| Current File : //usr/local/share/ghostscript/9.19/doc/pscet_status.txt |
# Report of Ghostscript's conformance against the Quality Logic CET suite
#
Test Status Comments
08-04- AOK This test is intentionally printer specific and is for information
only. Ghostscript completes all pages properly.
09-01-1 OK Fixed with gs_cet.ps initialization file
Default initial condition differs for currentblackgeneration and the
type and global of the default procedure differ.
09-01-2 OK Verified visually against Tek hardcopy. RJJ
matches CPSI.
09-02-1 OK Fixed rev 7102. RJJ
09-02-2 OK Colors corrected by rev 7172, fixed by rev. 7201
09-02-3 OK Corrected by 7152 and 7162. ADC
09-02-4 OK Corrected by 7152 and 7162. ADC
09-02-5 OK
09-03-1 OK Fixed rev 7102. RJJ
cpsi/09-03-1 is bogus. Acrobat 7 output obtained.
09-03-2 OK Fixed rev 7102. RJJ
cpsi/09-03-2 missing. Acrobat 7 output obtained.
09-03-3 OK Fixed by rev. 7202
09-03-4 OK Fixed rev 7102. RJJ
cpsi/09-03-4 missing. Acrobat 7 output obtained.
09-03-5 OK Verified visually against Tek hardcopy. RJJ
cpsi/09-03-5 missing. Acrobat 7 output obtained.
09-03-6 OK Verified visually against Tek hardcopy. RJJ
cpsi/09-03-6 missing. Acrobat 7 output obtained.
09-03-7 OK Verified visually against Tek hardcopy. RJJ
cpsi/09-03-7 missing. Acrobat 7 output obtained.
09-03-8 OK Fixed rev 7159.
cpsi/09-03-8 missing. Acrobat 7 output obtained.
09-03-9 OK
cpsi/09-03-9 missing. Acrobat 7 output obtained.
09-04-1 OK
09-04-2 OK
09-05-1 DIFF Device Dependent: GS, CPSI and Tek hardcopy all different.
probably can be made "closer" using gs_cet.ps assign Ray.
09-05-2 OK Device Dependent: GS matches Tek hardcopy. RJJ
09-05-3 OK Device Dependent: GS matches Tek hardcopy. RJJ
09-05-4 OK Device Dependent: GS matches Tek hardcopy. RJJ
09-05-5 OK Device Dependent: GS matches Tek hardcopy. RJJ
09-05-6 OK Device Dependent: GS matches Tek hardcopy after fixed by rev. 7203
(CPSI much different)
09-05-7 OK Device Dependent: GS matches Tek hardcopy. RJJ
09-05-8 OK
09-06-1 OK Fixed in rev. 7162. - Alex and RJJ
09-06-2 OK Fixed in rev. 7162. - Alex and RJJ
09-06-3 OK Fixed rev. 7159 and 7162. - Alex and RJJ
09-06-4 OK Fixed in rev. 7162. - Alex and RJJ
09-06-5 OK Fixed in rev. 7162. - Alex and RJJ
09-07-1 OK Rev. 7207 changes Ghostscript to match Tek.
The attributes of the arrays returned by currentcolortransfer
vary between PS interpreters. PLRM doesn't require any
particular attributes. Initial graphic state is created by an OEM,
not Adobe.
09-07-2 OK
09-07-3 OK
09-07-4 OK
09-08-1 OK Fixed in rev. 7207
09-08-2 OK
09-08-3 OK
09-09-1 OK Device Dependent: GS matches Tek hardcopy. CPSI always uses setflat == 1.
GS clamps setflat value minimum as 0.2 and maximum at 100.0 same as Tek.
09-09-2 OK Device Dependent: GS matches Tek hardcopy. CPSI always uses setflat == 1.
GS clamps setflat value minimum as 0.2 and maximum at 100.0 same as Tek.
09-09-3 OK
09-10-1 OK Fixed rev 7102. RJJ
09-10-2 OK "0" vs. "0.0" is fixed in rev. 7186.
09-10-3 OK Fixed rev 7102. RJJ
09-10-4 OK Colors corrected by rev 7172.
09-10-5 OK
09-11-1 AOK GS has different small numbers near zero (e-5) than CPSI Tek has 0.0
09-11-2 OK Corrected by revision 7162 (Alex). - ADC
09-11-3 OK Minor differences visually reviewed by RJJ
09-11-4 OK
09-11-5 AOK Implementation dependent: Tek and CPSI and GS all different. GS does
match 09-11-7 (Test B3) as the test comments state.
09-11-6 OK
09-11-7 AOK Implementation dependent: Tek and CPSI and GS all different. GS does
match 09-11-5 (Test B1) as the test comments state.
09-11-8 OK
09-11-9 OK
09-12-1 AOK Device Dependent. Every device has its own initial halftone.
Rev. 7208 makes currenthalftone return a read-only dictionary.
Adding a HalftoneType 9 as the Default entry to Resource/Halftone
path allows us to match Tek. Halftone isn't used with contone anyway.
09-12-2 AOK Device Dependent, Resolution dependent. Tek and CPSI differ
GS matches Tek hardcopy for SYNTAX 1 at 300dpi, SYNTAX 2
GS prints /Angle = 0, Tek shows /Angle = 0.0
09-12-3 OK GS matches Tek hardcopy. CPSI much different.
09-12-4 OK Resolution and Orientation dependent: GS matches Tek hardcopy at 300
dpi landscape.
09-12-5 OK GS at 300 dpi matches Tek hardcopy and checksum. Resolution dependent.
09-12-6 OK Resolution and Orientation dependent: GS matches Tek hardcopy at 300
dpi landscape.
09-12-7 OK Resolution and Orientation dependent: GS matches Tek hardcopy at 300
dpi landscape.
09-12-8 OK GS matches Tek hardcopy. CPSI much different.
09-12-9 OK
09-13-1 OK Fixed rev 7102. RJJ
09-13-2 OK Fixed again by rev 7215 after rev 7124 caused problem. Matches Tek
harcopy.
09-13-3 OK Fixed rev 7102. RJJ
09-13-4 OK
09-13-5 OK
09-14-1 OK Minor differences visually reviewed by RJJ
09-14-2 OK
09-15-1 OK Minor differences visually reviewed by RJJ
09-15-2 OK
09-16-1 OK
09-16-2 OK
09-16-3 OK
09-17-1 OK
09-17-2 OK
09-17-3 OK
09-18-1 OK
09-18-2 OK
09-19-1 OK Fixed rev 7102. RJJ
09-19-2 OK GS matches Tek hardcopy. CPSI gives 0.63 'B' value on first line of
third group -- Tek is probably correct. "0" vs. "0.0" is fixed in rev. 7186.
09-19-3 OK Fixed rev 7102. RJJ
09-19-4 OK
09-19-4 OK
09-19-5 OK
09-20-1 AOK Device Dependent: different default screen (OK)
GS matches Tek on the rest.
09-20-2 OK Device Dependent: Matches Tek hardcopy and checksum. RJJ
09-20-3 OK Device Dependent: Matches Tek hardcopy and checksum. RJJ
09-20-4 DIFF Device Dependent, Resolution dependent. Tek and CPSI differ
GS almost matches Tek hardcopy except "Actual Angle" when
AccurateScreens=true is 100.6 for GS (100.0 for Tek) and
GS shows "100 27" instead of "100.0 27.0" Assign: Ray
09-20-5 OK
09-21-1 OK
09-21-2 OK
09-21-3 OK
09-22-1 OK Rev. 7207 changes Ghostscript to match Tek.
See 09-07-01 for details.
09-22-2 OK
09-22-3 OK
09-22-4 OK
09-23-1 OK Fixed by gs_cet.ps RJJ
09-23-2 OK
09-24-1 OK
09-24-2 OK Fixed by rev 7617.
09-24-3 OK Fixed by r7123.
09-24-4 OK
09-25-1 OK
09-25-2 OK Fixed by rev 7617.
09-25-3 OK Fixed by r7123.
09-25-4 OK
09-26-1 OK Minor differences visually reviewed by RJJ
09-26-2 OK Fixed by rev 7617.
09-26-3 OK Fixed by r7123. Color corrected by rev 7172.
09-26-4 OK
09-27-1 AOK GS has different small numbers near zero (e-5) than CPSI Tek has 0.0
Device Dependent.
09-27-2 OK Minor differences visually reviewed by RJJ
09-27-3 OK Fixed by rev 7617.
09-27-4 OK Fixed by r7123. Color corrected by rev 7172.
09-27-5 OK
09-28-1 AOK Ghostscript is different from both Tek and PhotoPrint but
matches Distiller 5. Tek doesn't reset the dash array at
initgraphics. PhotoPrint redefines halftone operators as no-op,
which affects the test results. Fixed by rev. 7792. - Alex
09-28-2 OK Fixed as of rev. 7211
09-28-3 OK Fixed by rev 7617.
09-28-4 OK Imagemask -- Fixed by r7123. Tek and AD7 output differ.
09-28-5 OK Matches Tek hardcopy output. RJJ
09-29-1 AOK Device Dependent: GS has ProcessColorModel /DeviceRGB, Tek has non-standard
/DeviceCMY. GS matches Tek if BG, UCR set the same and ProcessColorModel is
forced to /DeviceCMY. CPSI shows ProcessColorModel /DeviceN. DeviceRGB
values fixed in rev. 7216
09-29-2 OK Fixed by rev 7617.
09-29-3 OK Imagemask -- Fixed by r7123. Tek and AD7 output differ.
09-29-4 OK
09-30-1 OK Colors corrected by rev 7172.
09-30-2 OK Error reports are fixed in rev. 7138. - Alex
09-30-3 OK Colors corrected by rev 7172. Fixed by rev 7617.
09-30-4 OK Fixed by r7123. Color corrected by rev 7172.
09-30-5 OK
09-31-1 OK Matches Tek hardcopy output. RJJ
cpsi/09-31-1 bogus. Acrobat 7 output obtained.
09-31-10 OK Matches Tek hardcopy output. RJJ
cpsi/09-31-10 missing. Acrobat 7 output obtained.
