Update pcl anomaly documentation for new problems discovered on the

Color Laserjet 4600.  Several problems previously identified as
"anomalies" have been addressed as well.

git-svn-id: http://svn.ghostscript.com/ghostscript/trunk@10789 a1074d23-0009-0410-80fe-cf8c14f379e6

1 files changed, 73 insertions, 358 deletions

diff --git a/pcl/Anomalies.txt b/pcl/Anomalies.txt
index 3a7406ffe..29c3ea46f 100644
--- a/pcl/Anomalies.txt
+++ b/pcl/Anomalies.txt
@@ -351,106 +351,48 @@ Difference in specific tests of the PCL 5c FTS
     proportionally spaced font. In the Artifex interpreter, this is a Times
     font, as it is for the Color LaserJet 5/5M, but not with the same spacing.
 
-65, 75, 85, 95, 103:
-
-    The raster in each of these tests is printed in a "simple" color space
-    (fixed palette). As described above, the colors in these color spaces
-    vary between geometric objects and rasters for the Color LaserJet 5/5M,
-    but not for the Artifex interpreter. Hence, the colors for the raster
-    will appear different in the two interpreters.
-
-140, 143:
-
-    The vertical bar character is from the bitmap lineprinter font on
-    the HP and from the Courier font on the Artifex interpreter, hence
-    the differences in the length of the bars.
-
-273, 274, 275, 276, 284, 285, 286:
-
-    Symbol set differences.
-
-310, 320:
-
-    These two panels are printed with the printer's "line printer" font.
-    On the HP DJ 1600C/CM and the CLJ 5/5M, this is a bitmap font that is
-    not scalable and is offered in only one size, hence its size does not
-    change with pitch. On the Artifex system, the scalable Courier font is
-    substituted for the line printer font. This font does change pitch as
-    requested in the tests.
-
-403, 412:
-
-    The Univers Condensed font provided with the Artifex PCL interpreter 
-    has all characters condensed. In the version of the same font provided
-    by HP, certain "universal" characters (characters that are independent
-    of font style), are not condensed.
-
-411 (et. al.):
-
-    The lower case 't' in the downloaded Park Avenue font contains a malformed
-    outline. This is handled differently by the Intellifont font scaler 
-    provided with the Artifex PCL interpreter than by the scaler provided with
-    HP PCL implementations.
-
-422:
-
-   The default vertical offset applied to floating underlines by the Artifex
-   PCL interpreter is not the same as that provided by HP PCL interpreters.
-
-441 - 443, 450 - 512:
-
-    The fonts provided with the Artifex PCL 5 interpreter are TrueType fonts,
-    which use the Unicode encoding scheme. The corresponding fonts in the
-    Color LaserJet 5/5M are Intellifont fonts, which use the MSL encoding
-    scheme. Thus, downloaded symbol sets that use the MSL coding scheme
-    can be used with the resident fonts of the Color LaserJet 5/5M, but not
-    with the resident fonts provided with the Artifex interpreter. Similarly,
-    downloaded symbol sets that use the Unicode encoding scheme can be used
-    with fonts provided with the Artifex interpreter, but not with those 
-    provided with the CLJ 5/5M.
-
-520, 521:
-
-    Symbol set discrepancies.
-
-680 - 682:
-
-    These rasters are generated in a simple color space, and thus have the color
-    anomalies described in the first section of this document.
-
-701:
-    The length of the bitarray for the third and fourth pair of raster
-    (magenta and yellow) differ from those on the HP CLJ 5/5M.
-
-    The output of the Artifex interpreter matches that of the HP DJ 1600C/CM.
-    The output of the CLJ 5/5M is contrary to the documentation, and likely
-    a limitation of the particular implementation due to the interaction
-    between rasters and the hardware dither method.
-
-714:
+292, 302:
+     
+     As indicated in the test, the font requested is for a bitmap
+     proportional font which Artifex renders and HP does not.  HP will
+     favor a scalable font over a perfectly matching bitmap font if
+     the resolution of the device is different than the bitmap font.
+     The file will print the same on each interpreter at 300 dpi.
+     Artifex does not support the resolution dependent HP behavior.
 
