gqmpeg-devel — Discussion about the development of GQmpeg.

[Gqmpeg-devel] [ANN] Release 0.91.2

[<mc...@ch...>]

Hi all,
finally, I made myself to prepare (at least basic) webpage and release 
version 0.91.2 of GQmpeg.
You can get it from http://members.chello.cz/mcermak/gqmepg/

What's new? Not much. It's mostly bug fix release (and prove of my 
ability to finish something).
Here's the list:
     * Fixed recursive playlist adding.
     * Fixed mpg123 message parsing for version 0.60+.
     * Fixed ogg123 end of playing recognition.
     * Fixed time base in io_xmp.
     - Added possibility to build without build-in skin editor.

Enjoy,
Milan Cermak

PS: John Ellis, if you read this, please, send me a note whether you 
plan on futher development or if I can take over.

[Gqmpeg-devel] [patch] xmp time too fast

[<mc...@ch...>]

Hi all,
announcing more and more patches can be really boring. And this one 
fixes xmp player module which noone seems to use. Otherwise it will be 
discovered very fast.

This patch fixes two scholar mistakes, both in xmp module. One is in 
parsing hexadecimal number, where the parsed number was treated as 
octal. The second one is in counting time where time base returns time 
milliseconds but next it's handled as being seconds.
Both bugs together made time running very fast (like 16 hours in 2 
minutes) while playing with xmp.

So, enjoy,
Milan Cermak

[Gqmpeg-devel] [patch] missed end of ogg123 playing depending on locale

[<mc...@ch...>]

Hi all,
here is another bugfix. Depending on locale, gqmpeg may miss end of 
ogg123 playing. The reason is that gqmpeg look for "Done." in ogg123 
stdout. With czech locale (which I'm using) ogg123 reports "Hotovo." 
instead which gqmpeg ignores and keeps waiting for "Done.".
My patch removes this dependency on english (or C) locale. End is now 
detected by end of stream announce (by ogg123) and "broken pipe" event 
occurence.

Regards,
Milan Cermak

[Gqmpeg-devel] [patch] GQmpeg will hang while playing mp3

[<mc...@ch...>]

Hi all,
this one is somewhat critical. When playing mp3 with misplaced or 
corrupted ID3v2 header, mpg123 will complain. Unfortunately GQmpeg will 
interpret this message as end of playing and will waitpid until mpg123 
finish.
As a result GQmpeg is completely frozen during playing the song.

Regards,
Milan Cermak

[Gqmpeg-devel] [patch] ignoring mpg123 messages

[<mc...@ch...>]

Hi all,
here is another patch. This one adds some more messages generated by 
mpg123 that can be safely ignored.

In case there will be no reply, I plan to release version 0.92 myself by 
the end of May.

Regards,
Milan Cermak

[Gqmpeg-devel] [patch] playlist and skin editor

[<mc...@ch...>]

Hi all,
I had a good day today and made another two patches for gqmpeg. The 
first one "gqmpeg.playlist.patch" fixes label of first column in 
playlist window. Originally the label is empty. Patch adds '#' there.
The reason is that gettext localization uses empty string translation 
for metainformation about PO file.

Second patch is bigger change. It adds switch to enable build-in skin 
editor to 'configure' script. My intention is to make skin editor 
optionally stand-alone.

Regards,
Milan Cermak

[Gqmpeg-devel] [patch] mpg123 version parsing

[<mc...@ch...>]

Hi all,
here I send patch which fixes parsing of mpg123 version for mpg123 0.60 
and later.

Original code made slider skip by minutes. This was observed on gqmpeg 
0.90 and 0.91.1.
With patch, slider moves as expected.

Regards,
Milan Cermak

[Gqmpeg-devel] Greetings

[<mc...@ch...>]

Hi all,
I just looked for new version of GQmpeg (cause the one in Debian Etch - 
0.90 - seemed broken in few details) and found out that there was no 
version for few years.
So my first question here is: Is there any development on GQmpeg since 
version 0.91.1?
(I really like the concept of separate gui running players on 
background. It looks to me more unix-like.)

The second question will be if this is the right place to send patches 
to? (I looked in the code and fixed version parsing for mpg123 0.60+.)

Regards,
Milan Cermak

[Gqmpeg-devel] Skinning ComboBox

[Gustavo V. <val...@gm...>]

Hi.
I'm new to SLIK and I've been looking the sample skins and I need to know
whether or not it's possible to skin a combo box or drop down list?

Thanks in advance.

