fontforge-devel — Bug reports, feature requests, etc.

Re: [Fontforge-devel] small pb in problems.c

[George W. <gw...@si...>]

On Sat, 2004-08-28 at 00:00, Pierre Hanser wrote:
> I've a font with custom encoding which segfaults in problem.c
> 
> I can prevent this with this small patch:
Thanks, applied.

Re: [Fontforge-devel] bug while merging fonts

[George W. <gw...@si...>]

On Wed, 2004-09-01 at 06:03, Pierre HANSER wrote:
> I've tracked down a bug which appears when using
> the MergeFonts script function, with core dumping.
> 
> It seems that in BitmapsCopy, the pointer is moved to
> next twice.
Oops. It certainly does. Applied, thanks.

Re: [Fontforge-devel] request for distinction of messages

[George W. <gw...@si...>]

On Wed, 2004-09-22 at 06:02, ka...@kh... wrote:
> This is a request to modify messages for distinction in Japanese translation.
> 
> str_Vertical is used twice, one as a flag in View menu of metrics view
> and another as a option in [Flip] operation of Transformation dialog.
> I need to distinguish them by the context, whether they are used for
> 'vertical writing' or not.
Both patch files have been applied.

Re: [Fontforge-devel] Skew() in Scripts

[George W. <gw...@si...>]

On Fri, 2004-08-27 at 06:41, Michael Gährken wrote:
> I did a little change to the function bSkew() in scripting.c, which 
> is called in scripts [via Skew()]. I wanted more than the available 
> one degree precision, so I implemented the numerator-denom-concept in 
> that function. The new additional call is:
> Skew(angle-numerator, angle-denom[,ox,oy])
Patch has been applied, thanks.

[Fontforge-devel] Re: Wishlist: font printing into pdf

[Johan V. <jvr...@sq...>]

"Arne G=C3=B6tje (.=AB.=E7.=C2=9B.=C2=8F.)" <ar...@li...> writes:

> Well, ps2pdf is in the gs-common package version 0.3.6 on debian=20
> unstable.

Can you try a more recent version of Ghostscript, e.g. 8.14?

-- Johan

[Fontforge-devel] request for distinction of messages

[<ka...@kh...>]

This is a request to modify messages for distinction in Japanese translation.

str_Vertical is used twice, one as a flag in View menu of metrics view
and another as a option in [Flip] operation of Transformation dialog.
I need to distinguish them by the context, whether they are used for
'vertical writing' or not.

In Japanese, the word 'yokogaki' is used for horizontal writing style 
and 'tategaki' for vertical writing style. Common suffix of them
('gaki') means 'writing' and omitting this in this context is unnatural.

In geometrical context for splines, I translated them as 'suihei' and
'suichoku' instead of 'yoko' and 'tate', which is somewhat confusing.
Because sometimes Japanese word 'yoko' means '(left and right) sides'. 

This patch replaces str_Vertical with str_VerticalW when it implies 
vertical writing and moved str_Vertical into the part of transformation 
menu.