-    On the Color LaserJet 5/5M, two of the white gaps between the cyan lines
-    appear to be wider than the others (after the 48th line and 98th cyan 
-    lines). These larger gaps are not present in the input data, and are 
-    probably due to a resonance phenomenon between the raster and the 
-    resolution enhancement technology used by the CLJ 5/5M. This behavior 
-    is not reproduced in the Artifex PCL 5c interpreter.
 
-721:
+401, 411:
+     
+     Certainly a bug in the HP printer, each pair of lines should be
+     identical.  The only change between the lines is the foundry
+     identifier within the typeface which would not have any effect
+     when selecting by attribute.
 
-    This raster are generated in a simple color space, and thus has the color
-    anomalies described in the first section of this document.
+692, 693, 694, 695, 814, 976
+     
+     The first raster of in this panel has inverted colors.  The
+     Artifex output is correct.  In some cases HP appears to ignore
+     the settings of the white and black points when using the device
+     rgb long form command.  The following cid sequence extracted from
+     the test should produce an rgb color space with white point and
+     black point swapped (white = 0 and black = 255).
+
+begin 644 cid.pcl
+8&RIV,3A7```#"`@(`````````/\`_P#_
+`
+end
 
-722:
+     Changing the white point and black point in the command and
+     running the test on the HP has no effect on the result as it
+     should by the specification.
 
-    On the Color LaserJet 5/5M, the first of the shorter, magenta raster lines
-    appears to be thinner than the others. In the input data all of the lines
-    are the same width, and there is no apparent reason for the thinness of
-    the first line. The Artifex PCL 5c interpreter does not replicate this
-    anomaly.
+     Similarly, for the second raster in each of these panels the
+     Artifex interpreter follows the HP specification for setting the
+     white and black point in device color space and the HP does not.
+     Clearly these rasters are incorrect they are simply black on the
+     HP.
 
 750:
-
     In the third row, the row with the greatest reduction in scale, the first
     (thinnest) bar of the raster completely disappears on the Color LaserJet
     5/5M, but not with the Artifex PCL 5c interpreter.
@@ -494,6 +436,9 @@ Difference in specific tests of the PCL 5c FTS
     of the CIE L*a*b* color space in the Color LaserJet 5/5M, which renders
     colors with the value of L* = 0 as red rather than black.
 
+    NOTE: this issue is not apparent in the default configuration as
+    these device independent color spaces are replaced with sRGB
+
 811
 
     Each raster in this example is rendered in a different color space, in the
@@ -502,6 +447,9 @@ Difference in specific tests of the PCL 5c FTS
     rendering dictionary provided. For the fourth raster, the different
     implementations of the CIE L*a*b* color spaces is also evident.
 
+    NOTE: this issue is not apparent in the default configuration as
+    these device independent color spaces are replaced with sRGB
+
 813:
     The first two rasters are printed in RGB and CMY color spaces, with the
     white and black points reversed. Due to an error in the implementation
@@ -516,296 +464,67 @@ Difference in specific tests of the PCL 5c FTS
     hence their appearance is dependent on the color rendering dictionary
     provided.
 
