[gs-devel] bug status report - 06-23-2004

Jack Moffitt jack at artifex.com
Wed Jun 23 06:43:16 PDT 2004

Total issues		:	179
Total bugs		:	130
Total enhancements	:	49
New bugs		:	16
Closed bugs		:	68
New potential bugs	:	00
Open potential bugs	:	01
Fixed potential bugs	:	00
New customer bugs	:	02
Open customer bugs	:	29
Fixed customer bugs	:	16
New top-ten bugs	:	00
Open top-ten bugs	:	07
Fixed top-ten bugs	:	01

Closers 		:	everyone (68)


#687520 - USER - Too few gray levels with WTS
assign: raph?

If WTS is enabled and sethalftone requests angles and rulings of an RT
screen, which perfectly fits into a very small halfone cell, then the
resulting screen has only relatively few gray levels.

#687521 - USER - Very slow .bindnow operator
assign: ray?

Execution of .bindnow operator (after DELAYBIND option) takes a lot of
time (about 5 seconds for binding operators from gs_*.ps files!).  Due
to this e.g. running a ps2ascii script is very slow.  Alex has a

#687523 - INTERNAL - Add determining the length of gx_color_index to genarch.c
submitted by: dan
assigned: ralph

gx_color_index can be either 32 or 64 bits.  It would be nice to have
something equivalent to the ARCH_SIZEOF_xx for this type.  This would
allow some conditional compiles to eliminate the code for the 64 bit
cases when the size of gx_color_index is only 32 bits.  Several
compilers (like the IBM C compiler on AIX RS6000) complain about the
64 bit code.

#687524 - USER - Make for 8.30 fails
assign: ralph?

User has trouble building 8.30.  Alex suggests that it is due to and
old zlib version which configure doesn't check for.

#687525 - USER - Slow rendering of bitmaps with Haftone type 6
assign: ray?

If the product of size of the Halftone threshold array and the
rendered bitmap is less than some value (about 10^11), or rendered
page has only `vector' elements, the rendering is very fast (even for
large resolutions).  But if threshold array is larger and the page
contains a bitmap, the rendering is propotional to the product of
length of the bitmap and length of the treshold array.

#687526 - CUSTOMER - /invalidfont in --show--
customer: #710
assign: ray?

The attached font from the customer fails after:
/BergellLET findfont 32 scalefont setfont 100 100 moveto (hello) show
with /invalidfont in --show--

#687527 - USER - Exception in GSAPI_INIT_WITH_ARGS with Delphi demo
assign: ??

The Delphi API demo (available from
throws an exception during the call to gsapi_ini_with_args.

#687528 - USER - using -sDEVICE=ljetplus only "print" a white paper
assign: ??

I used an old version of gs to printer I have updated to the lastest
version and now the printer (9022=ljetplus) only prints "while" pages.

#687529 - USER - Several diverse code fixes
assign: raph?

Fixes for compiler warnings and shell scripts

#687530 - USER - Corrupt conversion of JBIG2Decode enclosures in PDF
assign: ralph?

Converting a PDF file to a TIFFG4, that contains JBIG2 objects is,
creating corrupt output.

#687531 - USER - Duplex printing with ljet4d driver doesn't work
assign: ??

Code fixed in gdevdljm.c file (setting of page size -- bug 686815),
resets many settings: not only bad margins (bug 687255), but also
switch off duplex printing (tested on printer INFOTEC 4201MF -- but I
think, that all printers will behave the same).

#687532 - USER - Unable to find/search a text string in concatenated
		 PDF files
assign: igor?

When I concatenate two PDF files, I am not able to do a find or search
command on the resulting concatenated PDF file. The file opens
correctly in the Acrobat Reader. But when I search for a string/text,
I get zero occurances.

#687533 - USER - segmentation fault on this PS file
assign: ray?

User has a file which causes a segfault.

#687534 - INTERNAL - Unable to determine equivalent CMYK for spot colors
		     that are only used with overprinted
submitted by: dan
assigned: dan

The logic for determine an equivalent CMYK color for spot colors fails
if the spot color is only used when overprinting is enabled.  However
if the spot color is used with the fill_rectangle_hl_color method then
the equivalent CMYK color is determined.  This seems to be due to some
inconsistencies between the fill_rectangle_hl_color and the overprint

#687535 - CUSTOMER - thin lines after file rendered to tiff
customer: #770
assign: ray?

After running the attached test.ps with CVS HEAD (this in turn runs
the Sheet No 9.pdf) the resulting tiff has very thin lines in some
places.  These places are indicated in the SCAN0003.tif.

#458780 - raph - Illustrator 9 transparency rangecheck

#493348 - ray - Add image smoothing for better display
#578865 - jeong - Colors missing in AI9 file.
#610356 - raph - Banding in the gradient
#626464 - ray - /undefined in Adobe_level2_AI5
#634036 - ray - additional lines in gradient.
#661171 - ray - blends incorrect at high resolutions
#686824	- ray - borders are incorrect in pdf file
#686841 - ray - bbox device issue
#686859 - ray - currentfile file position off by 2048
#687157 - ray - file leaves interpreter in bad state
#687171 - ray - pdfmark issues
#687364 - ray - rangecheck in buildshading2 from Quark XPress PS file
#687380 - igor - 8.5MB PDF becomes 1GB PS with -dUseCIEColor
#687388 - ray - pdf files consume too much memory
#687434 - ray - /VMerror in --fill--
#687435 - ray - /ioerror in --token--
#687436 - igor - PDF output not rotated, but worked correctly in GS 8.11
#687444 - ray - incorrect bounding box with ccfont.dev
#687473 - ray - /invalidrestore in --restore--
#687474 - ray - damaged pdf files won't display with Ghostscript
#687475 - ray - font glyphs overlap (multimaster fonts)
#687476 - ray - /ioerror in --token--
#687510 - igor - pdfwrite: Acrobat 6.0 paints notdef instead spaces

#535366 - raph - /VMerror in .pushpdf14devicefilter
#611898 - raph - transparency rendering incorrect
#673124 - raph - /rangecheck in --.buildshading2--
#686789 - dan - gradients wrong after pdfwrite
#686980 - ray - stripes in pdf
#687367 - raph - customer needs patch for font outline bug in 8.10
#687394 - igor - Images not downsampled at some resolutions
#687401 - ray - Attempt to share (local) segments of path
#687496 - dan - transparency and CMYK support

More information about the gs-devel mailing list