09-31-2 OK Matches Tek hardcopy output. RJJ
cpsi/09-31-2 missing. Acrobat 7 output obtained.
09-31-3 OK Matches Tek hardcopy output. RJJ
cpsi/09-31-3 missing. Acrobat 7 output obtained.
09-31-4 AOK GS Matches Acrobat 7 output. Tek output has small dark rectangles
that are not correct.
cpsi/09-31-4 missing. Acrobat 7 output obtained.
09-31-5 OK Matches Tek hardcopy output. RJJ
cpsi/09-31-5 missing. Acrobat 7 output obtained.
09-31-6 OK Fixed revs 7159 and 7187.
NOTE: AD7 gets limitcheck as pre-7159 GS and matches that GS output.
cpsi/09-31-6 missing. Acrobat 7 output obtained.
09-31-7 OK Matches Tek hardcopy output. RJJ
cpsi/09-31-7 missing. Acrobat 7 output obtained.
09-31-8 OK Fixed by rev 7617.
cpsi/09-31-8 missing. Acrobat 7 output obtained.
09-31-9 OK Matches Tek hardcopy output. RJJ
cpsi/09-31-9 missing. Acrobat 7 output obtained.
09-32-1 OK Verified visually against Tek hardcopy by Sabrina / RJJ
cpsi/09-32-1 missing. Acrobat 7 output obtained.
09-32-2 OK Error codes are adjusted to match the checksum. rev. 7167 - Alex
cpsi/09-32-2 missing. Acrobat 7 output obtained.
09-32-3 OK Fixed by rev 7617.
cpsi/09-32-3 missing. Acrobat 7 output obtained.
09-32-4 OK Matches Tek hardcopy output. RJJ
cpsi/09-32-4 missing. Acrobat 7 output obtained.
09-32-5 OK Verified visually against Tek hardcopy by Sabrina / RJJ
cpsi/09-32-5 missing. Acrobat 7 output obtained (but differs).
09-33-1 OK
09-33-2 OK Matches Tek hardcopy output. RJJ
09-33-3 OK Fixed by rev 7617.
09-33-4 OK Fixed by r7123. Color corrected by rev 7172.
09-33-5 OK Fixed rev. 7217
09-34-1 OK Matches Tek hardcopy output. RJJ
09-34-10 OK Matches Tek hardcopy output. RJJ
cpsi/09-34-10 missing
09-34-11 OK Matches Tek hardcopy output. RJJ
cpsi/09-34-11 missing
09-34-12 AOK GS has smooth gradients, Tek hardcopy has dark gray stripe up
middle of gradients. GS matches AD7
I found that once I set embedded an sRGB colorrendering
dictionary, which is what AD7 uses for RGB matching,
GS and AD7 were nearly a perfect match. There was only
one relatively small rectangle area pointed out by cmpi in a
single ramp ( the CIEDefg [ffff] row). It appears that the TEK
output is the one that is incorrect. TIM
cpsi/09-34-12 missing
09-34-13 OK Fixed revs 7159 and 7187.
cpsi/09-34-13 missing
09-34-14 OK Matches Tek hardcopy output. RJJ
cpsi/09-34-14 missing
09-34-15 AOK Mostly matches Tek hardcopy output. GS has smooth gradients, Tek
has dark gray stripe in middle of 8th row. GS matched AD7
--
In my tests with GS, AD5 as well as the AD7 output, none show
the dark gray stripe mentioned. It is my opinion that TEK is the
one that is failing. TIM
cpsi/09-34-15 missing
09-34-16 AOK GS missing some graphics seen on Tek hardcopy.
Probably RGB vs. CMYK device color - RJJ.
I do not understand Ray's comment. Using the ppmraw device, I
see all of the graphics at the right side of the page. Please
note that the appearance of this page depends upon the process
color model of the output device. The page uses spot colors.
Thus the appearance of the page depends upon how the device
treats spot colors. One of the spot color is called C1 and it
is white. The C1 spot color is printed on top of the lower four
graphics. Thus if the output device supports spot colors then
this results in a white appearance for this section of the page.
The GS tiffsep device (which supports spot colors) shows this
section of the page as being white. The tiffsep device also
produces a separation for the C1 spot color which shows the area
covered by this ink. Devices which do not support spot colors
use the tint transform function to convert the spot colors into
process colors. Logically the tint tranform functions should
produced the same appearance as the spot colors. However this is
not the case with this page. Thus devices which do not support
spot colors such as the ppmraw (process color model = DeviceRGB)
or the tiff32nc (device process color = DeviceCMYK only) use the
tint transform functions are producing different appearances.
The same effect can be seen with Distiller/Acrobat 7.0. The
presence of the spot colors can be examined with the 'Output
Preview' feature. Disabling the 'Overprint Preview' changes
the process color model of Acrobat to DeviceRGB. This forces
the use of the tint transform functions by Acrobat. The
'missing' graphics then appear.
cpsi/09-34-16 missing
09-34-17 OK Fixed by rev 7617.
cpsi/09-34-17 missing
09-34-18 OK Fixed by r7123.
cpsi/09-34-18 missing
09-34-19 OK Matches Tek hardcopy output. RJJ
cpsi/09-34-19 missing
09-34-2 AOK GS has Yellow graphic, Tek shows Red for 'SeparateUnBlue setcolorspace'
(right of "1.0" text). Probably RGB vs. CMYK device color
NOTE: AD7 and Tek hardcopy output differ in same areas. - RJJ
Distiller/Acrobat 7.0 differ on four graphics. These are the five below
the text that says "( /HIVAL 4095 def )". Distiller/Acrobat shows the
first four of these as a grayish shade. The fifth is a paler shade of
gray. GS shows these as black, gray, burgandy, brown, and red respectively.
- ADC.
From what I can see the yellow graphic that GS creates is
correct. The color space is aptly named "SeparateUnblue" (see
further below), since "un-blue" is yellow, this lead me to
believe yellow is what should be produced. Upon further digging
into the color space and putting some debug messages in the
tintTransform procedure, it also seems to be correct. From what
I can see the tint tintTransform procedure takes the blue value
and does some math on it to vary it in the range of 0.0 to 1.0.
It also places two ones on the stack just ahead of this blue
value so that all values passed to the /DeviceRGB are in the
range 1,1,0 to 1,1,1, which are various shades of yellow.
--
When I first did this I didn't know about the AD7 output... This
further verifies that the yellow is correct as AD7 matches GS
for this part of the job.
--
There is a second unresolved issue below the section that says
"/HIVAL 4095 def". I believe again that the four non-matching
graphics produced by GS are correct. For example...
--
The burgandy and brown are produced by the following code, respectively:
/HIVAL 4095 def
...
(gs Indexed_RGB_Proc setcolorspace 2048 setcolor Q)
(gs Indexed_RGB_Proc setcolorspace 3072 setcolor Q)
With the following colorspace definition:
/Indexed_RGB_Proc
{
[
/Indexed
/DeviceRGB
HIVAL
{
dup
HIVAL
div
exch
HIVAL
2
idiv
dup
2
index
exch
mod
exch
div
exch
HIVAL
3
idiv
dup
3
-1
roll
exch
mod
exch
div
}
]
}
def
--
Basically this space produces as output RGB values where the B's
that are split into 3 ranges by B_out= ( Value_in mod (HIVAL/3)
) / (HIVAL/3), G's that are split into 2 ranges by G_out =
(Value_in mod (HIVAL/2) ) / (HIVAL/ 2) and the R's are in one
range by R_out = Value_in / HIVAL. So a input value of 1/2 of
the input range (determined by HIVAL) produces an output of
roughly R=0.5, G=0, B=0.5. An input value of .75 of the input
range produces and output of roughly R=.74,G=.5,B=.25
--
I modified the colorspace to dump the stack at the end of its
number crunching tintTransform procedure. I printed out the
values for the burgandy and brown graphic sections and here is
what I got back.
--
2048 setcolor
0.500366 // B
0.00048852 // G
0.500122 // R
Which is the burgandy color mentioned.
--
3072 setcolor
0.250549 // B
0.500733 // G
0.750183 // R
Which is the brown color mentioned.
This verified to me that GS is doing the right thing. TIM
cps/09-34-2 missing
09-34-3 AOK GS RGB has missing graphic, Tek shows Cyan for 'SeparateUncyan setcolorspace'
(right of "1.0" text). GS matches Tek if Cyan, Magenta, Yellow and Black
Separation colorspaces are known, as with the 'tiffsep' device. A printer
that is CMYK will usually have the Separations available for overprint
effects. RJJ
-
NOTE: AD7 and Tek hardcopy output differ in same areas. RJJ
GS produces output matching Tektronix if 'tiffsep' device since this
device does not use the tint transform on /Cyan separation used in the
SeparationUncyan section.
-
The 'SeparateUncyan' issue is another example of a separation color space
which (intentionally ?) has a tint transform function which produces different
colors when it is used. Thus the output depends upon whether or not the tint
transform function is used. This can be seen by disabling the 'Overprint
Preview' feature in Acrobat 7.0. GS produces results which are consistent with
Acrobat for both cases. (this is expected). ADC
cpsi/09-34-3 missing
09-34-4 OK Matches Tek hardcopy output. RJJ
cpsi/09-34-4 missing
09-34-5 OK Matches Tek hardcopy output. RJJ
cpsi/09-34-5 missing
09-34-6 AOK GS matches graphics colors, but the code equivalent to
/Pattern setcolorspace 16 16 true matrix { (jihad) } imagemask
leaves imagemask operands on the stack on Adobe interpreters.
This is an obvious bug that we don't want in Ghostscript.
cpsi/09-34-6 missing
09-34-7 AOK GS has different 'currentcolor' results and different color graphics
Probably RGB vs. CMYK device color
NOTE: AD7 and Tek hardcopy output differ in same areas. - RJJ
There are two issues with this page. The first issue is the colors of
of the various graphics on the page. The page has spot colors. It
also has tint transform functions which do not provide logically
consistent results. Thus the appearance of the graphics depend upon
if the tint transform functions are being used. The results with GS
match the results from Acrobat 7.0 as expected.