-830, 831:
-
-    A number of differences between the Artifex PCL 5c interpreter and the
-    CLJ 5/5M are visible in these tests:
-
-    1.  The Artifex PCL 5c interpreter, in its default form, supports the
-        rendering methods with predictable halftones, and does not support
-        the rendering methods with unpredictable or device-specific halftones.
-        The CLJ 5/5M has the reverse arrangement. In particular, the Artifex
-        interpreter supports user downloaded dithers, which the CLJ 5/5M
-        does not.
-
-    2.  The Artifex PCL 5c interpreter determines which pixels are "white"
-        for purposes of transparency prior to the effects of the color mapping
-        implied by the rendering algorithm; the CLJ 5/5M makes this same
-        determination after such color mapping. This is most visible in the
-        case of render algorithm 2 (prominent horizontal black lines in both
-        tests), where the output from the Artifex interpreter is transparent
-        while that of the CLJ 5/5M is opaque.
-
-    3.  The Artifex PCL 5c interpreter, in common with most HP implementations
-        of PCL 5c, implements render algorithm 1 (snap to primaries) after
-        applying gamma correction. The CLJ 5/5M is exceptional in this area in
-        snapping the colors to primaries before applying gamma correction.
-        Hence, when using render algorithm 1, the output of the Artifex
-        interpreter (and most HP PCL 5c interpreters) is always a solid
-        primary color, while for the CLJ 5/5M it may be a halftoned color.
-        This difference is visible in the latter part of test 831.
-
-    4.  The final two columns of tests 830 and 831 represent a non-solid
-        raster object combined with a non-solid foreground, using logical
-        operation 252. Especially for the larger gamma values, the appearance
-        of the result is critically dependent on the nature of the halftones
-        used for the raster and the foreground. In the Artifex PCL 5c
-        interpreter, the halftones used cause these regions to become white
-        more rapidly than is the case for the CLJ 5/5M.
-
-833:
-
-    Each pair of lines in this example constitutes a pair of rasters. The 
-    data for each raster pair is the same, only resolution is changed.
-    A bug in the CLJ 5/5M interpreter, also evident in test 701, results in
-    the improper rendering of raster output at 100 and 150 dpi in some 
-    situations. The Artifex PCL 5c interpreter does not mimic this bug,
-    hence the different appearance of the output.
-
-853:
-
-    The modified render algorithm referred to in the fourth line of this test
-    is render algorithm 10, monochrome user defined dither. This render method
-    is supported by the Artifex PCL 5c interpreter (in its default form), 
-    but not by the CLJ 5/5M. Hence the different appearance. Note also that 
-    the final foreground for the fourth line is used to draw the text for 
-    the final, fifth line of the test, so its appearance is modified as well.
-
-860 - 865:
-
-    See discussion for test 830 and 831.
-
-866:
+    NOTE: this issue is not apparent in the default configuration as
+    these device independent color spaces are replaced with sRGB
 
-    The different appearance of this test between the Artifex PCL 5c
-    interpreter and the Color LaserJet 5/5M is due to the different halftones
-    used.
+951:
 
-871:
-
-    The final three rows of rectangles in this example are rendered in a device
-    independent color space, hence their appearance will depend on the nature
-    of the color rendering dictionary provided. In the default configuration,
-    these rectangles have a yellow hue where white would be expected.
-
-883:
-
-    The output from the Artifex PCL 5c interpreter matches the documented
-    behavior, and the behavior of HP implementations other than the CLJ 5/5M.
-    For reasons that are not clear, the CLJ 5/5M will print a faint outline
-    of "Text" for the rightmost instance of this string in the first line
-    of the test. Examination under magnification reveals that this outline
-    is smaller than the design resolution of the device, so this may be a
-    print engine artifact.
-
-956:
-
-    The rectangle on the right is printed with a transparent pattern. This
-    produces the expected output on the HP 1600C/CM and with the Artifex PCL
-    interpreter. It does not produce the expected output for the HP CLJ 5/5M,
-    apparently due to a bug.
-
-
-974:
-
-    See the discussion for test 830. This particular example also demonstrates
-    a situation in which different color mapping methods are being used for the
-    foreground and the raster being printed. For reasons that are not yet
-    clear, this (highly unusual) combination is not properly supported by the
-    graphic library underlying the Artifex PCL 5c interpreter.
+    The Artifex interpreter renders a white (invisible raster) where
+    the HP renders a green raster.  Combining the color in the
+    graphics state with the color specified by the raster predicts a
+    white box.  We aren't sure how to emulate the Color Laserjet bug.
 
 980, 981:
 
