[gs-devel] Re: bug status report - 09-11-2002

Wed Sep 11 19:41:45 PDT 2002

Wow.  Great job on the bugs Igor.  Thanks.

At 09:47 AM 9/11/02 -0600, Jack Moffitt wrote:
>SUMMARY
>-------
>Total sourceforge bugs  :       105
>New sourceforge bugs    :       15
>Closed sourceforge bugs :       15
>New potential bugs      :       00
>Open potential bugs     :       01
>Fixed potential bugs    :       00
>New customer bugs       :       06
>Open customer bugs      :       21
>Fixed customer bugs     :       04
>New top-ten bugs        :       01
>Open top-ten bugs       :       06
>Fixed top-ten bugs      :       00
>
>Closers                 :       igorm (9), rayjj (2), ghostgum (3), 
>jackiem (1)
>
>NEW BUGS
>--------
>
>#604884 - INTERNAL - Many undefined variables in build
>submitted by: lpd
>assign: ??
>
>In the current sources, 44 undefined references in the standard build
>are reported, including a reference to errors_h from within lib.mak, which
>is a forbidden reference from the graphics library to the postscript
>interpreter.
>
>#604886 - INTERNAL - Many compile warnings, some serious
>submitted by: lpd
>assign: ??
>
>The current CVS fileset (9:35 PM 9/4/2002 Pacific time)
>compiled with gcc 2.96 produces 141
>non-allowed compile warnings. ("Non-allowed" means
>excluding "cast discards qualifiers", which result from
>casting a const pointer to a non-const pointer, and
>also excluding all warnings from third-party
>libraries.) 10 of these are type mismatches,
>which any reasonable compiler would treat as errors
>rather than warnings.
>
>#605200 - INTERNAL - pdf_close generates many path segments
>submitted by: igorm
>assign: ??
>
>pdf_close generates huge number of path segments
>while computing FontBBox for embedded fonts. This can
>cause VMerror being not handled properly by the device
>finalization. A better method wanted for computing
>FontBBox without allocating memory for path segments.
>
>#605830 - INTERNAL - DSC parser shouldn't parse fonts
>submitted by: ghostgum
>assign: ??
>
>The DSC parser is used by the pdfwrite device to look
>for DSC comments to set document info and page
>orientation. The DSC parser is not activivated until the
>input document is read, which is correct. However, the
>DSC parser is being given all DSC lines from fonts
>which are loaded by the input document. This is
>incorrect. Ghostscript should not process DSC
>comments from fonts and other files run from the input
>document.
>
>#606777 - CUSTOMER - Transparency effect missing from PS
>customer: #190
>assigned: rayjj
>
>interregeps.prn (Created from AbobePS4 driver)
>Original document is Powerpoint, with
>transparency inside.
>It can be ripped using x11alpha device and has
>correct viewing, but not with x11 or a printer device.
>
>#607425 - USER - setdistillerparams resets page device
>assign: rayjj?
>
>The execution of setdistillerparams seems to reset the
>current page device at least partly. E.g. the values for keys
>like /BeginPage and /EndPage are reset to their default
>values.
>
>#607480 - CUSTOMER - 'Ghost' text appears
>customer: #411
>assign: igorm?
>
>When converting the enclosed PostScript job to PDF
>then the resulting PDF has an upside-down outlined
>copy of the actual text just below the actual text.
>This bug does not seem to exist in 7.04 or 7.21, but
>does in 7.22 and HEAD.
>
>#607483 - CUSTOMER - Text missing
>customer: #411
>assign: igorm?
>
>The PDF generated from the enclosed EPS is missing
>the white text. The actual font is not embedded into the EPS, but
>subtituted when using for example GSView. The PDF
>does not contain any text.
>
>#607778 - INTERNAL - Unimplemented pdf_separation_name
>submitted by: igorm
>assign: ??
>
>With Dan's recommendation pdf_separation_name needs to access
>pgs->client_procs.get_colorname_string. Perhaps pdf_separation_name is
>called from device virtual functions (particularly from begin_typed_image), in
>which 'pgs' isn't accessible. Instead they get gs_imager_state instance,
>which is a subclass of gs_state.
>Can we specialize (i.e. can we narrow) ALL those device virtual functions
>with the subclass ?
>
>#607819 - INTERNAL - missing a dependency
>submitted by: igorm
>assign: ??
>
>Today's build crashed on Windows 2000. Rebuild-all repairs.
>I guess a dependecy is missed in makefiles. Please check with Unix make.
>
>#607850 - CUSTOMER - stream length incorrect in PDF
>cutomer: #1060
>assign: rayjj?
>
>File pr293775att.pdf gets an error in Ghostscript: PDF
>format error: stream length incorrect. Acrobat Reader
>views this file fine.
>
>#607862 - TOPTEN - Jagged fonts in PDF
>customer: #150
>assign: rayjj?
>
>File 14006-ROTELBRO2.JUL.pdf renders nicely in Acrobat
>Reader, but has jagged fonts (customer describes them
>as 'corrupted'.
>
>OPEN POTENTIAL ISSUES
>---------------------
>#458780 - raph - Illustrator 9 transparency rangecheck
>
>OPEN CUSTOMER ISSUES
>--------------------
>#493348 - rayjj - Add image smoothing for better display
>#513905 - rayjj - cas.ps does not work with recent gs
>#567487 - rayjj - Incorrect garbage collection
>#567773 - jeong - CIE color errors
>#570772 - rayjj - garbled image in native color mode
>#578838 - raph - /rangecheck in --.discardtransparencygro
>#578865 - jeong - Colors missing in AI9 file.
>#592837 - raph - tranparency rendering different
>#596009 - raph - stream operator not terminated by EOL
>#600281 - rayjj - mem_mapped2_copy_mono() incorrect
>#601336 - giles - MacOS Native Font Support
>
>OPEN TOP-TEN ISSUES
>-------------------
>#458775 - raph - Illustrator 9 shading/masking bugs
>#535366 - raph - /VMerror in .pushpdf14devicefilter
>#535752 - raph - incorrect pdf rendering with pdf1.4
>#576591 - rayjj - footnotes rendered incorrectly
>#600288 - rayjj - /undefined in --setcolorspace--

______________________________________________
Miles Jones		*	Artifex Software Inc.
President		*	101 Lucas Valley Road #110
tel (415)492-9861	*	San Rafael, CA  94903
fax (415)492-9862	*	e-mail: miles at artifex.com
URL:  http://www.artifex.com/