[gs-devel] bug status report - 04-24-2003

Thu Apr 24 09:07:55 PDT 2003

SUMMARY
-------
Total bugs		:	142
New bugs		:	15
Closed bugs		:	02
New potential bugs	:	00
Open potential bugs	:	01
Fixed potential bugs	:	00
New customer bugs	:	02
Open customer bugs	:	37
Fixed customer bugs	:	01
New top-ten bugs	:	00
Open top-ten bugs	:	10
Fixed top-ten bugs	:	00

Closers 		:	ray (1), ralph (1)

NEW BUGS
--------

#686828 - CUSTOMER - Ghostscripts renders PDF file onto too large a
		     page
customer: #440
assign: igor?

Ghostscript (both 8.00 and the current CVS Head) renders this file
with extra whitespace next to the actual page data.  Acrobat Reader
4.0 renders this file without the whitespace.  Russel has suggested
the -dUseCropBox option.

#686829	- CUSTOMER - Error: /undefined in /.jbig2makeglobalctx
customer: #440
assign: ralph?

The attached PDF file uses jbig2 compression.  When reading the file
using Ghostscript (CVS Head) the following error is reported:
Error: /undefined in /.jbig2makeglobalctx
Adobe Acrobat 5.0 renders this file correctly (earlier versions of
Acrobat Reader rebort an unkown filter name).

#686830 - INTERNAL - a trapezoid fill glitch
submitted by: igor
assigned: igor

Rendering -r72 -sDEVICE=ppmraw 001-01.ps (as well as any other Genoa
test) with scanline algorithm disabled. An extra pixel appears at the
lower left conner of the outer black rectangle. The pixel is not
painted with the scanline algorithm. Looks as a glitch of the
trapezoid algorithm. Discovered while analysing regressions (actually
progressions).

#686831 - INTERNAL - A different raster from debug and release build
		     (455690.pdf) 
submitted by: igor 
assigned: igor 

Comparing the debug and release build from MSVC for Intel processor,
the character '7' is painted differently with -r72 -sDEVICE=ppmraw
455690.pdf . The effect is not visible with other tests. It may be a
consequence of floating point expression optimization, which is not
necessary in the font rendering code.

#686832 - INTERNAL - FONTPATH searching fails with /undefined in
		     /.filenamedirseparator
submitted by: russell
assign: ray?

bin\gswin32c -sFONTPATH="c:\psfonts" -c "/UnknownFont findfont"
AFPL Ghostscript CVS PRE-RELEASE 8.01 (2002-12-06)
Copyright (C) 2002 artofcode LLC, Benicia, CA.  All rights reserved.
This software comes with NO WARRANTY: see the file PUBLIC for details.
Scanning c:\psfonts for fonts...Error: /undefined in /.filenamedirseparator
Operand stack:
   UnknownFont   Font   UnknownFont   (UnknownFont)   0   false   (c:\\psfonts)
  (*)   (c:\\psfonts)
This works in GS 8.00.

#686833 - INTERNAL - Imprecise pattern phase
submitted by: igor
assign: ??

-dBATCH -dNOPAUSE -sDEVICE=ppmraw -sOutputFile=cur.ppm -r300 012-09.ps
gives different raster than CPSI.

#686834 - INTERNAL - A pattern missed while PDF interpretation
submitted by: igor
assign: ray?

The attached file renders fine in Acrobat Reader, but with GS a pattern fill 
doesn't occur. Alex has a partial fix.

#686835 - INTERNAL - rangecheck in .buildcmap
submitted by: alex
assign: ray?

The sample file fails with rangecheck in .buildcmap on GS 8.00 and
current CVS versions.  The file runs fine on Distiller 5.

#686836 - INTERNAL - -r doesn't work
submitted by: alex
assign: ray?

-r no longer works in gswin32.exe compiled from today's CVS .  When
resolution >= 120 dpi GS silently quits without any error messages.

#686837 - USER - PDF with subscript rendered wrong
assign: ray?