The second issue (which is a problem) is that the currentcolor value
for the CIEBasedDEFG differs. The graphic for the same line is also
missing (which is probably related.) - ADC
--
The differences were releated to the colorrendering dictionary
in use. When I embedded an sRGB CRD the output matched AD5/AD7. TIM
cpsi/09-34-7 missing
09-34-8 OK Matches Tek hardcopy output. RJJ
cpsi/09-34-8 missing
09-34-9 OK Matches Tek hardcopy output. RJJ
cpsi/09-34-9 missing
09-35-1 OK Matches Tek hardcopy output. RJJ
09-35-2 OK Color difference (cmpi -c25 is OK, -c24 detects all pixels differ),
otherwise fixed by rev 7617.
Presumed to match Tek (different inks, etc.) since this test uses
'setcolortransfer' which is ignored by CPSI. RJJ
09-35-3 OK Fixed by r7123.
09-35-4 OK Minor color difference.
09-36-1 OK Minor differences visually reviewed by RJJ
09-36-2 OK Minor differences visually reviewed by RJJ
09-36-3 OK Fixed by rev. 7226
09-36-4 NF Dash pattern differences with GS.
Analyzed by Igor.
-
GS rounds coordinates to 1/256th of a pixel.
Due to that, when a curve is converted into a polygon,
its vertices are rounded, and therefore the length
of the polygon becomes few bigger than the ideal polygon.
Then applying the dash, it covers a smaller part
of the polygon than with CPSI.
-
One possible solution is :
Add floating coordinates to path segments.
The floating coordinates duplicate fixed ones,
and used only for computing the segment length.
The old fixed coordinated are used for anything else
as they do now.
09-36-5 DIFF In "Technical Paths", Ghostscript differs a lot, combination
of 09-24-2 linecap linejoin issues and dash issues. Defer
this until after 09-24-2 is addressed.
Need CPSI for investigations.
Assign: Ray
09-36-6 OK Fixed by r7123.
09-36-7 OK Fixed as of r7500.
09-37-1 AOK GS similar to Tek, but the value at which polygon edges changes from
16 to 8 differs (GS at 53, Tek at 19). Also Tek changes from 'smooth'
to 16 at 5, GS at 13. CPSI fixed to implement setflat changes from
smooth to 16 sides at 9 (GS at 13), CPSI never has less than 16 sides.
Resolution dependent (spec is 'output device pixels') and PLRM 3rd ed
states "If the flatness parameter is large enough to cause visible
straight line segments to appear, the result is unpredictable."
When GS is run at 300 dpi (same resolution as Tek), the transition
from smooth to 16 sides is at 7 (vs. 5 on the Tek) and from 16 to 8
sides at 27 (vs. 19 on the Tek).
-
Since Tek and CPSI differ, and GS generally uses more line segments,
the GS output conforms and is superior to the Tek output.
09-37-10 OK GS matches Tek and CPSI output now matches Tek (12/04/06).
Imagemask turkey fixed by r7123.
09-37-11 OK Matches Tek hardcopy output. RJJ
09-37-2 OK Matches Tek hardcopy output. RJJ
09-37-3 OK Matches Tek hardcopy output. RJJ
09-37-4 AOK GS similar to Tek, but the value at which polygon edges changes to
very coarse differs (GS at 73, Tek at 23). Also Tek changes from 'smooth'
to bumpy at 5, GS at 19. When GS is run at 300 dpi, the transitions
from smooth to bumpy is at 11, and from bumpy to very coarse at 39.
Resolution dependent (spec is 'output device pixels'). CPSI fixed to
implement setflat transitions to bumpy at 13 and to coarse at 33.
-
Since Tek and CPSI differ, and GS generally uses more line segments,
the GS output conforms and is superior to the Tek output.
09-37-5 OK Matches Tek hardcopy output. RJJ
09-37-6 OK Very small fill differences. RJJ
09-37-7 OK Very small fill differences. RJJ (CPSI ignores setflat, GS
similar to Tek hardcopy)
09-37-8 DIFF GS similar to Tek, but GS differs in the exact polygon shapes.
Resolution dependent (spec is 'output device pixels'). GS at 300
dpi has less edges on large outline in top two images. CPSI (at
600 dpi) also smoother than GS on top two outlines.
assign: Ray
09-37-9 DIFF "Stroked_Objects" and "Filled_Objects" differ slightly with
number of polygonal edges. Tek shows strokepath of "Technical
Paths" IS subject to setflat -- GS has smoother round cap/join.
Resolution dependent (spec is 'output device pixels'). CPSI fixed
to implement setflat matches Tek ouptut for "Technical paths".
Assign: Ray
09-38-1 OK Minor differences visually reviewed by RJJ
09-38-2 OK Fixed by rev 7617.
09-38-3 OK Fixed by r7123.
09-38-4 OK
09-39-1 OK Matches Tek hardcopy output. RJJ
09-39-2 OK matches CPSI
09-39-3 OK matches CPSI. Imagemask fixed by r7123.
09-39-4 OK
09-40-1 AOK Matches Tek hardcopy output. RJJ (CPSI ignores settransfer)
GS is contone, Tek is dithered
09-40-10 OK
09-40-11 OK
09-40-12 OK
09-40-13 OK
09-40-14 OK
09-40-15 OK
09-40-16 OK
09-40-17 OK
09-40-18 OK
09-40-19 OK Matches Tek hardcopy output. RJJ (CPSI ignores settransfer)
09-40-2 AOK GS is contone, Tek is dithered, otherwise GS matches Tek.
CPSI ignores transfer function.
09-40-20 OK Minor differences visually reviewed by RJJ
09-40-21 AOK GS and Tek give 'rangecheck' for last three "Test I". GS
matches Tek except GS is contone, Tek is dithered.
09-40-22 OK
09-40-23 OK Fixed by rev 7617.
09-40-24 OK Fixed by r7123. Color corrected by rev 7172.
09-40-25 OK
09-40-3 OK
09-40-4 OK
09-40-5 OK
09-40-6 OK
09-40-7 OK
09-40-8 OK GS matches Tek since rev. 7168 - Alex
09-40-9 OK
09-41-1 OK
09-41-2 OK Fixed by rev 7617.
09-41-3 OK Fixed by r7123.
09-41-4 OK Fixed by r7124.
09-42-1 OK
09-42-2 OK
09-42-3 OK Minor differences visually reviewed by RJJ
09-42-4 OK
09-42-5 OK Fixed by rev 7617.
09-42-6 OK Fixed by r7123.
09-42-7 OK
09-43-1 OK
09-43-2 OK
09-43-3 OK Minor differences visually reviewed by RJJ
09-43-4 OK
09-43-5 OK Fixed by rev 7617.
09-43-6 OK Fixed by r7123.
09-43-7 OK
09-44-1 OK Minor differences visually reviewed by RJJ
09-44-2 OK
09-44-3 OK Minor differences visually reviewed by RJJ
09-44-4 OK
09-44-5 OK Fixed by rev 7617.
09-44-6 OK Fixed by r7123.
09-44-7 OK
09-45-1 OK
09-45-2 OK
09-45-3 OK
09-45-4 OK
09-45-5 DIFF Partly fixed by rev 7617. Remaining diff is width of miter
extension in "Stroked Objects". Assign: Ray
09-45-6 OK Fixed by r7123.
09-45-7 OK
09-46-1 OK Minor differences visually reviewed by RJJ
09-46-2 OK
09-47A-1 OK Matches Tek hardcopy output. RJJ
09-47A-2 OK Matches Tek hardcopy output. Minor color differences. RJJ
09-47A-3 OK Minor differences visually reviewed by RJJ
09-47A-4 OK Matches Tek hardcopy output. RJJ
09-47A-5 OK Matches Tek hardcopy output. RJJ
09-47A-6 OK Matches Tek hardcopy output. RJJ
09-47A-7 OK Matches Tek hardcopy output. RJJ
09-47A-8 DIFF Resolution dependent: GS 'pattern' has double line joined, Tek
and CPSI do not although Tek and CPSI differ with pattern phase.
Assign: Ray
09-47A-9 OK Fixed by rev 7169.
09-47B-1 DIFF GS has different pattern phase. Tek and CPSI appear to be the same.
Assign: Ray
09-47B-10 DIFF Test01 : aok
CPSI problem. The tensor patch has corner colors like this :
00 55
ff aa
Thus it must have a horizontal line of a constant color 80
near the middle. The line is observable with GS and it is not with CPSI.
Analyzed by Igor.
-
Test09 - a color mapping problem.
Analyzed by Igor.
CIEBasedDEF colors are given with a sampled
(2 args, 2 results) function like this :
000000 ff0000
00ff00 ffffff
With GS the brightness of magenta is asame as cyan.
CPSI paints a lighter magenta.
I guess something is wrong with WhitePoint in GS,
because an assimmetry red|blue must come from it.
The assymmetry is observing with CPSI and not with GS.
Assign to Tim
-
Test10: AOK
Match when using common CRD.
-
Test11: AOK
Match when using common CRD.
09-47B-11 DIFF Test02, Test03 : aok. Same as 09-47B-10 Test11.
-
Test04: aok
Analyzed by Igor.
Likely CPSI problem.
The test is buggy - insufficient data for function.
It needs 16 samples, but 12 are only provided.
GS adds zeros at end.
000000ff 000000ff
00ffffff 00000000
Thus the right end is Magenta+Yellow+Black =
(1-green + 1-blue)*(1-black) =
(1 + red)*(1-black).
It is observing with GS.
CPSI likely due to occasional data added at stream end.
See also Test11.
-
Test05: AOK
Analyzed by Igor.
A bad color CMYK=[1 1 1 1] in the test.
GPSI and GS differently convert [v v v v] CMYK colors to RGB.
Matches fine in 7736 with CPSI_mode = true.
Tim
-
Test06: AOK
Analyzed by Igor.
A bad color CMYK=[0.7 0.7 0.7 0.7] in the test.
GPSI and GS differently convert [v v v v] CMYK colors to RGB.