-    1.  The raster operation mechanism of the CLJ 5/5M is broken, and thus
-        cannot be used as a reliable comparison (see notes at the start of
-        this document).
-
-    2.  The implementation of opaque text for scalable fonts in the
-        Artifex PCL 5c interpreter is somewhat imperfect. The implementation
-        requires both inside and "outside" of a path to be filled, in
-        independent operations. The two filled regions do not join exactly,
-        hence a faint outline of the original character is usually visible.
-
-    3.  When compared to the documented behavior, the behavior of the Artifex
-        interpreter varies in some of the opaque source, transparent pattern
-        cases. We have not yet analyzed why this is the case, but do not
-        expect it to cause difficulty in practice.
-
-983:
-
-    The behavior of the Artifex PCL 5c interpreter matches the documented
-    behavior, and the behavior of HP devices other than the CLJ 5/5M. Raster
-    operations on the CLJ 5/5M are broken.
-
-1053:
-
-    The behavior of the Artifex PCL 5c interpreter matches the documented
-    behavior, and the behavior of HP devices other than the CLJ 5/5M. The
-    behavior of the CLJ 5/5M is apparently a bug.
-        
-1074 and 1084:
-
-     The artifex interpreter's implementation of fill type 3 and 4
-     (FT) differs slightly from HP.  Artifex gl/2 vector fills always
-     radiate away from the origin in plotter units space.  This can
-     cause slight discrepancies as evidenced in these two frames.
-
-1070:
-    The artifex interpreter always fills along +x and +y for hpgl/2
-    vector fills.  HP apparently fills in a direction related to the
-    direction of the graphical objects.  We believe our interpretation
-    is follows the HP spec.
-
-1272:
-
-   The gl/2 interpreter places scalable characters slightly to the
-   left, the discrepancy for a font scaled to 2" is about 1/16".
-   Attachment 1 below illustrates the problem.  The GL/2 Label is
-   drawn then we return to pcl maintaining the current "y" coordinate,
-   select the same font and print the same string.  The Artifex
-   interpreter produces the same output for each string, the expected
-   result.  HP does not.
-
-begin 664 gl2chbug.pcl
-M&T4;)3!"24X[4%<P.T94-"PV,RXU+#`[4$$P+#`[4E(U,#`P,"PU,#`P,#M0
-M5S`N,C4[1E0T+#4P."PP.U!!,"PP.U)2-3`P,#`L-3`P,#`[4%<P+C4[1E0T
-M+#$P,38L,#M003`L,#M24C4P,#`P+#4P,#`P.QLE,4))3CM34#$[25`[4%<P
-M+C`T.U-$-"PQ-#0L,BPQ+#<L-#$P,3M003`L-3`P,#M,0E!#3`-003`L,#L;
-5)3%!&RAS,30T5ALH<S%04$-,&T4*
-`
-end
-
-1402 (et. al.):
-
-    The filling of GL/2 polygon regions with lines in the Artifex PCL 5c
-    interpreter is not completely accurate. Line space and phasing are both
-    incorrect. These bugs are present in many other GL/2 tests as well.
-
-1422, 1423, 1424, 1445, 1446, 1462, 1463, 1801, 1803, 1811, 1817, 1894, 1895, 1941, 1946,
-1950, 1951, 1960, 1961, 1962, 1963, 1964, 1965, 1966, 2210 (page 1), 2220 (page 2)
-
-    Artifex GL/2 contains minor discrepancies with anchoring line
-    fills.  The problem is more evident in cases where scaling is
-    specified.  Low priority.
-
-1463, 1463:
-
-    The patterned line fills used in these two tests vary in appearance
-    between the CLJ 5/5M and the Artifex PCL 5c interpreter because
-    the latter does not properly set the phase of GL/2 line fills.
-
-1601, 1602:
-
-    The GL/2 label printing command ("LB") in the Artifex PCL 5c interpreter
-    does not set the initial position of text correctly when printing in the
-    vertical direction.
+    The raster operation mechanism of the is broken, and thus cannot
+    be used as a reliable comparison.  Graphic one renders red text
+    using raster operation 17 (not Destination or Source), the
+    destination background is white so the text is rendered black.
 
 1641:
 
-    The handling of horizontal tab in GL/2 differs between the Artifex PCL 5c
-    interpreter and the CLJ 5/5M.
-
-1642, 1650:
-
-    The position of the label origin in GL/2 varies between the Artifex PCL 5c
-    interpreter and the CLJ 5/5M.
-
-1680:
+    The handling of the stick font's horizontal tab in GL/2 differs
+    between the Artifex PCL 5c interpreter and the HP CLJ.
 