User created a PDF with chemical formula for water.  In AR5, one sees
the subscript 2 as expected.  In Ghostscript, both the 2 and the O are
subscripted.  I've confirmed the bug in CVS HEAD.

#686838 - INTERNAL - Ghostscript doesn't do anything
submitted by: igor

This looks like a dupe of #686836.  I suggest we close it as such. Or
close the other one as a dupe of this.

#686839 - USER - Patch to ignore pdf errors in outlines and links
assign: raph?

I found several pdf files with damaged or incorrect annotations
(outlines, links) pointing to non existing pages or even pointing to
nodes of incorrect type.  When using the pdf interpreter with
-dDOPDFMARKS, such errors cause an execution error in the pdf
interpreter The attached patch ignores such errors.  The assumption is
that it is better to produce botched pdfmarks than producing an error
message and be unable to process the file.  The patch is with respect
to gs7.05.  The problem also exists in gs8.

#686840 - CUSTOMER - incorrect letters at end of words using ASCII85
customer: #131
assign: ray?

The test file renders incorrect characters at the end of sequences on
page 2.  The customer claims it is usually shifted one letter of the
alphabet.  The customer thinks this may be an error in our ASCII85
decoder since this error does not occur when printing to a real
PostScript printer.

#686841 - CUSTOMER - bbox device issue
customer: #770
assign: ray?

Customer reports that converting the test file with
gswin32c.exe -r400 -sDEVICE=tiffg4 -r400 -sOutputFile=1.tif alliant.ps
produces output with proper page orientation, but improper image
rotation.  They are attempting to use the bbox values to determine the
orientation of the image, but the bbox values are flipped in relation
to the paper orientation.

#686842 - USER - Macintosh lacks a command line interpreter
assign: ralph?

User has provided a patch for a commandline interpreter under classic
MacOS.

OPEN POTENTIAL ISSUES
---------------------
#458780 - raph - Illustrator 9 transparency rangecheck

OPEN CUSTOMER ISSUES
--------------------
#493348 - ray - Add image smoothing for better display
#578838 - raph - /rangecheck in --.discardtransparencygro
#578865 - jeong - Colors missing in AI9 file.
#592837 - raph - tranparency rendering different
#601336 - ralph - MacOS Native Font Support
#607850 - ray - Problem processing CIDFontType2 -- undefined in get
#610356 - raph - Banding in the gradient
#617077 - ray - /limitcheck in --setpagedevice--
#626464 - ray - /undefined in Adobe_level2_AI5
#628110 - raph - conversion to tiff produces black page
#634036 - ray - additional lines in gradient.
#651644 - ray - mixed italic and nonitalic letters
#661171 - ray - blends incorrect at high resolutions
#667301 - raph - x11cymk crash
#674417 - ray - Problem with embedded device being removed by GC
#680301 - jack - pdfwrite font problems
#681469 - raph - image mask not working
#686768 - russell - problems with CHECK_INTERRUPTS on macos
#686770 - ralph - non-carbon code in gdevmacxf.c 
#686779 - ray - /ioerror in --token--
#686782 - raph - PDF 1.4 transparency incorrect and slow
#686788 - ray - line disappears at high resolutions
#686796 - dan - tiffg4 device and incorrect halftone frequency
#686798 - dan - /typecheck in --setcolorspace-- with
		-sDEVICE=x11cmyk
#686803 - ray - Font sheet problem
#686824	- ray - borders are incorrect in pdf file
#686825	- ray - Media size wrong in pdf2ps conversion

OPEN TOP-TEN ISSUES
-------------------
#458775 - raph - Illustrator 9 shading/masking bugs
#535366 - raph - /VMerror in .pushpdf14devicefilter
#535752 - raph - incorrect pdf rendering with pdf1.4
#600288 - dan - /undefined in --setcolorspace--
#611898 - raph - transparency rendering incorrect
#664484 - raph - transparency rendered incorrectly
#673124 - raph - /rangecheck in --.buildshading2--
#677573 - ray - 7.04 crashes in cs_next_packed_value
#686778 - ray - /undefined in --get-- in postscript file
#686789 - dan - gradients wrong after pdfwrite