[Gqmpeg-devel] Feature Request?

[Chris A. <ch...@mo...>]

I'm a "long time" user of gqmpeg. I really enjoy the software. I use  
it in my carputer project ( http://moose.ca/node/49 ). I've been  
using 0.91.1 ever since it hit the Gentoo portage tree.

I dunno if gqmpeg is being further developed but if it is...

Would it be possible to incorporate a way to read/load song files and  
playlists from an iPod directly in gqmpeg? libgpod allows access to  
iPods but I don't know if it is practical. Here is the link to the  
libgpod home page: http://www.gtkpod.org/libgpod.html

What do you think?

TIA

Re: [Gqmpeg-devel] Compiling gqmpeg/gqradio

[John E. <jo...@ve...>]

Dzuy Nguyen wrote:
> 
> Hello,
> 
> I just wonder if it is possible to compile gqmpeg/gqradio without the
> skin editor?
> Would that trim down the size of the binaries?

Attached is a patch for GQradio to allow disabling the skin editor.

Binary before:

-rwxrwxr-x  1 johne johne 1478082 Jun 17 14:02 gqradio (debug)
-rwxrwxr-x  1 johne johne 473840  Jun 17 14:12 gqradio (strip)

and after:

-rwxrwxr-x  1 johne johne 1260709 Jun 17 14:02 gqradio (debug)
-rwxrwxr-x  1 johne johne 388112  Jun 17 14:13 gqradio (strip)


ui2_editor.c could have a few more of the ui_edit_* convenience functions
disabled, but preferences.c will need the ui_edit_list_* utils. GQmpeg should
see a similar size decrease.



A while back I started a lite version of GQradio, it works but the main window
interface using standard gtk widgets was never completed:

-rwxrwxr-x  1 johne johne 527276 Jun 17 14:20 gqradio-lite (debug)
-rwxrwxr-x  1 johne johne 213876 Jun 17 14:21 gqradio-lite (strip)



Greetings,
John


-- 
John Ellis <jo...@ve...>

http://gqview.sourceforge.net <GQview> | http://hideseek.sourceforge.net
http://gqmpeg.sourceforge.net <GQmpeg> |     <Preferences Hide and Seek>

[Gqmpeg-devel] Compiling gqmpeg/gqradio

[Dzuy N. <dz...@in...>]

Hello,

I just wonder if it is possible to compile gqmpeg/gqradio without the 
skin editor?
Would that trim down the size of the binaries?

Thanks.

Dzuy

Re: [Gqmpeg-devel] Extended characters in skins

[John E. <jo...@ve...>]

Emil Nowak wrote:
> 
> Hi again,
> In the SKIN-SPECS there are described two types of images for font's (section:
> fonts). And when the non-extended one is used, all other characters are
> converted to whitespaces. It would be nicer to convert them into "base
> character" (like "e with tail" for e). I think it can be done by adding
> additional check (if skin has extended characters) and if not, it should
> use g_utf8_normalize() function. Probably exist conversion would strip
> accents so the string will be more readable. (I'm just guessing it can be
> done that way).

The conversion is done in ui2_text.c, look for the use of the iso_ascii
variable. This broke, however, when porting to GTK2, as all strings are now
stored internally as utf-8.

The attached patch for ui2_text.c will convert the text to ascii (ISO-8859-1)
when drawing the bitmapped characters so that the conversion will work properly.

Note:
New skins should only be using bitmapped characters when necessary, for example
when attempting to mimmick the text display of real-world equipment (car/home
stereo, lcd display, etc.). All other cases are encouraged to use the newer font
description (description = XXX) instead of an image file (image = XXX), for
example:

[text_WIDGETNAME]
description = Sans 12
width = 130
red = 0
green = 0
blue = 0
alpha = 255
...

You can play with this in the skin editor
(press tilde '~' to open skin editor -> add... button -> text tab).


This fix will be included in the next release.


Greetings,
John

-- 
John Ellis <jo...@ve...>

http://gqview.sourceforge.net <GQview> | http://gqapplets.sourceforge.net
http://gqmpeg.sourceforge.net <GQmpeg> |                  <panel applets>

[Gqmpeg-devel] Extended characters in skins

[Emil N. <em...@go...>]

Hi again,
In the SKIN-SPECS there are described two types of images for font's (section:
fonts). And when the non-extended one is used, all other characters are
converted to whitespaces. It would be nicer to convert them into "base
character" (like "e with tail" for e). I think it can be done by adding
additional check (if skin has extended characters) and if not, it should
use g_utf8_normalize() function. Probably exist conversion would strip
accents so the string will be more readable. (I'm just guessing it can be
done that way).