--- metricsview.c.orig	Thu Aug 12 07:27:45 2004
+++ metricsview.c	Fri Sep 17 10:08:43 2004
@@ -2793,7 +2793,7 @@
     { { (unichar_t *) _STR_Hidegrid, NULL, COLOR_DEFAULT, COLOR_DEFAULT, NULL, NULL, 0, 0, 0, 0, 0, 0, 0, 1, 0, 'G' }, '\0', ksm_control, NULL, NULL, MVMenuShowGrid, MID_ShowGrid },
     { { (unichar_t *) _STR_Antialias, NULL, COLOR_DEFAULT, COLOR_DEFAULT, NULL, NULL, 0, 0, 1, 0, 0, 0, 0, 1, 0, 'A' }, '5', ksm_control, NULL, NULL, MVMenuAA, MID_AntiAlias },
     { { NULL, NULL, COLOR_DEFAULT, COLOR_DEFAULT, NULL, NULL, 0, 0, 0, 0, 0, 1, 0, 0, }},
-    { { (unichar_t *) _STR_Vertical, NULL, COLOR_DEFAULT, COLOR_DEFAULT, NULL, NULL, 0, 0, 1, 0, 0, 0, 0, 1, 0, '\0' }, '\0', ksm_control, NULL, NULL, MVMenuVertical, MID_Vertical },
+    { { (unichar_t *) _STR_VerticalW, NULL, COLOR_DEFAULT, COLOR_DEFAULT, NULL, NULL, 0, 0, 1, 0, 0, 0, 0, 1, 0, '\0' }, '\0', ksm_control, NULL, NULL, MVMenuVertical, MID_Vertical },
     { { NULL, NULL, COLOR_DEFAULT, COLOR_DEFAULT, NULL, NULL, 0, 0, 0, 0, 0, 1, 0, 0, }},
     { { (unichar_t *) _STR_OutlineMn, NULL, COLOR_DEFAULT, COLOR_DEFAULT, NULL, NULL, 0, 0, 1, 0, 0, 0, 0, 1, 0, 'O' }, '\0', ksm_control, NULL, NULL, MVMenuShowBitmap, MID_Outline },
     { NULL },			/* Some extra room to show bitmaps */
--- nomen-en.c.orig	Mon Aug 23 06:30:00 2004
+++ nomen-en.c	Fri Sep 17 10:05:56 2004
@@ -473,8 +473,8 @@
 static char str_Outline[] = "Outline";
 static char str_OutlineMn[] = "Outline";
 static unichar_t mnemonic_OutlineMn[] = 'O';
-static char str_Vertical = "Vertical";
-static unichar_t mnemonic_Vertical = 'V';
+static char str_VerticalW[] = "Vertical";
+static unichar_t mnemonic_VerticalW = 'V';
 static char str_Showgrid[] = "Show Grid";
 static unichar_t mnemonic_Showgrid[] = 'G';
 static char str_Hidegrid[] = "Hide Grid";
@@ -2963,6 +2963,7 @@
 static char *str_DegreesClockwise   = "\260 Clockwise";
 static char *str_DegreesWithershins = "\260 Withershins";	/* deiseal */
 static char *str_Horizontal = "Horizontal";
+static char *str_Vertical = "Vertical";
 static char *str_PercentMark = "%";
 static char *str_DegreeMark = "\260";
 static char *str_RotationAboutXAxis = "Rotation about X Axis";


My update of nomen-ja.c reflects this modification. 
I also fixed many mistranslation and refined many entries to
match the translation of manuals in progress...

[Fontforge-devel] Re: Wishlist: font printing into pdf

[Arne ) <ar...@li...>]

On Wednesday 22 September 2004 15:40, Johan Vromans wrote:
> Personally, I'd say somwthing is wrong with the ps generation, or
> Arne uses an old version of ps2pdf. I'll try to get some more info
> today.

Well, ps2pdf is in the gs-common package version 0.3.6 on debian=20
unstable.
The ps file displays fine in gv and only in gv.
I was trying to print my CJK unicode font, the ps file looks ok, all=20
glyphs are there. When I convert the file with ps2pdf and then view it=20
in Acrobat Reader, the codepoint range from U+4F00 to U+4FFF is=20
completely missing and in the CJK-Compatible range I get the error=20
message that the pdf file is corrupted and the font cannot be found.=20
This happens only with one font (I have two fonts with identical=20
codepoints but different glyphs), the other one displays correctly.
Both fonts were made by me with fontforge and in that specific area=20
(CJK-Compatibility) I used references for the glyphs. However=20
references in other part of the fonts display correctly, so it's not a=20
general reference problem...

If someone wants to test, the fonts can be downloaded from
http://tavi.debian.org.tw/index.php?page=3DUnifonts
The ps file in case of the ming font is > 100 MB...
The page where Acrobat reports the error is page 63 (of 64).

Thanx for help.

Cheers
Arne
=2D-=20
Arne G=C3=B6tje (=E9=AB=98=E7=9B=9B=E8=8F=AF) <ar...@li...>
PGP/GnuPG key: 1024D/685D1E8C
=46ingerprint: 2056 F6B7 DEA8 B478 311F  1C34 6E9F D06E 685D 1E8C
Key available at wwwkeys.pgp.net.   Encrypted e-mail preferred.

