headers
Urs Janßen | 3 Nov 2010 14:57

[tin 1.9.x] new snapshots

I've made new snapshots including some minor fixes for the !USE_CURSES
case. I'll release tin 1.9.6 in about two weeks.

<ftp://ftp.tin.org/pub/news/clients/tin/unstable/snapshots/≥
Permalink | Reply | 0 comments |
headers
Dennis Preiser | 6 Nov 2010 13:52
Picon
Picon

[PATCH] small fix for save.c:show_attachment_page()

In save.c:show_attachment_page() we determine the largest
content-type/encoding... -part of all attachments (info_len). The
current loop starts at attmenu.first (first line/attachment on screen)
which is wrong. The loop has to start at 0 to consider all attachments,
otherwise the screen layout may change during scrolling.

Dennis

diff -urp tin-1.9.6/src/save.c tin-1.9.6_r1/src/save.c
--- tin-1.9.6/src/save.c	2010-10-29 17:17:43.000000000 +0200
+++ tin-1.9.6_r1/src/save.c	2010-11-06 13:27:26.000000000 +0100
 <at>  <at>  -1524,7 +1524,7  <at>  <at>  show_attachment_page(
 		attmenu.curr = 0;

 	info_len = max_depth = 0;
-	for (i = attmenu.first; i < attmenu.max; ++i) {
+	for (i = 0; i < attmenu.max; ++i) {
 		part = get_part(i);
 		snprintf(buf, sizeof(buf), _(txt_attachment_lines), part->line_count);
 		tmp_len = strwidth(buf);
Permalink | Reply | 0 comments |
headers
Dennis Preiser | 13 Nov 2010 10:16
Picon
Picon

col_inverse_bg=0 && col_inverse_fg=7

While playing with colors I found the following issue which happens in
the USE_CURSES && HAVE_USE_DEFAULT_COLORS case:

It's not possible to set col_inverse_bg=0 (black) and col_inverse_fg=7
(white). tin uses the (terminal) default fg an bg colors instead and
thus the inverse bar gets 'invisible'. Steps to reproduce:

open 'xterm -fg yellow -bg blue', start tin
    'M'enu
	Draw -> instead of highlighted bar  : OFF
	Use inverse video for page headers  : ON
    Standard foreground color           : Default
    Standard background color           : Default
    Color for inverse text (foreground) : White
    Color for inverse text (background) : Black

-> the inverse bar is 'invisible'.

What happens: During startup use_default_colors() (in
tcurses.c:InitScreen()) assign terminals fg and bg to color pair 0.
Later on, color.c:set_colors() checks if the given fcolor and bcolor
matches COLOR_WHITE and COLOR_BLACK. If so, color pair 0 is used.
Unfortunately, for HAVE_USE_DEFAULT_COLORS, color pair 0 may hold other
color informations. For col_inverse_bg=0 && col_inverse_fg=7 this
results in an inverse bar with standard fg/bg -> the bar is 'invisible'.

The following fixes this for me, alas I'm not an curses expert so I
might be wrong:

diff -urp tin-1.9.6/src/color.c tin-1.9.6_r2/src/color.c
(Continue reading)

Permalink | Reply | 3 comments |
headers
Thomas Dickey | 13 Nov 2010 13:48
Picon

Re: col_inverse_bg=0 && col_inverse_fg=7

On Sat, 13 Nov 2010, Dennis Preiser wrote:

> While playing with colors I found the following issue which happens in
> the USE_CURSES && HAVE_USE_DEFAULT_COLORS case:
>
> It's not possible to set col_inverse_bg=0 (black) and col_inverse_fg=7
> (white). tin uses the (terminal) default fg an bg colors instead and
> thus the inverse bar gets 'invisible'. Steps to reproduce:
>
> open 'xterm -fg yellow -bg blue', start tin
>    'M'enu
> 	Draw -> instead of highlighted bar  : OFF
> 	Use inverse video for page headers  : ON
>    Standard foreground color           : Default
>    Standard background color           : Default
>    Color for inverse text (foreground) : White
>    Color for inverse text (background) : Black
>
> -> the inverse bar is 'invisible'.
>
> What happens: During startup use_default_colors() (in
> tcurses.c:InitScreen()) assign terminals fg and bg to color pair 0.
> Later on, color.c:set_colors() checks if the given fcolor and bcolor
> matches COLOR_WHITE and COLOR_BLACK. If so, color pair 0 is used.
> Unfortunately, for HAVE_USE_DEFAULT_COLORS, color pair 0 may hold other
> color informations. For col_inverse_bg=0 && col_inverse_fg=7 this
> results in an inverse bar with standard fg/bg -> the bar is 'invisible'.
>
> The following fixes this for me, alas I'm not an curses expert so I
> might be wrong:
(Continue reading)

Permalink | Reply | 2 comments |
headers
Dennis Preiser | 13 Nov 2010 14:29
Picon
Picon