Matches fine in 7736 with CPSI_mode = true.
Tim
-
Test11: AOK
This is the same problem as Test04. Bug in test.
Insufficient data for function.
/CIEBasedDEFG as /DeviceCMYK in Test04 needs 16 samples,
and only 12 are provided.
Tim
-
Test12:
CIEBasedDEFG shading does not match CPSI.
Assign to Igor
09-47B-12 AOK GS has wrong color Test06 GS not smooth Test07, Test08.
NOTE: Tek gets rangecheck error on Test03 Test10, CPSI and GS do not.
.
According to my reading of the spec, test6 should rangecheck,
but CPSI returns garbage data instead. r7130 improves match to
Tek output, but ultimately rendering of this page will be
device dependent, hence aok. - Raph
09-47B-2 DIFF GS has different pattern phase. Tek and CPSI appear to be the same.
Assign: Ray
09-47B-3 DIFF GS has different pattern phase. Tek and CPSI appear to be the same.
Assign: Ray
09-47B-4 OK
09-47B-5 OK Minor color difference, reviewed visually. RJJ
09-47B-6 OK Fixed by rev 7169.
09-47B-7 AOK Test01, Test02 :
Briefly analyzed by Igor.
Resolved by using common CRD. Tim
-
Test10: aok
CPSI problem : a non-monotonic color space is rendered as monotonic.
09-47B-8 AOK Test01, Test02, Test03, Test05, Test08, Test09, Test10 : aok
Analized by Igor.
CPSI problem : a non-monotonic color is rendered as monotonic.
-
Test04, Test11: aok
Analized by Igor.
CPSI problem same as 09-47B-10 Test01.
09-47B-9 AOK Test03, Test04, test05, Test06, Test07, Test10, Test11: aok
Analized by Igor.
CPSI problem : a non-monotonic color is rendered as monotonic.
-
Test08 : aok
Analized by Igor.
CPSI has a worse smoothness.
-
Test12 : aok
Analized by Igor.
CPSI problem same as 09-47B-10 Test01.
09-47C-1 OK Fixed by rev 7221.
09-47C-2 DIFF Rev. 7740 adds 'null' to the message.
Adobe shows: 'F --show-- rangecheck null'
At 144 dpi GS shows 'F --.type1execchar-- rangecheck null'.
At 72 dpi GS shows the same message as Adobe. assign: Peter
09-47D-1 OK Fixed by rev 7169.
09-47D-2 OK Fixed by rev. 7149.
09-47D-3 OK Fixed by rev 7169.
09-47D-4 OK Fixed by rev 7169.
09-47D-5 OK Fixed by rev 7169.
09-47D-6 OK Fixed by rev. 7149.
09-47E-1 OK Fixed by rev 7169.
09-47E-2 OK Fixed by rev 7149.
09-47E-3 OK Fixed by rev 7169.
09-47F-1 OK Fixed by rev 7169.
09-47G-1 AOK Test06: ok.
rev. 7176 now properly sets errorinfo.
-
Test07: aok
Analyzed by Igor.
The test is buggy : insufficient stream data,
the effect does not correspond to the comment in the test.
Due to insufficient data GS adds zeros at end,
and the function becomes like this :
000000 ff000000 ff00ff ffff0000
000000 00000000 000000 00000000
The DeviceN maps it to DeviceRGB with dropping the last 4 components,
so we've got corner colors :
black magenta
black black
That's what we observe with GS.
There is a comment in gsdsrc.c ln 47 :
/* Access data from a string or a byte object. */
/* Does check bounds, and returns 0 data oob. Spec calls for rangecheck,
but CPSI implementation silently gives (bogus) data. */
Tek paints black magenta
black red - likely an occasional data at stream end.
-
Test10: aok
Analyzed by Igor.
A bug in the test : excessive stream data,
the effect does not correspond to the comment in the test.
The real effect is a linear sampled function like this :
Cyan: 00 00
00 ff
Magenta: 00 00
00 ff
Yellow: 00 00
00 ff
Black: 00 00
00 ff
Thus it paints bad CMYK colors [v v v v] for various v.
Same as 09-47B-11 Test05.
I believe that [1 1 1 1] is very black, right ?
Thus CPSI looks wrong.
GS looks close to Tek.
09-47G-2 AOK Test01: ok
Fixed with rev 7199.
-
Test10: aok
Analyzed by Igor.
The linear sampled function is :
000000 ff0000
00ff00 ffffff
The function's domain by the 2nd argument
is twice bigger than the shading domain.
Therefore the function linear is cropped to this :
007700 ff7777
00ff00 ffffff
With CPSI it looks like this (wrong) :
000000 ff0000
00ff00 ffff77
With GS it looks like this (correct) :
007700 FF7777
00ff00 777777
CPSI appears wrong, GS appears correct.
-
Test11: aok
The function domain is cropped by both coordinates.
First convert to 3x3 :
000000 770000 ff0000
007700 777733 ff7777
00ff00 77ff77 ffffff
Then crop with the "positive" quadrant :
777733 ff7777
77ff77 ffffff
It looks close to what GS paints.
CPSI paints something like this :
770000 ff0000
ffff77 7777ff
CPSI appears wrong, GS appears correct.
09-47G-3 AOK Test01: aok
The function 000000 ff0000 black red
00ff00 ffffff green white
It is truncated with BBox/Matrix = 0.909, 0.933
So by theory we get:
000000 e70000
e60000 d1d1d1
It looks so with GS.
Don't know why CPSI misses green.
-
Test02 : Same as 09-47G-2 Test11.
09-47H-1 OK Fixed by rev 7149.
09-47H-2 OK Fixed by rev 7149.
09-47H-3 OK Partially fixed by rev. 7169, finished by rev. 7176.
09-47H-4 OK Fixed by rev 7206
09-47I-1 OK Fixed by rev 7149.
09-47I-2 OK Fixed by rev 7175.
09-47I-3 OK Test01, Test02, Test03, Test04, Test05, Test06, Test07: ok
Fixed by rev 7175.
-
Test11: OK. Looks fine in rev 7718. Tim.
09-47I-4 OK Fixed as of rev 7211.
09-47I-5 OK Fixed by rev 7175.
09-47I-6 OK Fixed by rev 7175.
09-47I-7 OK Test02, Test03, Test04: ok Fixed by rev 7175.
Test11: ok Fixed with rev 7179.
09-47I-8 OK Test02, Test04: ok Fixed with rev 7179.
-
Test07, Test11: ok Fixed by rev 7175.
-
Test06, Test08: ok Fixed with rev 7218.
- -
Test12: ok Fixed by rev 7175.
09-47J-1 DIFF Graphic differences in Test01 (partly drawn) and Test12
(green line across the shading). Assign: Igor
09-47J-10 OK
09-47J-2 OK
09-47J-3 AOK Test01 CPSI misses an imagemask with a shading color.
09-47J-4 OK
09-47J-5 OK
09-47J-6 OK
09-47J-7 OK
09-47J-8 AOK Test01 CPSI misses an imagemask with a shading color.
09-47J-9 OK
09-47K-1 AOK Test01 CPSI yields a rangecheck while rendering
an imagemask or a character with a shading color.
and misses the rest of graphics.
I don't see any legal reason for rangecheck.
09-47K-2 OK
09-47K-3 OK
09-47K-4 OK
09-47K-5 OK
09-47K-6 OK
09-47K-7 AOK Test01 CPSI yields a rangecheck while rendering
an imagemask or a character with a shading color.
I don't see any legal reason for rangecheck.
Graphics appearance matches CPSI visually.
09-47K-8 OK
09-47L-1 AOK Test01 CPSI misses the character and imagemask with
a shading color.
09-47L-2 OK
09-47L-3 OK
09-47L-4 OK
09-47L-5 DIFF Test01 CPSI misses the character and imagemask with
a shading color.
Test07, Test08 -- GS missing 'wedges' in center region
of shading.
09-47M-1 AOK Test01: The corner colors are : black red
green white
GS paints entire shading.
CPSI misses for imagemask and character. AD7 paints character,
but misses imagemask.
Analyzed by Igor.
09-47M-2 OK Fixed by rev 7149.
09-47M-3 OK Fixed by rev 7169.
09-47M-4 AOK Test01 : ok, similar to 09-47M-1. (CPSI misses imagemask and
character, AD7 misses imagemask).
-
Test04: ok - Fixed with rev 7210.
Note: AD7 gets "drawing error" on the PDF it creates from this
as it is drawing Test04. GS is able to display entire Page and
output matches GS rendering of 09-47M.PS. Bug in Acrobat 7
Reader!
-
Test09: ok - Fixed with rev 7210.
-
Test10, Test11, Test12: ok - Fixed by rev 7206
09-47M-5 AOK GS has smoother shadings (CPSI has triangle 'glitches'
with self-intersectint tensor: Test05, 06
Test10-12). NOTE: Tek shapes and colors differ greatly.
09-47N-1 DIFF A different flattening. Assign: Ray
09-47N-2 OK Fixed by r7123.
09-47N-3 AOK Really minor phase difference in pattern. RJJ
09-48-1 OK Fixed in rev. 7138 - Alex
09-48-2 OK Fixed by rev 7617.
09-48-3 OK Fixed by r7123.
09-48-4 OK
09-49A-1 OK
09-49B-1 OK
09-49B-2 AOK Wrong value of AccurateScreens is fixed by rev. 7741.
CPSI-based PhotoPrint product ignores all attempts to change
the halftone and cannot be used as a reference.
GS is close to Tek Special Test B except for resolution
dependent numbers for ActualFrequency. Test C ActualFrequency is
also Device Dependent. - Alex
09-49C-1 OK Matches Tek hardcopy and checksum. RJJ (CPSI differs because it
ignores setscreen).
09-49D-1 OK Fixed by rev 7617.
09-49D-2 OK Fixed by r7123. Color corrected by rev 7172.
09-49D-3 OK Fixed by r7450
09-50-1 DIFF GS has different line spacing. CPSI and Tek agree. GS gets same
results for "false setstrokeadjust" "true setstrokeadjust".