[Fontforge-devel] Re: Wishlist: font printing into pdf

[Johan V. <jvr...@sq...>]

George Williams <gw...@si...> writes:

> On Tue, 2004-09-21 at 22:46, Arne wrote: 
>>  So, for convenience, would it be possible to add an option to the
>>  printing dialog to print into pdf directly?
> An interesting thought. I don't really know very much about building pdf
> files, to do this right will take considerable research on my part so it
> is not going to happen fast (at least not by me... if someone else wants
> to work on it that would be great).  I will keep it in mind though.

Personally, I'd say somwthing is wrong with the ps generation, or Arne
uses an old version of ps2pdf. I'll try to get some more info today.

-- Johan

[Fontforge-devel] Re: Small doc bug in Cygwin instructions

[George W. <gw...@si...>]

>I found one small bug in the Cygwin installation instructions:
 
>First you move the man page to /usr/local/man but add /usr/local/share/man 
>to the MANPATH.
Oops, thanks for pointing that out. It will be fixed in the next
release.

***** I find that while I was gone my mailbox filled up with 1000 ******
***** copies of the same (non-spam) message (but in some server)  ******
***** the result is that the original message never got to me and ******
***** I had to find it in the archives. But the archives don't    ******
***** have the full email address. Could someone send me (off list******
***** the email address of the sender of this message?            ******
*****                Thanks                                       ******

[Fontforge-devel] Re: Wishlist: font printing into pdf

[George W. <gw...@si...>]

On Tue, 2004-09-21 at 22:46, Arne wrote: 
>  So, for convenience, would it be possible to add an option to the
>  printing dialog to print into pdf directly?
An interesting thought. I don't really know very much about building pdf
files, to do this right will take considerable research on my part so it
is not going to happen fast (at least not by me... if someone else wants
to work on it that would be great).  I will keep it in mind though.

[Fontforge-devel] Fontforge problems

[George W. <gw...@si...>]

METAFONT bitmaps in the .pk format
	Fixed (ff didn't read the "long format" header properly)
	pk.patch

Flex sequences (how to suppress?)
	ff should be better about adding flexes now.
	flex.patch

Merging short segments
	Code for approximating spline keeping end-point slopes fixed is
	better (now uses a perturbation approach rather than least
	squares on errors)
	approx_by_perterbation.patch
Inaccurate simplify
	Same patch as above
Path with inflection
	I don't think there needs to be an additional point at the
	inflection problem goes away with the new approximation
	algorithem.
No subdivision
	I don't think subdivision is needed. The above patch improves
	things

Unseparated (and hidden) lineto
	It isn't a lineto. The slope of the spline makes an abrupt
	change at the end-points (ie the control points are very close
	to the end points, and point far from the other endpoint).
	Separation isn't relevant either, the problem is the abrupt
	change in slope.
	Here is a patch which alters the splines slightly before
	stroking them.
	tinycps.patch
Strange approximation
	This is the same problem, but more extreme.

Zero width (METAFONT parameterization)
	Ok, for the particular case of a fill & 0 width stroke I will
	remove the stroke.
	zerowidthstroke.patch

Bad fit of short (circular) curves
	Wasn't able to figure out what you meant here.

Import without simplify?
	Import is not expected to simplify. Not sure what you mean.

Removing overlap
	Remove overlap is documented to have bugs. I don't need
	examples, I have plenty.

[Fontforge-devel] Wishlist: font printing into pdf

[Arne G. <ar...@li...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi list,

currently fontforge supports printing a font into a postscript file.
However, when I print a truetype font, a temporary postscript version
of the font will be generated and then used to create the ps file.
So far so good. But when I try to convert the ps file into pdf using
ps2pdf, the resulting pdf gets corrupted and Acrobat Reader complains
about missing glyphs and corrupt font.
I guess for the conversion from ps to pdf, the postscript font is needed
again.
So, for convenience, would it be possible to add an option to the
printing dialog to print into pdf directly?

Cheers
Arne
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)

iD8DBQFBT1/Tbp/QbmhdHowRAjf2AKD8DM3oPWkyef7rM49PIspG6AVtrQCg/bfc
kPVPmmdk8Evwz+m/oK4nag8=
=j625
-----END PGP SIGNATURE-----

[Fontforge-devel] Small doc bug in Cygwin instructions

[Jan W. <ja...@wi...>]

Great tool! Actually the only free tool I found to convert Type1 fonts to 
TrueType.

