/--maxptys/max_ptys option specifies max pty windows (1 - 32).
          If you specify 32 , you can use 32 pty windows in one process.

* ChangeLog
[20020203]
* cursor disappears under backscroll mode. fixed. (Thanks to Sakamoto Hironori san)
* -K/--maxptys/max_ptys option is added. You can use more than 5 pty windows in one
  process.
  Note that max_ptys option accepts numbers between 1 and 32.

[20020202]
* --shade/shade_ratio option is renamed to --bright/brightness.
* brightness option works under transparent mode.

[20020202]
* -H/--shade/shade_ratio option is added.

[20020202]
* if number of terminal rows is double as many as that of cols , mlterm dumps core. fixed.
* mongol vertical mode is added.
* "Font" Tab is added to mlconfig.

[20020201]
* -o/--lsp/line_space option is added.

[20020131]
* If both vertical view and bidi are used at the same time , terminal window
  can become too large. fixed.
* under vertical mode settings in etc/tfont and etc/taafont are used.

--
kiken
j00...@ip...

Re: [Mlterm-dev-en] 2.2.0 segfault

[Araki K. <j00...@ip...>]

	Hi,

        Thanks for your report and sorry for my late reply.

	Subject: [Mlterm-dev-en] 2.2.0 segfault
	From: Noah Levitt <nl...@co...>
	Message-ID: <200...@co...>
	Date: Fri, 1 Feb 2002 17:40:08 -0500

> $ uname -a
> SunOS stealth 5.8 Generic_108528-05 sun4u sparc
>
> Configured with no options (except --prefix=$HOME/local)
>
> It segfaults immediately after I start:
>
> Program received signal SIGSEGV, Segmentation fault.
> 0xff3c56d4 in ?? ()
> (gdb) bt
> #0  0xff3c56d4 in ?? ()
> #1  0xff3cdf8c in ?? ()
> #2  0xff3ce1b8 in ?? ()
> #3  0xff3ce2e8 in ?? ()
> #4  0xff3ce41c in ?? ()
> #5  0xff1fce78 in _XsunRegisterIMInstantiateCallback ()
>    from /usr/openwin/lib/libX11.so.4
> #6  0xff1deba4 in XRegisterIMInstantiateCallback ()
>    from /usr/openwin/lib/libX11.so.4
> #7  0x28388 in ml_xim_init (display=0x5c8e0) at ml_xim.c:325
> #8  0x18eec in ml_window_manager_init (win_man=0xffbef308, 
>     disp_name=0x37008 "") at ml_window_manager.c:48
> #9  0x30518 in ml_term_manager_init (term_man=0xffbef308, argc=1, 
>     argv=0xffbef628) at ml_term_manager.c:793
> #10 0x2b384 in main (argc=1, argv=0xffbef624) at main.c:35
>
> This happens even if I start with --openim=false --xim=false

        I have no idea why this happens , sorry.
        
        Would you show me your XMODIFIERS and LC_CTYPE environmental variables ?
        And if you use an attached patch , mlterm starts normally ?
        
        # Perhaps all the versions of mlterm have the same problems ?
        
--
kiken
j00...@ip...

Index: ml_xim.c
===================================================================
RCS file: /home/ken/cvsroot/mlterm/src/ml_xim.c,v
retrieving revision 1.65
diff -u -r1.65 ml_xim.c
--- ml_xim.c	2002/01/11 15:49:22	1.65
+++ ml_xim.c	2002/02/02 11:52:24
@@ -322,8 +322,10 @@
 		}
 	}
 
+#if  0
 	XRegisterIMInstantiateCallback( xim_display , NULL , NULL , NULL ,
 		xim_server_instantiated , NULL) ;
+#endif
 		
 	return  1 ;
 }

[Mlterm-dev-en] 2.2.0 segfault

[Noah L. <nl...@co...>]

Hello,

$ uname -a
SunOS stealth 5.8 Generic_108528-05 sun4u sparc

Configured with no options (except --prefix=$HOME/local)

It segfaults immediately after I start:

Program received signal SIGSEGV, Segmentation fault.
0xff3c56d4 in ?? ()
(gdb) bt
#0  0xff3c56d4 in ?? ()
#1  0xff3cdf8c in ?? ()
#2  0xff3ce1b8 in ?? ()
#3  0xff3ce2e8 in ?? ()
#4  0xff3ce41c in ?? ()
#5  0xff1fce78 in _XsunRegisterIMInstantiateCallback ()
   from /usr/openwin/lib/libX11.so.4
#6  0xff1deba4 in XRegisterIMInstantiateCallback ()
   from /usr/openwin/lib/libX11.so.4
#7  0x28388 in ml_xim_init (display=0x5c8e0) at ml_xim.c:325
#8  0x18eec in ml_window_manager_init (win_man=0xffbef308, 
    disp_name=0x37008 "") at ml_window_manager.c:48
#9  0x30518 in ml_term_manager_init (term_man=0xffbef308, argc=1, 
    argv=0xffbef628) at ml_term_manager.c:793
#10 0x2b384 in main (argc=1, argv=0xffbef624) at main.c:35

This happens even if I start with --openim=false --xim=false

Noah

[Mlterm-dev-en] commit log [2002/01/30]

[Araki K. <j00...@ip...>]

       Hi,

       I committed changes below.

* vertical view support is added.(-G/--vertical/vertical_mode option)
* actual screen size can be specified with -1/--wscr/screen_width_ratio and
  -2/--hscr/screen_height_ratio options regardless of console columns and rows.

       Now mlterm supports vertical view which is usual in cjk.
       mlterm -G cjk enables vertical view.

       Here is a screen shot.
       http://www.geocities.co.jp/SiliconValley-Cupertino/6461/tate.png

       -2 50 option which makes the height of terminal window half is recommended
       under vertical view mode.

       Of cource , these options can be changed from mlconfig.

       Best Regards.
  
--
kiken
j00...@ip...

Re: [Mlterm-dev-en] 2.2.0 released

[Behdad E. <be...@ba...>]

Hi,

Sorry for not saying it earlier, please use fribidi-0.10.1, no much 
enhancements, but a small deference in configurations, with 
fribidi-0.10.0 you should #include <fribidi.h>, while with all other 
versions you need to #include <fribidi/fribidi.h>, so for backward 
compatibility I changed it to old style in 0.10.1.

Yours,
behdad


On Tue, 29 Jan 2002, Araki Ken wrote:

> 	$B9SLZ$G$9(B:-)
> 
>         2.2.0 $B%j%j!<%9$$$?$7$^$7$?!#(B
> 
>         $B%@%&%s%m!<%I(B => http://prdownloads.sourceforge.net/mlterm/mlterm-2.2.0.tar.gz
> 
>         $B$D$$$G$K!"%5%s%W%k2hA|$b99?7$7$^$7$?!#(B
> 
>         $B0J2<$O!"(B2.2.0 $B$N?75!G=$J$I$G$9!#(B
>         
> * variable length column is supported.
> * ISCII support codes are imported(but is not still supported formally).
> * fribidi 0.10.0 is supported.
> * Full Reset button is added to mlconfig.
> * color fading is supported.
> * big5 unstandard characters can be converted to and from ucs4.
> * cursor form is changed when window is focused or unfocused.
> * Background of color-reversed characters is drawn with fg color event under wall paper 
>   or tranparent mode.
> * ESC ] 20 ; pt BEL sequence is supported.
> * mlterm configuration protocol is decided and implemented.
> * w3mmlconfig(w3m local-CGI) is contributed by Sakamoto Hironori san.
> * mlconf_curses(perl + curses) is contributed by Minami Hirokazu san.
> * etc/{font|vfont|aafont|vaafont} format is changed.
>   (Dynalab font can be shown correctly by using this extension)
> * some characters of CP932 based true type fonts couldn't be shown , but fixed.
> * CP932 gaiji characters (IBM , NEC , NECIBM) <=> UCS conversion is supported.
> * and many minor bugs are fixed.
> 
> 	$B$G$O(B
> --
> kiken
> j00...@ip...
> 
> _______________________________________________
> Mlterm-dev-en mailing list
> Mlt...@li...
> https://lists.sourceforge.net/lists/listinfo/mlterm-dev-en
> 

-- 
Behdad Esfahbod				9 Bahman 1380, 2002 Jan 29 
<behdad at bamdad dot org>		[Finger for Geek Code]

You have to know how to accept rejection and reject acceptance.
	-- Ray Bradbury, advice to writers 

[Mlterm-dev-en] 2.2.0 released

[Araki K. <j00...@ip...>]

	Hi,

        I released mlterm 2.2.0.

        Download => http://prdownloads.sourceforge.net/mlterm/mlterm-2.2.0.tar.gz
        
        Screen shot is also updated.
          