Assign: Ray
09-50-2 OK
09-50-3 DIFF GS missing radial lines filling arcs. Tek and CPSI agree. May
be related to 09-24-2. Assign: Ray
09-50-4 OK Fixed by r7123.
09-50-5 OK
09-51-1 OK Matches Tek hardcopy. RJJ (CPSI ignores settransfer).
09-51-2 OK Fiexd by rev 7617. GS matches colors of Tek hardcopy, CPSI
shows MUCH different colors.
09-51-3 OK Fixed by r7123. GS matches colors of Tek hardcopy, CPSI
shows MUCH different colors.
09-51-4 OK Minor color difference (looks black to me). USE_ADOBE_CMYK_RGB issue
assign: Ray
09-52-1 AOK Device Dependent: GS has default ProcessColorModel /DeviceRGB
CPSI has /DeviceN, Tek has /DeviceCMY (not a standard one??)
09-52-2 OK Fixed by rev 7617.
09-52-3 OK Fixed by r7123.
09-52-4 OK Fixed by gs_cet.ps RJJ
09-53-1 OK Fixed by r7155 - Raph
09-53-2 OK Fixed by r7155 - Raph
09-53-3 OK Minor differences visually reviewed by RJJ
09-53-4 OK GS has different numbers (off by 1.0 or 2.0 out of ~120.0). Related
to cliprestore.
Fixed by r7155 - Raph
09-53-5 OK Fixed by rev 7617.
09-53-6 OK Fixed by r7123. Color corrected by rev 7172.
09-53-7 OK
09-54-1 AOK GS has > 1000 limit on clipsave. Tek and CPSI limit at 34.
Device Dependent.
09-54-2 OK Fixed in r7155 - Raph
09-54-3 OK Fixed in r7155 - Raph
09-54-4 OK Fixed in r7155 - Raph
09-54-5 OK Fixed in r7155 - Raph
09-54-6 OK Fixed by rev 7617.
assign: Ray
09-54-7 OK Fixed by r7123. Color corrected by rev 7172.
09-54-8 OK Matches Tek hardcopy and CPSI. RJJ
09-55-1 AOK Implementation dependent: GS allows small smoothness values,
Tek clamps lower value to 0.005, CPSI to 0.002.
09-55-2 OK
09-56-1 AOK Device Dependent: Tek and Adobe show '/findcolorrendering'
instead of '--findcolorrendering--' with errors. I've checked
in a fix for the setcolorrendering issue. /GenericImagesetter
and /AppleMonitor are CRD's provided by Adobe to OEM as
examples. They must be removed from the final product but
Scanvec-Amiable forgot.
-
The checksum depends on the order of resource enumeration,
which depends on the order of dictionary enumeration and
directory enumeration for external halftones.
09-56-2 OK revisin 7234 (peter)
09-56-3 OK Ghostscript matches the checksum for Tek360 version of the test.
PhotoPrint has extra '1' in GRAY33 that tests that whether the
change in colorspace affects CRD. Surprisingly, it does.
Tek scan and Distillier 5 have '1' in MEMO 03 and 04, i.e.
have different default CRD and a CRD returned by /Saturation
findcolorrendering. CRD selection is implementation dependent
and is normally left to OEM.
09-57-1 AOK Fixed by rev 7169.; GS has a better smoothing.
09-57-10 OK Fixed by rev 7169.
NOTE: Tek is MUCH different (wrong??)
09-57-11 OK Fixed by rev 7169.
NOTE: Tek is all blank graphics (wrong??)
09-57-12 OK Matches Tek hardcopy. RJJ (CPSI differs)
09-57-13 AOK Test 09-12: With bad smoothness CPSI shadings are worse.
Analyzed by Igor.
09-57-14 OK Fixed by rev 7617.
09-57-15 OK Fixed by r7123.
09-57-16 OK
09-57-2 AOK Fixed by rev 7169.; GS has a better smoothing.
09-57-3 AOK Fixed by rev 7169.; GS has a better smoothing.
09-57-4 AOK Fixed by rev 7169.; GS has a better smoothing.
09-57-5 AOK Fixed by rev 7169.; GS has a better smoothing.
09-57-6 AOK Test04: With bad smoothness CPSI shadings are worse.
09-57-7 OK
09-57-8 OK Reviewed by RJJ. (something must have fixed this)
09-57-9 OK Fixed by rev 7169.
09-58-1 AOK Device Dependent: GS returns /none (4th line), Tek and CPSI return
/DefaultHalftone. GS matches Tek for some lines, CPSI for others.
09-59-1 AOK Device Dependent: CPSI and Tek hardcopy differ.
09-60-1 OK
10-01-1 OK
10-01-2 OK
10-01-3 OK
10-01-4 OK Minor differences in positions and character shapes - ADC
10-01-5 OK Minor differences visually reviewed by RJJ
10-01-6 AOK GS matches Tek hardcopy. CPSI differs in gradient in section
labeled 'Fountain_image' in the upper left.GS also
matches the appearance of of Distiller/Acrobat 7.0 - ADC
10-01-7 OK Color corrected by rev 7172.
10-02-1 OK
10-02-2 OK
10-02-3 OK
10-02-4 OK
10-02-5 OK
10-02-6 OK Minor differences in positions and character shapes - ADC
10-02-7 OK Minor differences visually reviewed by RJJ
10-02-8 AOK Differences between CPSI and GS in gradient in section
labeled 'Fountain_image' in the upper right. However GS
matches the appearance of of Distiller/Acrobat 7.0. See
10-01 page 6. - ADC
10-02-9 OK
10-03-1 AOK First section contains values dependent upon the initial CTM.
Different initial matrices for all three interpreters.
Notes: Acrobat and CPSI are rotated 180 degrees relative to
each other. Ghostscript changes the initial CTM based upon
the output device.
Distiller 7.0 reports: [ 8.333 0.0 0.0 -8.333 -2550.0 3301.0 ]
CPSI reports: [ -8.333 0.0 0.0 8.333 2550.0 -3301.0 ]
Ghostscript reports: [ 8.333 0.0 0.0 -8.333 0.0 6600.0 ]
10-03-2 OK
10-03-3 OK
10-04-1 OK
10-04-2 OK
10-05-1 OK
10-05-2 OK
10-06-1 OK
10-06-2 OK
10-07-1 OK
10-07-2 OK
10-08-1 OK Minor differences in positions and character shapes - ADC
10-08-2 OK Minor differences in positions and character shapes - ADC
10-09-1 OK
10-09-2 OK
10-09-3 OK
10-09-4 OK
10-10-1 OK
10-10-2 AOK Checksum 63395 does not match CPSI or Distiller/Acrobat 7.0.
The pages (and checksums) match Distiller 7.0 and CPSI except
for the 'much22' test. This test is specified by:
(/much22{ xy-t {0 -3 6 3 setbbox 3 0 1 0 360 arc 3 0 2 0 360 arc} ineofill not iff xyt })
The operation {10 100 itransform} is run before and after
running the test. The results should match. However Distiller
7.0 shows a -3.05176e-05 difference in the x value (307.2).
This appears to be a minor math problem in Distiller 7.0. CPSI
also shows an 'error' with this test. I do not have a copy of
CPSI to verify its values. Ghostscript produces matching values
(which is mathematically correct). - ADC
10-11-1 OK
10-11-2 OK
10-12-1 OK
10-12-2 OK
10-12-3 OK Minor differences visually reviewed by RJJ
10-12-4 AOK Differences between CPSI and GS in gradient in section
labeled 'Fountain_image' in the upper right. However GS
matches the appearance of of Distiller/Acrobat 7.0. See
10-01 page 6. - ADC
10-12-5 AOK Checksum 0 matches Distiller/Acrobat 7.0, CPSI 18736 matches
CET expected value. - ADC
10-13-1 OK
10-13-2 OK
10-13-3 OK Minor differences in positions and character shapes - ADC
10-13-4 OK Minor differences visually reviewed by RJJ
10-13-5 AOK Differences between CPSI and GS in gradient in section
labeled 'Fountain_image' in the upper right. However GS
matches the appearance of of Distiller/Acrobat 7.0. See
10-01 page 6. - ADC
10-13-6 OK Color corrected by rev 7172.
10-14-1 OK
10-14-2 OK
10-14-3 OK
10-14-4 AOK Difference between CPSI and GS with several objects:
1) The location of rectangles in items 13 and 14 in the
column labeled 'LOOK'. 2) The shape/locations of hooked
shape features in items 15, 16, and 18 in the same column.
3) The length of a diagonal line in item 19 in the same
column. 4) The position of line and arc in items 22 and
23 in the same column. 5) The position of diamond shaped
gradient at position 36 in the same column. 6) The position
of triangles at items 3 and 4 of a column labeled 'MEMO'.
7) The positions of rectangles at items 22, 23, and 24, in
the column labeled 'CASH'. 8) The position of a larger
triangle at item 5 in the column labeled 'MUCH'. 9) The
position of trangles at items 8, 9, and 27 in the same column.
10) The size of an 'angled' shaped objects at items 15 and
30 in the same column. 11) The position of the triangle
at item 27 in the column labeled 'MORE'. 12) The location
of shaded squares in item 27 of the column labeled 'GRAY'.
13) The location of rectangles in item 28 in the same column.
14) The location of a trangle in item 32 in the same column.
For all of the above objects, the positions match between GS
and Distiller/Acrobat 7.0.
The one item that has some question, item 31 in the column
labeled 'GRAY' This is a triangle outline apparently drawn
with a pattern. All three intpreters differ. The GS output
looks like CPSI except that the pattern is mirrored about a
45 degree diagonal. (Note: The pattern is only visible in
the lines that make up the lines of the triangle. Thus there
is very little information. The differences may be a minor
position shift.) Distiller/Acrobat differs from both. - ADC
10-15-1 OK
10-15-2 AOK The diffs in items 'much22' and 'much24' (user paths / insideness testing)
Analyzed by Igor.
This is a CPSI problem : the insideness testing operators
slightly disturb CTM.ty .
A simplified test case is placed to
peeves:/home/igor/pscet/mytests/10-15-2-Igor01.PS .