Re: col_inverse_bg=0 && col_inverse_fg=7

On Sat, Nov 13, 2010 at 07:48:02AM -0500, Thomas Dickey wrote:
> On Sat, 13 Nov 2010, Dennis Preiser wrote:
>>diff -urp tin-1.9.6/src/color.c tin-1.9.6_r2/src/color.c
>>--- tin-1.9.6/src/color.c	2010-05-07 16:00:53.000000000 +0200
>>+++ tin-1.9.6_r2/src/color.c	2010-11-12 18:01:32.000000000 +0100
>> <at>  <at>  -110,8 +110,12  <at>  <at>  set_colors(
>>		if (bcolor > 0)
>>			bcolor %= COLORS;
>>
>>+#		ifdef HAVE_USE_DEFAULT_COLORS
>>+		if (fcolor != default_fcol || bcolor != default_bcol) {
>>+#		else
>>		/* curses assumes white/black */
>>		if (fcolor != COLOR_WHITE || bcolor != COLOR_BLACK) {
>>+#		endif /* HAVE_USE_DEFAULT_COLORS */
>>			struct LIST *p;
>>			t_bool found = FALSE;
>>
>>default_fcol and default_bcol are either white/black or, if
>>use_default_colors() succeeds, -1/-1 (default colors from terminal).
> 
> something like that (I'd be inclined to fold the two if-lines together
> by defining DEFAULT_FCOL and DEFAULT_BCOL somewhere, and using those
> for cases like this).

Thanks. Initially, default_fcol is 7 and default_bcol is 0. This changes
to -1/-1 only when use_default_colors() succeeds. So the diff shrinks
to:

diff -urp tin-1.9.6/src/color.c tin-1.9.6_r2/src/color.c
(Continue reading)

Permalink | Reply | 1 comment |
headers
Thomas Dickey | 13 Nov 2010 14:30
Picon

Re: col_inverse_bg=0 && col_inverse_fg=7

On Sat, 13 Nov 2010, Dennis Preiser wrote:

> On Sat, Nov 13, 2010 at 07:48:02AM -0500, Thomas Dickey wrote:
>> On Sat, 13 Nov 2010, Dennis Preiser wrote:
>>> diff -urp tin-1.9.6/src/color.c tin-1.9.6_r2/src/color.c
>>> --- tin-1.9.6/src/color.c	2010-05-07 16:00:53.000000000 +0200
>>> +++ tin-1.9.6_r2/src/color.c	2010-11-12 18:01:32.000000000 +0100
>>>  <at>  <at>  -110,8 +110,12  <at>  <at>  set_colors(
>>> 		if (bcolor > 0)
>>> 			bcolor %= COLORS;
>>>
>>> +#		ifdef HAVE_USE_DEFAULT_COLORS
>>> +		if (fcolor != default_fcol || bcolor != default_bcol) {
>>> +#		else
>>> 		/* curses assumes white/black */
>>> 		if (fcolor != COLOR_WHITE || bcolor != COLOR_BLACK) {
>>> +#		endif /* HAVE_USE_DEFAULT_COLORS */
>>> 			struct LIST *p;
>>> 			t_bool found = FALSE;
>>>
>>> default_fcol and default_bcol are either white/black or, if
>>> use_default_colors() succeeds, -1/-1 (default colors from terminal).
>>
>> something like that (I'd be inclined to fold the two if-lines together
>> by defining DEFAULT_FCOL and DEFAULT_BCOL somewhere, and using those
>> for cases like this).
>
> Thanks. Initially, default_fcol is 7 and default_bcol is 0. This changes
> to -1/-1 only when use_default_colors() succeeds. So the diff shrinks
> to:
(Continue reading)

Permalink | Reply | 0 comments |
headers
Urs Janßen | 14 Nov 2010 19:38

[PATCH] fix some (hypothetic) null pointer dereference

--- src/cook.c	2010-10-03 10:58:09 +0000
+++ src/cook.c	2010-11-14 10:53:47 +0000
 <at>  <at>  -455,7 +455,7  <at>  <at> 
 				if (max_line_len < strlen(buf) + 2) {
 					max_line_len = strlen(buf) + 2;
 					line = my_realloc(line, max_line_len);
-				};
+				}
 				strcpy(line, buf);

 				/*

--- src/filter.c	2010-09-08 21:25:42 +0000
+++ src/filter.c	2010-11-12 23:41:57 +0000
 <at>  <at>  -517,11 +517,11  <at>  <at> 
 							FreeIfNeeded(ptr[i].subj);
 							ptr[i].subj = my_strdup(subj);
 						}
-					}
 #ifdef DEBUG
-					if (debug & DEBUG_FILTER)
-						debug_print_file("FILTER","buf=[%s]  Gsubj=[%s]", ptr[i].subj, glob_filter.filter[i].subj);
+						if (debug & DEBUG_FILTER)
+							debug_print_file("FILTER","buf=[%s]  Gsubj=[%s]", ptr[i].subj, glob_filter.filter[i].subj);
 #endif /* DEBUG */
+					}
 					break;
 				}

--- src/nntplib.c	2010-11-03 19:17:32 +0000
(Continue reading)

Permalink | Reply | 0 comments |
headers
Dennis Preiser | 19 Nov 2010 16:13
Picon
Picon

doc/config-anomalies cleanup

5 former tinrc variables can be removed and auto_cc + auto_bcc have been
merged into auto_cc_bcc.