        And these are new features of 2.2.0
        
* variable length column is supported.
* ISCII support codes are imported(but is not still supported formally).
* fribidi 0.10.0 is supported.
* Full Reset button is added to mlconfig.
* color fading is supported.
* big5 unstandard characters can be converted to and from ucs4.
* cursor form is changed when window is focused or unfocused.
* Background of color-reversed characters is drawn with fg color event under wall paper 
  or tranparent mode.
* ESC ] 20 ; pt BEL sequence is supported.
* mlterm configuration protocol is decided and implemented.
* w3mmlconfig(w3m local-CGI) is contributed by Sakamoto Hironori san.
* mlconf_curses(perl + curses) is contributed by Minami Hirokazu san.
* etc/{font|vfont|aafont|vaafont} format is changed.
  (Dynalab font can be shown correctly by using this extension)
* some characters of CP932 based true type fonts couldn't be shown , but fixed.
* CP932 gaiji characters (IBM , NEC , NECIBM) <=> UCS conversion is supported.
* and many minor bugs are fixed.

        and these are change logs after 2002/01/26.

[20020129]
* mlcurses_conf is updated(20020127 version).(thanks to Minami Hirokazu san)

[20020129]
* copy&paste with kterm fails. fixed. (thanks to Sakamoto Hironori san)
* minus value specified for unsigned int is treated as error. (thanks to Minami Hirokazu san)

[20020128]
* variable column width and character combining are turned on in ISCII encoding.

[20020128]
* doc/en/README.iscii is added.
* fribidi-config is used to check libs and cflags for fribidi.
* -h help messages are improved.(thanks to Kubota Tomohiro san)
* font_larger_smaller_size option is renamed to step_in_changing_font_size.
* man/mlterm.1 manual is updated.(thanks to Kubota Tomohiro san)

--
kiken
j00...@ip...

[Mlterm-dev-en] 2.2.0 released

[Araki K. <j00...@ip...>]

	荒木です:-)

        2.2.0 リリースいたしました。

        ダウンロード => http://prdownloads.sourceforge.net/mlterm/mlterm-2.2.0.tar.gz

        ついでに、サンプル画像も更新しました。

        以下は、2.2.0 の新機能などです。
        
* variable length column is supported.
* ISCII support codes are imported(but is not still supported formally).
* fribidi 0.10.0 is supported.
* Full Reset button is added to mlconfig.
* color fading is supported.
* big5 unstandard characters can be converted to and from ucs4.
* cursor form is changed when window is focused or unfocused.
* Background of color-reversed characters is drawn with fg color event under wall paper 
  or tranparent mode.
* ESC ] 20 ; pt BEL sequence is supported.
* mlterm configuration protocol is decided and implemented.
* w3mmlconfig(w3m local-CGI) is contributed by Sakamoto Hironori san.
* mlconf_curses(perl + curses) is contributed by Minami Hirokazu san.
* etc/{font|vfont|aafont|vaafont} format is changed.
  (Dynalab font can be shown correctly by using this extension)
* some characters of CP932 based true type fonts couldn't be shown , but fixed.
* CP932 gaiji characters (IBM , NEC , NECIBM) <=> UCS conversion is supported.
* and many minor bugs are fixed.

	では
--
kiken
j00...@ip...

[Mlterm-dev-en] commit log [2002/01/26]

[Araki K. <j00...@ip...>]

	Hi,

--- ChangeLog
[20020126]
* big5_buggy option didn't work(is always true). fixed.
* XIM fg/bg color may be faded. fixed.
* man/mlterm.1 manual is updated.

[20020125]
* mlconf_curses is updated(20020125 version) and LICENCE,Makefile,mlcc.html
  files are added. (thanks to Minami Hirokazu san)
* CP932 gaiji characters (IBM , NEC , NECIBM) <=> UCS conversion is supported.
* -z/--largesmall/font_larger_smaller_size option is added.

[20020124]
* use_cp932_ucs_for_xft option is true by default.
* screen is corrupt when scrolled. fixed.
* if font_size_range is not set , mlterm behaves unexpectedly. fixed.
* mlconf_curses(2002/01/24 version) is added. (thanks to Minami Hirokazu san)
* doc/ja/FAQ is added.

        I'll release 2.2.0 in a few days.
        Bug reports are sincerely welcome.

--
kiken
j00...@ip...