.
With this test CPSI prints :
GlobInterProc1=2467.17
GIFBoundProc1=false
{749.8 1433.41 749.8 1433.41 setbbox 749.8 1433.41 moveto}
false
iff in : gif_assertion=true
iff out : gif_assertion=true
GifProcInstance=false
1
GlobInterProc2=2467.17
GlobInterProc2=false
.
Note the value 2467.17 looks same due to rounding,
but the equality test fails (GlobInterProc2=false).
The effect is resolution dependent and the page size dependent.
10-16-1 OK
10-16-2 OK
10-16-3 OK
10-16-4 OK Minor differences visually reviewed by RJJ
10-16-5 OK GS matches Tek hardcopy. CPSI differs in gradient in section
labeled 'Fountain_image' in the upper right. GS also
matches the appearance of of Distiller/Acrobat 7.0. See
10-01 page 6. - ADC
10-16-6 OK Minor differences in positions and character shapes - ADC
11-01-1 AOK Matches Tek. CPSI slightly differs in few coordinates.
11-01-2 OK
11-01-3 OK
11-01-4 OK
11-01-5 OK
11-01-6 AOK rev 7617 : look30, much34 look like a saw due to "strokepath stroke".
CPSI and GS rev 7617 generate individual contours for line joins,
which look like a quadrangle with a very acute corner.
Then it's a subject of miter limit, which depends on the angle,
which depends on the flattening factor of the original curve.
With CPSI look30 has a smaller number of peaks.
11-02-1 AOK Same as 11-01-1
11-02-2 OK Minor differences visually reviewed by RJJ
11-02-3 OK
11-02-4 OK
11-02-5 OK
11-02-6 AOK Same as 11-01-6.
11-03-1 OK Minor differences visually reviewed by RJJ
11-03-2 OK
11-03-3 OK Rev. 7171 works for me. - Alex
11-03-4 OK Fixed in rev. 7171. - Alex
11-03-5 OK Minor differences visually reviewed by RJJ
11-03-6 OK Minor differences visually reviewed by RJJ
11-03-7 OK Fixed as of rev 7618.
11-04-1 OK Minor differences visually reviewed by RJJ
11-04-2 OK
11-04-3 OK Fixed iv rev. 7154. - Alex.
11-04-4 OK
11-04-5 OK Minor differences visually reviewed by RJJ
11-04-6 OK Minor differences visually reviewed by RJJ
11-04-7 OK Fixed as of rev 7618.
11-05-1 OK Minor differences visually reviewed by RJJ
11-05-2 OK Minor differences visually reviewed by RJJ
11-05-3 OK Minor differences visually reviewed by RJJ
11-05-4 OK
11-05-5 OK Minor differences visually reviewed by RJJ
11-05-6 OK fixed in r7228 - Raph
11-05-7 AOK Same as 11-01-6. Color corrected by rev 7172.
11-06-1 OK Minor differences visually reviewed by RJJ
11-06-2 OK Fixed in r7155 - Raph
11-06-3 OK Minor differences visually reviewed by RJJ
11-06-4 OK Minor differences visually reviewed by RJJ
11-06-5 AOK CPSI problem.
GS has extra line: MUCH 15, 30, 31. Tek and CPSI do not.
Analyzed by Igor.
The following test discovers a problem in CPSI :
.
72 72 moveto 72 72 rlineto clip
0 0 200 200 rectfill
showpage
72 72 moveto 72 72 rlineto clip
0 0 200 200 rectclip
clippath currentpoint
showpage
.
The test installs aclipping path from a single line.
The first page paints a line, so
the clipping is not empty.
The second page first installs same clipping,
and then applies rectclip over it.
The intersection isn't empty by theory,
but with CPSI it appears empty and
"clippath currentpoint" fails with nocurrentpoint.
11-07-1 OK
11-07-2 OK Minor differences visually reviewed by RJJ
11-07-3 OK Minor differences visually reviewed by RJJ
11-07-4 OK Minor differences visually reviewed by RJJ
11-07-5 OK Fixed by rev 7617.
11-08-1 OK Minor differences visually reviewed by RJJ
11-08-2 OK Minor differences visually reviewed by RJJ
11-08-3 OK Minor differences visually reviewed by RJJ
11-08-4 OK Minor differences visually reviewed by RJJ
11-08-5 OK Fixed as of rev. 7531.
11-09-1 AOK a better stroke adjustment; not important. Analyzed by Igor.
11-09-2 OK
11-09-3 OK
11-09-4 AOK linewidth different.
Analyzed by Igor.
GS paints a "thin line" whenever the line wdth is lesser than 1 pixel.
The decision is taken in gxstroke.c ln 453, 468, 482
with the threshold half width smaller than half pixel.
Well I can set a smaller threshold for CPSI compatibility,
but I need CPSI to measure its constant.
11-09-5 OK Minor differences visually reviewed by RJJ
11-09-6 OK
11-09-7 OK
11-09-8 AOK look30, much34 : AOK same as 11-01-6.
-
more21: aok - CPSI appears ignoring the bad flattening precision.
May be related to 11-09-4 assign: Ray.
11-10-1 AOK Test measures number of subdivisions with different values of
setflat. Differences are expected (resolution dependent). -Raph
11-10-2 OK Minor differences visually reviewed by RJJ
11-10-3 OK Minor differences visually reviewed by RJJ
11-10-4 DIFF setflat problem maybe - curves smooth in cpsi assign: Raph.
We have very coarse flattenpath results. - Raph
11-10-5 DIFF same as 11-10-4 assign: Raph.
Same as 11-10-4
11-10-6 OK Fixed as of rev 7618.
11-11-1 OK Fixed with rev 7212.
11-11-2 OK
11-11-3 OK
11-11-4 OK Minor differences visually reviewed by RJJ
11-11-5 OK Minor differences visually reviewed by RJJ
11-11-6 OK Fixed as of rev 7534.
11-12-1 OK
11-12-2 OK
11-12-3 OK
11-12-4 OK Minor differences visually reviewed by RJJ
11-12-5 OK Minor differences visually reviewed by RJJ
11-12-6 OK Minor differences visually reviewed by RJJ
11-13-1 OK Minor differences visually reviewed by RJJ
11-13-2 OK
11-13-3 OK
11-13-4 OK Works in rev. 7527.
11-14-1 OK Fixed with rev 7213.
11-14-2 OK
11-14-3 OK Minor differences visually reviewed by RJJ
11-14-4 OK Minor differences visually reviewed by RJJ
11-14-5 OK fixed in r7228 - Raph
11-14-6 OK Color corrected as of rev 7172.
11-15-1 OK Fixed as of rev 7534.
11-15-2 OK Minor differences visually reviewed by RJJ
11-15-3 OK Minor differences visually reviewed by RJJ
11-15-4 OK
11-15-5 OK Minor differences visually reviewed by RJJ
11-15-6 OK Minor differences visually reviewed by RJJ
11-15-7 OK Minor differences visually reviewed by HAS
11-16-1 OK Minor differences visually reviewed by HAS
11-16-2 OK
11-16-3 OK
11-16-4 OK Minor differences visually reviewed by HAS
11-16-5 OK
11-16-6 OK
11-16-7 OK
11-16-8 AOK
11-17-1 OK Fixed as of rev 7598.
11-17-2 OK Fixed as of rev 7598.
11-17-3 OK Minor differences visually reviewed by RJJ
11-17-4 OK Minor differences visually reviewed by RJJ
11-17-5 OK fixed in r7228 - Raph
11-17-6 AOK GS has small square in upper right of border line, CPSI has
similar square in lower right. Tek has small square in lower
right. Device Dependent: Uses device coordinates, GS, Tek
and CPSI all have different default CTM. RJJ
11-18-1 OK Fixed with rev 7212.
11-18-2 OK
11-18-3 OK
11-18-4 OK Minor differences visually reviewed by RJJ
11-18-5 OK Minor differences visually reviewed by RJJ
11-18-6 AOK Adobe interpreters (CPSI, Tek, Distiller 5) have obvious
rendering problems, which are hard to simulate in GS.
1. The long diagonal line is caused by "clippath 0 0 rlineto
stroke showpage". Different Adobe interpreters draw it
fron the center of the page to different corners.
2. The line drawn by "10 0 rlineto" is not horizontal in
LOOK 30 and MUCH 34 cases. This bug can be demonstrated
by this short program "300 300 moveto 10 10 rlineto 8 8
rlineto strokepath 10 0 rlineto stroke showpage". - Alex
11-19-1 OK
11-19-2 OK
11-19-3 OK
11-19-4 OK Minor differences visually reviewed by RJJ
11-19-5 OK Minor differences visually reviewed by RJJ
11-19-6 AOK cash21, look30 : ustrokepath ending point, Analyzed by Igor.
I believe it is CPSI problem.
I placed some tests to peeves:/home/igor/pscet/mytests/11-19-Igor01.zip .
The ZIP contains 2 PS files, which are made from 11-19.PS
with removing unrelated things and inserting a debug printing.
The 2 files differ in a small thing : wheter it access
currentpoint after ustrokepath, or not.
The rasters (included into the ZIP) appear to be different !
The log files (included into the ZIP) tell us that no error happen,
because the currenpoint coordinates are successfully printed
(in the order Y X).
The position of the 2nd cross depends on the
currentpoint after ustrokepath,
but 'currentpoint' appears to cnahge it !
I believe that CPSI actually didn't pass the "interference" test,
which 1-19-6 is designed for.
As we noticed from other 20 tests (unpublished), this behavior
appears pretty occasional, so we're unable to analyze it completely
and to reproduce it in Ghostscript.
Tek matches CPSI though - I can't know why.
Welcome anybody who can improve this analyzis.
11-20-1 OK Minor differences visually reviewed by HAS
11-20-2 OK Differences resolved as of rev 7531.
11-20-3 OK Minor differences visually reviewed by RJJ
11-20-4 OK
11-20-5 DIFF Analyzed by Igor.
I believe it is CPSI problem.
A simplified test is stored into peeves/home/igor/pscet/mytests/11-20-5-Igor01.ps .