I found one small bug in the Cygwin installation instructions:

First you move the man page to /usr/local/man but add /usr/local/share/man 
to the MANPATH.

Thanks for Fontforge,
Jan

[Fontforge-devel] Bugs with simplification

[Andrey V. P. <pa...@ca...>]

Hello.

Sometimes ff after applying of "Simplify" function creates artefacts. I have 
attached a font with examples.

In "H", point 282.09,340 after simplification control point is moved to 
another direction (rotated about 180 degrees against skewed line). In "yen", 
point 326.69,325.66 after simplification control point is created also in 
wrong direction.

-- 
Andrey V. Panov
panov /@/ canopus.iacp.dvo.ru

Re: [Fontforge-devel] problem with a Mac OS/X distribution. Version 24-Aug-2004

[<gw...@si...>]

Quoting Alexander Zatko <ale...@ef...>:
> Hi, after installation, when I attempt to launch the app I get the 
> following error:
> 
> dyld: /usr/local/bin/fontforge can't open library: 
> /usr/lib/libiconv.2.dylib  (No such file or directory, errno = 2)
> Trace/BPT trap
You need to upgrade to OS/X 10.3
(or you can build from source on your machine and it will be configured 
correctly)

I need to update my documentation to mention this.

Re: [Fontforge-devel] compile with gtk

[<gw...@si...>]

Quoting Behnam Esfahbod <be...@es...>:

> I wanted to compile source (fro cvs) with gtk enabled, so first I 
> configured like this:
This is under development. Don't try to use it yet.

[Fontforge-devel] problem with a Mac OS/X distribution. Version 24-Aug-2004

[Alexander Z. <ale...@ef...>]

Hi, after installation, when I attempt to launch the app I get the 
following error:

dyld: /usr/local/bin/fontforge can't open library: 
/usr/lib/libiconv.2.dylib  (No such file or directory, errno = 2)
Trace/BPT trap

thanks

A.

[Fontforge-devel] bug while merging fonts

[Pierre H. <Pie...@sx...>]

hello

I've tracked down a bug which appears when using
the MergeFonts script function, with core dumping.

It seems that in BitmapsCopy, the pointer is moved to
next twice.

The following patch seems to take care of this.

diff -C3 fvfonts.c.orig fvfonts.c
*** fvfonts.c.orig      Wed Sep  1 14:40:54 2004
--- fvfonts.c   Wed Sep  1 14:41:21 2004
***************
*** 240,248 ****
--- 240,250 ----
            t_bdf = t_bdf->next;
            f_bdf = f_bdf->next;
        } else if ( t_bdf->pixelsize < f_bdf->pixelsize )
+             {
            t_bdf = t_bdf->next;
        
            f_bdf = f_bdf->next;
+             }
      }
  }
  

-- 
	Pierre

[Fontforge-devel] small pb in problems.c

[Pierre H. <ha...@cl...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

hello

I've a font with custom encoding which segfaults in problem.c

I can prevent this with this small patch:

*** fontforge/problems.c.ph     2004-08-28 08:45:27.000000000 +0200
- --- fontforge/problems.c        2004-08-28 08:48:45.000000000 +0200
***************
*** 2981,2990 ****

~      sf = p.fv->sf;

!     if ( advancewidthval==0 && (sf->charcnt >= ' ') && sf->chars['
']!=NULL )
~        advancewidthval = sf->chars[' ']->width;
- -     else
- -         printf("charcnt: %d ' ': %d\n", sf->charcnt, ' ');
~      sprintf(awidthbuf,"%g",advancewidthval);
~      rlabel[2].text = (unichar_t *) awidthbuf;
~      rlabel[2].text_is_1byte = true;
- --- 2981,2988 ----

~      sf = p.fv->sf;

!     if ( advancewidthval==0 && sf->chars[' ']!=NULL )
~        advancewidthval = sf->chars[' ']->width;
~      sprintf(awidthbuf,"%g",advancewidthval);
~      rlabel[2].text = (unichar_t *) awidthbuf;
~      rlabel[2].text_is_1byte = true;

the printf was intended for debug only and says:
charcnt: 27 ' ': 32


- --
	Pierre
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.2 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFBMC2cMQ2/jYUwmmwRAkGnAKChj662OeuiI0ZZtK7l4dvZ9+qMpgCfSNgV
5+1Cc15EoLEZxa8pTOr1dEk=
=h9gh
-----END PGP SIGNATURE-----