Re: [Gqmpeg-devel] errors in gettext support

[John E. <jo...@ve...>]

Emil Nowak wrote:
> 
> Hi,
> I was trying to make new translation (for Polish languge), but after I have
> translated few strings i made msgfmt. All msg's containing polish special
> characters doesn't appear in gui, instead of them I got empty space or original
> message. The real mess is on the playlist window the po-header is here :-P. You
> can see all this errors on screenshot:
> http://emil.kicks-ass.org/zonker/gqmpeg_error1.png
> I was trying to use po file in iso-8859-2 and in utf-8 (still same errors).

Thanks for reporting this.


I will release a fixed GQmpeg shortly, till then the attached patch will fix the
utf-8 conversion warnings, errant list title in the playlist window, and pass
the correct filename encoding to mpg123.

Greetings,
John

-- 
John Ellis <jo...@ve...>

http://gqview.sourceforge.net <GQview> | http://gqapplets.sourceforge.net
http://gqmpeg.sourceforge.net <GQmpeg> |                  <panel applets>

[Gqmpeg-devel] cannot read iso-8859-2 paths

[Emil N. <em...@go...>]

Hi,
There is some big problem opening and playing files with polish characters.=
 I
got my locales set to iso-8859-2 and all directories and filenames are
in this encoding. Runing gqmpeg from console I got this messages:

(gqmpeg:22263): Gdk-WARNING **: Error converting from UTF-8
to'ISO-8859-2': Invalid byte sequence in conversion input

(gqmpeg:22263): Gdk-WARNING **: Error converting from UTF-8 to 'ISO-8859-2':
Invalid byte sequence in conversion input Error, mpg123
reported:"/segate/muzyka/pl/KSU - Pod Pr=C4d/01 - pod pr=C4d.mp3: No such f=
ile or
directory"

[Gqmpeg-devel] errors in gettext support

[Emil N. <em...@go...>]

Hi,
I was trying to make new translation (for Polish languge), but after I have
translated few strings i made msgfmt. All msg's containing polish special
characters doesn't appear in gui, instead of them I got empty space or original
message. The real mess is on the playlist window the po-header is here :-P. You
can see all this errors on screenshot:
http://emil.kicks-ass.org/zonker/gqmpeg_error1.png
I was trying to use po file in iso-8859-2 and in utf-8 (still same errors).

[Gqmpeg-devel] changing player commands?

[Evren Y. <eyu...@tu...>]

Hi

wouldnt it be nice if we could define lets say how mpeg123 will be called
from a menu? I have trouble with kde artsd monopolizing the dsp, there is
a solution to use "artsdsp mpeg123 blah blah" but it is impossible to set
in gqmpeg as far as I could see???

Evren

[Gqmpeg-devel] ogginfo vorbistools 1.0 patch

[Duane P. <pa...@ro...>]

Here is a patch I created to fix ogg support in gqmpeg 0.19.0 when using
the vorbistools 1.0 final release ogginfo tool.  The patch also applies
cleanly to 0.18.0.  The patch needs to be applied from within the 
gqmpeg-0.19.0/src directory.

I've tested the patch on Solaris 7 and 8 SPARC.

The other patch I've applied was to libvorbis 1.0 lib/psy.c from Jason
Bruce to fix oggenc core dumping during file creation due to a buggy qsort
in the Solaris libc.  This patch should not affect gqmpeg at all, just an
FYI for anyone trying to create vorbis ogg files on Solaris. You can find
Jason's patch at

http://www.xiph.org/archives/vorbis-dev/200207/att-0182/01-libv.patch

Happy Ogging!

Duane Paddock

--- CUT HERE ---
--- io_ogg123.c.orig	Sat Jul 20 20:04:42 2002
+++ io_ogg123.c	Thu Nov 14 07:51:16 2002
@@ -132,7 +132,7 @@
 	gchar *ptr, *ptr2;
 
 	ptr = text;
-	while (*ptr != '=' && *ptr != '\0') ptr++;
+	while (*ptr != '=' && *ptr != '\0' && *ptr != ':' ) ptr++;
 	if (*ptr == '\0') return NULL;
 	ptr++;
 
@@ -179,6 +179,8 @@
 	gint child_pid;
 	int child_pipe[2];
 	char buf[128];
+        gint minutes;
+        gint seconds;
 
 	if (!ogg123_info_found) return NULL;
 