It prints the pathbbox before and after "3 3 rlineto",
as well as the current point before and after it :
bbb0
711.0
289.037
691.0
80.3225
xxx0
691.0
85.535
bbb1
711.0
289.037
691.0
39.5
xxx1
694.0
88.535
11-20 GSTATE = 31293 Graphic 0 ms
.
Note: pathbbox and currentpoint are printed in the reverse order.
The starting and the ending point of the line appear
within the initial pathbbox.
In same time the lower X of pathbbox changes from 80.3225 to 39.5
without a visible reason. With GS it doesn't change.
Note: my copy of CPSI probably is some older,
but it renders same 11-20-5.ppm as one on peeves.
Passing to Ray for further processing.
assign: Ray.
11-20-6 DIFF In special test B, Adobe has rangechecks, we don't. In
special test A, we have one rangecheck they don't.
The test is about interaction of setbbox with save,
gsave and other operators. - assign Raph
11-20-7 OK
11-20-8 OK Minor differences visually reviewed by RJJ
11-20-9 DIFF '4's and '1's added in gs output.
The problem (at least partly) is caused by a small bounding box
set by setbbox operator that cannot accomodate long strings of text.
Adobe appears to extend path bounding box on demand instead of
enforcingwhen the limits when the current point comes from a
different gstate, for instance:
"0 0 moveto gsave 0 0 5 5 setbbox 10 10 lineto stroke showpage"
assign Alex
11-21-1 OK Fixed as of rev 7598.
11-21-10 AOK Differences in look20, look30, look33 are caused by CPSI problem :
rlineto, rcurveto count wrong coordinates after strokepath.
It looks as the starting point for relative coordinate is not equal to
the current point. CPSI paints a self-intersecting curve with
"9 0 9 -9 0 -9 rcurveto stroke", which must not self-intersect.
A simplified test is placed to
peeves:/home/igor/pscet/mytests/11-21-10-Igor01.ps
11-21-2 OK Minor differences visually reviewed by RJJ
11-21-3 OK Fixed by rev 7617.
11-21-4 DIFF Analyzed by Igor.
There are several stroking problems.
1. linewidth - same as 11-09-4.
2. setflat - same as 09-09-1.
3. CPSI doesn't roll a dash pattern at some corners.
An example is at the lower right corner of the page.
The dash at one corner looks longer than the distance to the corner.
IMO it's another CPSI bug.
Passing to Ray do decide what to to with it.
assign: Ray.
11-21-5 OK Minor differences visually reviewed by HAS
11-21-6 AOK Resolution dependent: GS gives pathbox exact (not enlarged
slightly). Tek and CPSI differ.
-
curve saw : AOK - same as 11-01-6.
11-21-7 OK Fixed by rev. 7701 - Alex
11-21-8 AOK Device Dependent / Resolution Dependent. CPSI ignores setflat
Tek and CPSI differ.
11-21-9 AOK flattness differ. Tek and CPSI differ.
11-22-1 OK
11-22-2 OK Minor differences visually reviewed by RJJ
11-22-3 AOK CPSI is wrong with the invalid test "User path must start with absolute point".
11-22-4 OK
11-22-5 DIFF GS has different number. Although this is resolution dependent,
the test comment indicates what should happen. Tek has 0.12
values, CPSI has 0.06 (approx.). Assign: alex
11-22-6 OK shape has a small 'point' with CPSI, Ghostscript matches Tek
hardcopy [2nd row down, right side, upper curve.]
11-22-7 OK fixed in r7228 - Raph
11-22-8 OK Fixed as of rev 7618.
11-23-1 AOK GS (apparently) does not implement upath caching. Tek and CPSI
do, but differ. Device Dependent (as comment says).
11-23-2 OK Minor differences visually reviewed by RJJ
11-23-3 OK ucache typecheck error in gs not seen in cpsi.
fixed in r7229 - Raph
11-23-4 OK
11-24-1 OK
11-24-2 OK
11-24-3 AOK user path :
CPSI doesn't follow PLRM page 199 about user paths :
"the first operator after setbbox must be an absolute positioning operator
(moveto, arc, or arcn)."
"All coordinates specified as operands must fall within the bounds
specified by setbbox".
When an user path starts with rmoveto, CPSI looks using current point
from the graphic state when it is set, and probably the setbbox lower coordinates
when current point is not set. However arc, arcn in the user path
do not paint a line from the graphic state current point.
Also painting outside setbbox somethimes throws error, sometimes not -
we could not figure out what does it depend on.
Due to all that CPSI's algorithm isn't clear and we're unable to reproduce it.
Will follow PLRM instead.
11-24-4 OK
11-24-5 OK
11-24-6 OK
11-24-7 OK Minor differences visually reviewed by RJJ
11-24-8 OK
11-25-1 OK
11-25-2 OK
11-25-3 AOK user paths : same as 11-24-3.
11-25-4 OK
11-25-5 OK
11-25-6 OK
11-25-7 OK Minor differences visually reviewed by RJJ
11-25-8 OK
11-26A-1 OK Minor differences visually reviewed by HAS
11-26A-2 OK Minor differences visually reviewed by RJJ
11-26A-3 OK Fixed as of rev 7598.
11-26B-1 OK Minor differences visually reviewed by RJJ
11-26B-2 DIFF precision discrepancies assign Alex
11-26C-1 OK Minor differences visually reviewed by RJJ
11-26C-2 OK Minor differences visually reviewed by RJJ
11-26C-3 OK Minor differences visually reviewed by RJJ
11-26D-1 OK Minor differences visually reviewed by RJJ
11-26D-2 OK fixed in rev 7228 - Raph
11-26D-3 OK Minor differences visually reviewed by HAS
11-27-1 OK Minor differences visually reviewed by RJJ
11-27-2 OK
11-27-3 AOK user paths : same as 11-24-3.
11-27-4 OK
11-27-5 OK
11-27-6 OK
11-27-7 OK Minor differences visually reviewed by RJJ
11-27-8 OK
11-28-1 DIFF GS has much different stroke pattern in circles/wedges. Tek
and CPSI agree. May be setflat issue. Assign: Ray
11-28-10 AOK same as 11-21-8.
11-28-11 AOK same as 11-21-9.
11-28-12 AOK same as 11-21-10.
11-28-2 OK Fixed by rev 7617.
11-28-3 AOK curve saw : AOK - same as 11-01-6.
11-28-4 AOK CPSI stroking problem. CPSI and Tek differ.
11-28-5 OK Fixed as of r7231.
11-28-6 OK Minor differences visually reviewed by RJJ
11-28-7 OK Minor differences visually reviewed by HAS
11-28-8 DIFF same as 11-21-4 assign: Ray.
11-28-9 OK Fixed by rev. 7701 - Alex
12-01-1 OK Minor text outline differences - ADC
12-01-2 OK Minor differences visually reviewed by RJJ
12-01-3 OK Minor differences versus CPSI- ADC. PDF produced by Distiller 7.0
shows teal colored circles. (GS and CPSI render sea foam green.)
12-01-4 OK Minor differences visually reviewed by RJJ.
12-01-5 DIFF Magenta in pattern in lower right corner of GS output. Not seen in
CPSI output. Strange result in Distiller 7.0 output. Acrobat 7.0
reports incorrect number of values for setcolor operator. Page is
blank.
See also 18-02G-1.
Partial analysis from Jan (11/29/2006)
1. An uncolored pattern is created, and used to draw the region on
the left. The pattern consists of a 9 unit x 9 unit tile, in the
center of which is a 7 unit x 7 unit colored square. The pattern
receives its color at the time the setpattern operator is invoked.
For this test, that color is black.
--
2. Next, a colored pattern is created, and used to draw the figure
on the right. This pattern also consists of a 9 unit x 9 unit tile,
whose PaintProc operates as follows:
- create a path consisting of a 5 unit x 5 unit square centered
in the 9 unit x 9 unit tile
- save graphic state
- set an approximately magenta color and fill the square
- restore the graphic state
- outline the same square with an approximately 4 unit wide line;
this line obscures all but a 1/2 unit x 1/2 unit square of the
magenta region
The key step is the last one, because the PaintProc does not set a
color for drawing the line. In this case it inherits the color
from the graphic state in effect when the makepattern operator was
invoked, which is the uncolored pattern created in Step (1).
--
3. Ghostscript renders the uncolored pattern into a gx_color_tile
structure when drawing the left region, and caches that strcuture.
When the second region is to be rendered, another gx_color_tile
is created, and the cached tile of the uncolored pattern is used
(in part) to render this new tile. The caching accumulator
device is used for this purpose.
--
4. The problem arises because the cached gx_color_tile for the
uncolored pattern has its phase information set for the original
device, not the accumulator device being used to create the tile
for the colored pattern.
--
As we noted earlier, Ghostscript renders patterns to
gx_color_tile structures of minimum size. Such a mechanism
cannot handle the general case in which one pattern makes use of
another pattern. Either the tile size must be increased to the
least common multiple of the patterns' step sizes, or some more
elaborate compositing technique must be used for final rendering.
--
However, in this case both patterns have the same step sizes (and
the same tile size), so Ghostscript should be able to handle the
situation correctly. We thought the problem was that the
uncolored pattern was already cached, and so we modified the code
to discard cached patterns just before executing a pattern
PaintProc. That resolves the phase issue, but the "inner"
pattern (the uncolored pattern in this case) is rendered using
the wrong transformation matrix. We haven't yet had time to go
back and determine why this is happening.
Assign: Tim
12-01-6 OK Minor differences in positions and character shapes - ADC
12-02-1 OK Minor differences in positions and character shapes - ADC
12-02-10 OK Minor differences in positions and character shapes - ADC
12-02-11 OK Minor differences visually reviewed by RJJ
12-02-12 OK Minor differences, colors more vibrant in shadings in GS.
12-02-13 OK Minor differences in positions and character shapes - ADC
12-02-14 OK Corrected by 7152 and 7162. ADC
12-02-15 OK Minor differences in positions and character shapes - ADC
12-02-16 OK Minor differences in positions and character shapes - ADC
12-02-17 OK Minor differences in positions and character shapes - ADC
12-02-18 AOK Squares in the top row are darker in GS vs CPSI.