-    The character sets supported by the fonts used in Artifex PCL 5c
-    interpreter, in its default form, are not the same as those supported by
-    the fonts in the CLJ 5/5M.
-
-1752, 1753 (et. al.):
-
-    The lower-case 'a' in the GL/2 stick font provided with the Artifex PCL 5c
-    interpreter is a different style than that of the stick font in the
-    CLJ 5/5M. This difference is also evident in other tests.
-
-1860, 1862, 1863:
-
-    The position of the GL/2 label origin in the Artifex PCL 5c interpreter is
-    not accurate. This is particularly noticeable for arbitrary transformations
-    of text, as in the first two tests cited.
-
-2080:
-
-    The scaling of the rightmost imported image in this test is
-    different than the HP printer.  The artifex interpreter complies
-    with the HP PCLTRM.  The following test shows that simply entering
-    and leaving HPGL/2 <ESC>%OB<ESC>%0A has the side effect of
-    resetting HPGL/2 scaling.  This should not happen according to
-    PCLTRM 18-17.
-
-    <ESC>E
-    <ESC>*c720X
-    <ESC>*c0T
-    <ESC>*c8.3K
-(1) <ESC>%0B   /* enter hpgl2/2 */
-(2) <ESC>%0A   /* enter pcl */
-    <ESC>*c420Y
-    <ESC>*c0T
-    <ESC>%1BSP2;PW0;IN;SP6;SC0,100,0,100;PA50,50;SC;SP7;CI450;
-    <ESC>%1A
-    <ESC>E
-
-    the two commands (1) and (2) change the behavior the HP scaling
-    but not Artifex Scaling.  If (1) and (2) are removed both HP and
-    Artifex produce the same results.
-
-2101, 2102:
-
-    See comments for test 830 and 831.
-
-2106:
+1650:
+    
+    Label origin 9 differs slightly.
 
-    Logical operations are broken on the CLJ 5/5M. The behavior of the Artifex
-    PCL 5c interpreter follows the documentation.
+1772:
+  
+    Rotated character is clipped, low impact.
 
-2120, 2122:
+1950-1967:
 
-    HPGL/2 pixel placement is implemented using the postscript
-    operator setfilladjust.  The HP command definition is not consistant with
-    the behavior of the implementation and different devices (Color
-    Laserjet & 6MP) produce different results for each of these test files.
+    Our interpreter does not exactly emulate the HP when using HPGL/2
+    user defined fill type with adaptive lines.  The dash pattern
+    length should be a fuction of the the length of the segment
+    clipped to the graphic object being filled.  We don't emulate this
+    unusual filling style which is a relic from old HP pen plotters.
+    We don't see the filling style in real world applications and feel
+    our approximation is sufficient.
 
 2210 and 2220 (fourth page):
 
-    SI, SR discrenpancy.
-
-
-2322, 2323:
-
-    Because the Artifex PCL 5c interpreter determines which regions are "white"
-    for purposes of transparency before taking into account any color lookup
-    table modifications, the white arrows in the center of these images are 
-    considered transparent even in the center case (the inverting color 
-    lookup table). Hence, for the inverting color lookup table case, the 
-    central rectangle of the raster will be completely white.
+    SI, SR discrenpancy, obscure and low impact.
 
 2330:
 
     The CLJ 5/5M does not support the viewing illuminant command; the Artifex
     PCL interpreter does.
 
-2331:
-
-    These rasters are rendered in a device independent color space, hence
-    their appearance will be dependent on the color rendering dictionary
-    provided.
-
 2351, 2352:
 
     The driver configuration command linearly maps saturation and
     lightness arguments to pcl's gamma function.  It is expected that
     vendors will provide product specific code for this function.
+    This functionality appears to be broken on the Color Laserjet
+    4600.
 
 2412:
 
-    After downloading a color lookup table, the monochrome palettes at the
-    bottom of the first and second pages of this test have one additional
-    black entry when rendered with the Artifex PCL 5c interpreter, as compared
-    to the output of the CLJ 5/5M. This problem has not been analyzed.
+    We do not know why the text below the the palette is not printed
+    on the Color Laserjet.
 
 PCL5c ATS Anomalies
 ===================
@@ -931,7 +650,3 @@ in the X dimension.
 
 	   HP incorrectly renders 4 of the rasters on the page 
 Black triangles appear in a regular pattern across the raster  
-
-
-
-