[Fontforge-devel] Skew() in Scripts

[Michael <mga...@gm...>]

Hi,

I did a little change to the function bSkew() in scripting.c, which 
is called in scripts [via Skew()]. I wanted more than the available 
one degree precision, so I implemented the numerator-denom-concept in 
that function. The new additional call is:
Skew(angle-numerator, angle-denom[,ox,oy])

This is the function bSkew, starting in line 2327 in file scripting.c:
------------------------------------
static void bSkew(Context *c) {
     /* Arguments:
     2 => angle
     3 => angle-numerator, angle-denom
     4 => angle, origin-x, origin-y
     5 => angle-numerator, angle-denom, origin-x, origin-y
     */

     real trans[6];
     int otype = 1;
     BVTFunc bvts[2];
     double a;

     if ( c->a.argc==1 || c->a.argc>5 )
     error( c, "Wrong number of arguments");
     if ( c->a.vals[1].type!=v_int || (c->a.argc==3 && 
c->a.vals[2].type!=v_int) ||
         (c->a.argc==4 && (c->a.vals[2].type!=v_int || 
c->a.vals[3].type!=v_int )) ||
         (c->a.argc==5 && (c->a.vals[2].type!=v_int || 
c->a.vals[3].type!=v_int ||
                           c->a.vals[4].type!=v_int)))
     error(c,"Bad argument type in Skew");
     if (c->a.argc==3 || c->a.argc==5)
         a = c->a.vals[1].u.ival / (double) c->a.vals[2].u.ival;
     else {
         if ( (c->a.vals[1].u.ival %= 360)<0 ) c->a.vals[1].u.ival += 360;
         a = c->a.vals[1].u.ival;
     }
     a = a *3.1415926535897932/180.;
     trans[0] = trans[3] = 1;
     trans[1] = 0; trans[2] = tan(a);
     trans[4] = trans[5] = 0;
     if ( c->a.argc==4 ) {
         trans[4] = 
c->a.vals[2].u.ival-(trans[0]*c->a.vals[2].u.ival+trans[2]*c->a.vals[3].u.ival);
         trans[5] = 
c->a.vals[3].u.ival-(trans[1]*c->a.vals[2].u.ival+trans[3]*c->a.vals[3].u.ival);
         otype = 0;
     }
     if ( c->a.argc==5 ) {
         trans[4] = 
c->a.vals[3].u.ival-(trans[0]*c->a.vals[3].u.ival+trans[2]*c->a.vals[4].u.ival);
         trans[5] = 
c->a.vals[4].u.ival-(trans[1]*c->a.vals[3].u.ival+trans[3]*c->a.vals[4].u.ival);
         otype = 0;
     }
     skewselect(&bvts[0],trans[2]);
     bvts[1].func = bvt_none;
     FVTransFunc(c->curfv,trans,otype,bvts,true);
}
------------------------------------

Please note, that in the new call no check about the size of the 
requested angle is made, because I wasn't sure how to do it (I'm used 
to program in Pascal, not C)

Michael

[Fontforge-devel] compile with gtk

[Behnam E. <be...@es...>]

Hi list,

I wanted to compile source (fro cvs) with gtk enabled, so first I 
configured like this:

$ ./configure --prefix /home/behnam/opt/fontforge --with-gtk

But when I finished make and make install, I noticed that it doesn't use 
gtk, and there's no deference between default configuration and with-gtk.


Then I took a look at source and defined FONTFORGE_CONFIG_GTK manually 
in fontforge/configure_pfaedit.h.  But it makes this error:

gcc -I/usr/include/freetype2/ -I/usr/include/libxml2/ -I../inc -I. -I. 
-Wmissing-prototypes -Wunused -Wimplicit -Wreturn-type -Wparentheses 
-Wformat -Wchar-subscripts -I/usr/X11R6/include -DPACKAGE_NAME=\"\" 
-DPACKAGE_TARNAME=\"\" -DPACKAGE_VERSION=\"\" -DPACKAGE_STRING=\"\" 
-DPACKAGE_BUGREPORT=\"\" -DSTDC_HEADERS=1 -DHAVE_SYS_TYPES_H=1 
-DHAVE_SYS_STAT_H=1 -DHAVE_STDLIB_H=1 -DHAVE_STRING_H=1 
-DHAVE_MEMORY_H=1 -DHAVE_STRINGS_H=1 -DHAVE_INTTYPES_H=1 
-DHAVE_STDINT_H=1 -DHAVE_UNISTD_H=1 -DHAVE_DLFCN_H=1 -DHAVE_ICONV_H=1 
-D_NO_XKB=1 -D_HAS_LONGLONG=1 
'-DSHAREDIR="/home/behnam/opt/fontforge/share/fontforge"' 
-DLIBDIR='"/home/behnam/opt/fontforge/lib"'   -c -o autohint.o autohint.c
In file included from autohint.c:27:
pfaedit.h:32:24: gtkbasics.h: No such file or directory
pfaedit.h:33:22: gtk/gtk.h: No such file or directory
pfaedit.h:34:19: gwwv.h: No such file or directory


It'll find <gtk/gtk.h> if with-gtk works correctly, but I couldn't find 
"gtkbasics.h" anywhere.


Are these known bugs?  Is any binary for gtk back-end?


Regards,
-- 
     '
     '     Behnam Esfahbod
    '
   *  ..
  *  `  *  http://behnam.esfahbod.info/
   * o *

Re: [Fontforge-devel] little but on web page

[George W. <gw...@si...>]

On Sun, 2004-08-22 at 06:45, Simon Goodwin wrote:
> Source distribution
> Or you can go to the cvs tree and get the most recent version. (This is
> currently delayed by 24 hours)
> 
> 
> the cvs tree link still points to pfaedit! maybe point it to fontforge
> :-)
Oops. Thanks.
Corrections will appear when the next build is posted.

[Fontforge-devel] little but on web page

[Simon G. <si...@gm...>]

Source distribution
You can download a source RPM: Version: 08-Aug-2004
Or you can download the source treein one .tgz file: Version
08-Aug-2004docs 
  * Unicode widget library source (19-Jan-2002)
-- docs
  * Pfaedit source (19-Jan-2002)
-- docs
-->
Or you can go to the cvs tree and get the most recent version. (This is
currently delayed by 24 hours)


the cvs tree link still points to pfaedit! maybe point it to fontforge
:-)

Re: [Fontforge-devel] Re: [Pfaedit-devel] I'm impressed about FForge!

[Danilo S. <ds...@gm...>]

Today at 0:37, Danilo Segan wrote:

>> On Tue, 2004-08-17 at 05:09, Krzysztof D=C4=85bski wrote:
>>> In fact sometimes I've problem with message:"Naruszenie ochrony=20
>>> pamieci" but it obvious I think.=20
> ...
> Based on the similarity of all Slavic languages, I guess this is
> something like "Segmentation fault"

I remembered that I have a libc CVS checkout around, and I peeked into
po/pl.po, and I guessed it correctly:

#: sysdeps/generic/siglist.h:38 stdio-common/../sysdeps/unix/siglist.c:37
msgid "Segmentation fault"
msgstr "Naruszenie ochrony pami=C4=99ci"

I don't know how much this will help, but hey... ;-)

Cheers,
Danilo

[Fontforge-devel] Re: [Pfaedit-devel] I'm impressed about FForge!

[Danilo S. <ds...@gm...>]

(I moved the CC to fontforge-devel)

Yesterday at 21:55, George Williams wrote:

> On Tue, 2004-08-17 at 05:09, Krzysztof D=C4=85bski wrote:
>> In fact sometimes I've problem with message:"Naruszenie ochrony=20
>> pamieci" but it obvious I think.=20
> I'm afraid I do not speak that language, so whatever problems the
> message may have are not obvious to me, could you explain further?

Based on the similarity of all Slavic languages, I guess this is
something like "Segmentation fault" (pamieci - mind, memory,
naruszenie - error, "something bad done with", ochrony I cannot guess
what it might be, though it might be something like "store" or "save",
as in "pohrani" in Serbian and Croatian ;).

If I'm correct, I guess you wouldn't be able to do much without
testcases, backtraces, etc.

Cheers,
Danilo