@@ -195,45 +197,71 @@
 		{
 		const gchar *value;
 		value = info_line_value(buf);
+                if (debug_mode > 2) printf("buf: %s\n", buf );
 
-		if (info_line_key_test(buf, "filename"))
+		if (info_line_key_test(buf, "Vorbis headers parsed"))
 			{
 			info->pass_integrity = TRUE;
+                        if (debug_mode) printf("Vorbis headers parsed: %d\n", info->pass_integrity );
 			}
-		else if (info_line_key_test(buf, "length"))
+		else if (info_line_key_test(buf, "\tPlayback length"))
 			{
-			info->length = (gint)strtol(value, NULL, 10);
+                        (void) sscanf (value, "%ld%*c%*c%02ld", &minutes, &seconds );
+			info->length = (gint) (minutes * 60)  + seconds;
+                        if (debug_mode) printf("Playback length: %ldm:%02lds or %d seconds\n", \
+                                minutes, seconds, info->length);
 			}
+#if 0
 		else if (info_line_key_test(buf, "header_integrity") ||
 			 info_line_key_test(buf, "stream_integrity"))
 			{
 			info->pass_integrity = info->pass_integrity & (strncmp(value, "pass", 4) == 0);
 			}
-		else if (info_line_key_test(buf, "bitrate_nominal"))
+#endif
+		else if (info_line_key_test(buf, "Nominal bitrate"))
 			{
 			info->bitrate_nominal = (gint)strtol(value, NULL, 10);
+                        if (debug_mode) printf("Nominal bitrate: %d\n", info->bitrate_nominal);
 			}
-		else if (info_line_key_test(buf, "bitrate_average"))
+		else if (info_line_key_test(buf, "\tAverage bitrate"))
 			{
 			info->bitrate_average = (gint)strtol(value, NULL, 10);
+                        if (debug_mode) printf("Average bitrate: %d\n", info->bitrate_average);
 			}
-		else if (info_line_key_test(buf, "bitrate_upper"))
+		else if (info_line_key_test(buf, "Upper bitrate"))
 			{
-			if (strncmp(value, "none", 4) == 0)
-				info->bitrate_upper = 0;
-			else
-				info->bitrate_upper = (gint)strtol(value, NULL, 10);
+                            if ( value != NULL )
+                            {
+        	    		if (strncmp(value, "not set", 7) == 0)
+	        			info->bitrate_upper = 0;
+		        	else
+			        	info->bitrate_upper = (gint)strtol(value, NULL, 10);
+                            }
+                            else
+                            {
+                                info->bitrate_upper = 0;
+                            }
+                            if (debug_mode) printf("Upper bitrate: %d\n", info->bitrate_upper);
+			}
+		else if (info_line_key_test(buf, "Lower bitrate"))
+			{
+                            if ( value != NULL )
+                            {
+        	    		if (strncmp(value, "not set", 7) == 0)
+	        			info->bitrate_lower = 0;
+		        	else
+			        	info->bitrate_lower = (gint)strtol(value, NULL, 10);
+                            }
+                            else
+                            {
+                                info->bitrate_lower = 0;
+                            }
+                            if (debug_mode) printf("Lower bitrate: %d\n", info->bitrate_lower);
 			}
-		else if (info_line_key_test(buf, "bitrate_lower"))
-			{
-			if (strncmp(value, "none", 4) == 0)
-				info->bitrate_lower = 0;
-			else
-				info->bitrate_lower = (gint)strtol(value, NULL, 10);
-			}
-		else if (info_line_key_test(buf, "channels"))
+		else if (info_line_key_test(buf, "Channels"))
 			{
 			info->channels = (gint)strtol(value, NULL, 10);
+                        if (debug_mode) printf("Channels: %d\n", info->channels);
 			}
 		}
 
@@ -637,9 +665,14 @@
  *			Time: 00:01.50 [04:32.13] of 04:33.63, Bitrate: 122.3
  * 1.0 rc3:
  *			Time: 00:03.92 [04:48.17] of 04:52.09  (140.9 kbps)  Output Buffer 75.0%
+ * 1.0:
+ *                      Time: 00:04.76 [03:12.11] of 03:16.87  (233.1 kbps)  Output Buffer   3.1%  
+ *
+ *
  *
  * as of right now this only effects parse_bitrate()
  */
+
 
 	if (strncmp(buffer, "Time: ", 6) == 0)
 		{
--- CUT HERE ---