Dennis

--- tin-1.9.6/doc/config-anomalies	2010-03-19 23:35:49.000000000 +0100
+++ tin-1.9.6_r3/doc/config-anomalies	2010-11-19 16:01:04.000000000 +0100
 <at>  <at>  -51,11 +51,6  <at>  <at>  prompt_followupto
 pos_first_unread
 show_only_unread_arts
 kill_level                   *
-tab_goto_next_unread         *
-space_goto_next_unread       *
-pgdn_goto_next               *
-full_page_scroll             *
-show_last_line_prev_page     *
 group_catchup_on_exit
 thread_catchup_on_exit
 thread_articles
 <at>  <at>  -88,8 +83,7  <at>  <at>  hide_uue                     *
 news_quote_format
 mail_quote_format            *
 xpost_quote_format           *
-auto_cc
-auto_bcc
+auto_cc_bcc
 default_filter_days          *
 getart_limit                 *
 recent_time                  *
(Continue reading)

Permalink | Reply | 0 comments |
headers
Urs Janßen | 24 Nov 2010 18:42

questions about tin-1.5.6 to tin-1.5.7 misc.c:copy_body() change

between tin 1.5.6 and tin 1.5.7 misc.c:copy_body() was changed -
does anyone remeber why? it may leed to wrong indented underlining
                                        ^^^^^^^^^^^^^^^^^^^^^^^^^^

--- tin-1.5.6/src/misc.c	2000-08-03 15:49:22.000000000 +0200
+++ tin-1.5.7/src/misc.c	2000-10-19 19:46:46.000000000 +0200
 <at>  <at>  -174,6 +174,8  <at>  <at> 
  * copy the body of articles with given file pointers,
  * prefix (= quote_chars), initials of the articles author
  * with_sig is set if the signature should be quoted
+ *
+ * TODO: rewrite from scratch, the code is awfull
  */
 void
 copy_body (
 <at>  <at>  -244,10 +246,9  <at>  <at> 
 			} else	/* line is empty */
 					retcode = fprintf (fp_op, "%s\n", (tinrc.quote_empty_lines ? prefixbuf : ""));
 		} else {		/* no initials in quote_string, just copy */
-			if ((buf[0] != '\n') || tinrc.quote_empty_lines) {
-				/* use blank-stripped quote string if line is already quoted */
-					retcode = fprintf (fp_op, "%s%s", ((buf[0] == '>') ? prefixbuf : prefix), buf);
-			} else
+			if ((buf[0] != '\n') || tinrc.quote_empty_lines)
+				retcode = fprintf (fp_op, "%s%s", ((buf[0] == '>' || buf[0] == ' ') ? prefixbuf : prefix), buf);  /* use
blank-stripped quote string if line is already quoted or beginns with a space */
+			else
 				retcode = fprintf (fp_op, "\n");
 		}
 		if (retcode == EOF) {
(Continue reading)

Permalink | Reply | 0 comments |
headers
Urs Janßen | 26 Nov 2010 16:40

[ANNOUNCE] tin 1.9.6 released

I've just released tin 1.9.6 which is the last 1.x.er release. The
next release will be tin 2.0.0. It would be nice if someone could
look over the documentation and translation files, do some test
builds and code checking, ...

<ftp://ftp.tin.org/pub/news/clients/tin/v1.9/tin-1.9.6.tar.gz>
<ftp://ftp.tin.org/pub/news/clients/tin/v1.9/tin-1.9.6.tar.bz2>
<ftp://ftp.tin.org/pub/news/clients/tin/v1.9/tin-1.9.6.tar.lzma>

-- 1.9.6 release 20101126 "Burnside" --

U025) Matt Anton <tin <at> syrius.org>
      ADD. updated french translation
      FIX. fr.po

U024) Toomas Soome <Toomas.Soome <at> microlink.ee>
      ADD. updated estonian translation
      FIX. et.po

U023) Joe Hansen <joedalton2 <at> yahoo.dk>
      ADD. updated danish translation
      FIX. da.po

U022) Thomas E. Dickey <dickey <at> invisible-island.net>
      BUG. _XOPEN_SOURCE was defined on darwin
      ADD. CF_SIGWINCH - compile check for SIGWINCH
      FIX. aclocal.m4, configure[.in], Makefile[.in]

U021) Dennis Preiser <onkelp <at> gmx.de>
      BUG. SIGWINCH at page level after a connection timeout lead to a crash
(Continue reading)

Permalink | Reply | 0 comments |

Gmane