This is exactly the same issue as 12-07B-10. After embedding
CPSI's default CRD into the job, the output matches. TIM
12-02-19 OK Minor differences in positions and character shapes - ADC
12-02-2 OK Minor differences in positions and character shapes - ADC
12-02-3 OK Minor differences in positions and character shapes - ADC
12-02-4 OK Minor differences in positions and character shapes - ADC
12-02-5 OK Minor differences in positions and character shapes - ADC
12-02-6 OK Corrected by 7152 and 7162. ADC
12-02-7 OK Minor differences in positions and character shapes - ADC
12-02-8 OK Minor differences in positions and character shapes - ADC
12-02-9 OK Minor differences in positions and character shapes - ADC
12-03-1 OK Corrected by 7152 and 7162. ADC
12-03-2 OK Minor differences visually reviewed by RJJ
12-03-3 OK Minor differences in positions and character shapes - ADC
12-03-4 OK Minor differences visually reviewed by RJJ
12-03-5 DIFF Magenta in pattern at lower right in GS. Not present in CPSI.
See 12-01-5. Assign Tim
12-03-6 OK Minor differences in positions and character shapes - ADC
12-04-1 OK Verified by ADC
12-04-2 OK Corrected by 7152 and 7162. ADC
12-04-3 OK Minor differences visually reviewed by RJJ
12-04-4 DIFF Magenta in pattern at lower right in GS. Not present in CPSI.
See 12-01-5. Assign: Tim
12-04-5 AOK GS has extra center line on symbols: LOOK 3, 12 , MUCH 9, MORE 19
Analyzed by Igor :
I believe it is a CPSI problem.
eofill is applied to two overlapping horizontal lines.
In most cases CPSI paints them, and sometimes doesn't.
Tek paints always.
GS matches Tek.
12-05-1 OK Minor differences in positions and character shapes - ADC
12-05-2 OK
12-05-3 AOK Page is gray in GS. White in CPSI.
This "{.75 mul} settransfer erasepage showpage" gives
a gray page in GS.
--
I ran this job through Acrobat Distiller 5.0 as well as Apple's
Preview, they match the GS output. The Red Book states that
erasepage is influenced "if an atypical transfer function has
been defined"... Since the transfer function is {.75 mul} a gray
value of R=191,G=191,B=191 should be expected from all that I
can see and reproduce here. I noticed it is noted several times
that "CPSI ignores settransfer". I believe this is another case
where whatever version of CPSI is used to generated these
PPM's, settransfer looks to be ignored. TIM
12-06-1 OK Minor differences in positions and character shapes - ADC
12-06-2 OK Corrected by 7152 and 7162. ADC
12-06-3 OK
12-06-4 DIFF Magenta in pattern at lower right in GS. Not present in CPSI.
See 12-01-5. Assign: Tim
12-06-5 OK Minor differences in positions and character shapes - ADC
12-07A-1 DIFF? Different blurring in squares in 7th column compared to CPSI.
GS blurring matches appearance of Distiller/Acrobat 7.0
assign: Tim
12-07A-10 OK Minor differences visually reviewed by RJJ
12-07A-11 OK Minor differences visually reviewed by RJJ
12-07A-12 OK Adobe acceptts images in the pattern color space.
Since rev. 7597 we do the same in the compatibility mode. - Alex
12-07A-13 OK Minor differences in positions and character shapes - ADC
12-07A-14 OK Minor differences in positions and character shapes - ADC
12-07A-15 OK Minor differences in positions and character shapes - ADC
12-07A-2 OK Minor differences in positions and character shapes - ADC
12-07A-3 DIFF Squares are blurred in GS. No blurring in CPSI. Distiller/Acrobat
7.0 shows blurring similar to GS.
assign: Tim
12-07A-4 OK Minor differences in positions and character shapes - ADC
12-07A-5 DIFF Page consists of blurred squares. GS position shifted down slight
at the top of the page vs CPSI and Distiller/Acrobat 7.0. - ADC
assign: Tim
12-07A-6 OK Minor differences in positions and character shapes - ADC
12-07A-7 OK Minor differences in positions and character shapes - ADC
12-07A-8 OK Minor differences visually reviewed by RJJ
12-07A-9 OK Minor differences visually reviewed by RJJ
12-07B-1 OK Minor differences in positions and character shapes - ADC
12-07B-10 AOK Some grays in the top row are lighter in GS versus CPSI
(after 7152). - ADC
This is exactly the same issue as 12-02-18. After embedding
CPSI's default CRD into the job, the gray rectangles in the top
row match. TIM
12-07B-11 DIFF Some grays in the top row are lighter in GS versus CPSI
(after 7152). - ADC Assign Tim
12-07B-12 DIFF Blurred images render a slight position shift relative to CPSI and
Distiller/Acrobat 7.0 - ADC assign: Tim
12-07B-13 DIFF Several differences. Color squares are missing in GS. The
Distiller/Acrobat 7.0 shows these squares blurred. They are not
in both GS and CPSI. - ADC assign: Tim
12-07B-14 DIFF Several differences. Color squares are missing in GS. The
Distiller/Acrobat 7.0 shows these squares blurred. They are not
in both GS and CPSI. - ADC assign: Tim
12-07B-15 DIFF Several differences. Dark squares are missing in GS. There are
dfferences in the blurring between Distiller/Acrobat 7.0 vs GS
and CPSI. - ADC assign: Tim
12-07B-2 OK Minor differences in positions and character shapes - ADC
12-07B-3 OK Minor differences in positions and character shapes - ADC
12-07B-4 OK Minor differences in positions and character shapes - ADC
12-07B-5 OK Minor differences in positions and character shapes. Some
minor differences in shades of gray. - ADC
12-07B-6 OK Minor differences in positions and character shapes. Some
minor differences in shades of gray. - ADC
12-07B-7 OK Minor differences in positions and character shapes - ADC
12-07B-8 OK Minor differences in positions and character shapes - ADC
12-07B-9 OK Minor differences in positions and character shapes. Some
minor differences in shades of gray. - ADC
12-07C-1 OK Minor differences in positions and character shapes - ADC
12-07C-10 DIFF White and black rectangles missing in GS. Differences between
CPSi and Distiller/Acrobat 7.0 - ADC assign Tim
12-07C-11 DIFF White and black rectangles missing in GS. Differences between
CPSi and Distiller/Acrobat 7.0 - ADC assign Tim
12-07C-12 DIFF White and black rectangles missing in GS. Differences between
CPSi and Distiller/Acrobat 7.0 - ADC assign Tim
12-07C-13 DIFF White, black, and cyan rectangles missing in GS. Differences
between CPSi and Distiller/Acrobat 7.0 - ADC assign Tim
12-07C-14 OK Minor differences in positions and character shapes - ADC
12-07C-15 OK
12-07C-16 OK
12-07C-17 OK
12-07C-2 OK Minor differences in positions and character shapes - ADC
12-07C-3 OK Minor differences in positions and character shapes - ADC
12-07C-4 OK Minor differences visually reviewed by RJJ
12-07C-5 OK Minor differences visually reviewed by RJJ
12-07C-6 OK Minor differences in positions and character shapes - ADC
12-07C-7 OK GS matches Tek -- no thin vertical lines.
12-07C-8 DIFF GS missing leftmost vertical column of shaded data in the
test block at the lower right of the page - ADC. assign Tim
12-07C-9 OK Color corrected by rev 7172. Something fixed the error message.
12-07D-1 OK Minor differences in positions and character shapes - ADC
12-07D-2 OK Minor differences in positions and character shapes - ADC
12-07D-3 OK Minor differences visually reviewed by RJJ
12-08A-1 DIFF Differences in the second and third test pattern in the rightmost
column. GS pattern matches Distiller/Acrobat. CPSI is different.
Adobe shows these a blurred. Both CPSI and GS are crisp. - ADC
assign Tim
12-08A-2 OK Minor differences in positions and character shapes - ADC
12-08A-3 DIFF Different pattern between GS and CPSI. Distiller/Acrobat shows
the same pattern as GS. Adobe shows the image blurred. Both
CPSI and GS shows a crisp image. (The page says interpolate
true in the upper left corner.) - ADC
assign Tim
12-08A-4 OK Color corrected by rev 7172.
12-08A-5 AOK Minor differences in positions and character shapes.
12-08A-6 AOK Minor differences in positions and character shapes.
12-08A-7 OK Color corrected by rev 7172.
12-08A-8 OK Color corrected by rev 7172.
12-08B-1 OK Minor differences in positions and character shapes - ADC
12-08B-2 OK Accept multiple data sources in imagemask by rev. 7181.
12-08B-3 OK Minor differences in positions and character shapes - ADC
12-08C-1 OK Minor differences in positions and character shapes - ADC
12-08C-2 OK Color corrected by rev 7172.
12-08C-3 OK Minor differences in positions and character shapes - ADC
12-08C-4 OK Minor differences in positions and character shapes - ADC
Color corrected by rev 7172.
12-08D-1 OK Minor differences visually reviewed by RJJ
12-08D-2 DIFF Magenta in pattern at lower right in GS. Not present in CPSI.
See 12-01-5. Assign: Tim
12-08D-3 OK Minor differences in positions and character shapes - ADC
12-09-1 OK Minor differences in positions and character shapes - ADC
12-09-2 OK Minor differences in positions and character shapes - ADC
12-09-3 OK Minor differences in positions and character shapes - ADC
12-09-4 OK Minor differences in positions and character shapes - ADC
12-10-1 OK Minor differences in positions and character shapes - ADC
12-10-2 OK Color corrected by rev 7172.
12-10-3 OK Minor differences visually reviewed by RJJ
12-10-4 OK Color corrected by rev 7172.
12-10-5 OK Minor differences visually reviewed by RJJ
12-10-6 OK Minor differences in positions and character shapes - ADC
12-10-7 OK Minor differences in positions and character shapes - ADC
12-10-8 DIFF Magenta in pattern at lower right in GS. Not present in CPSI.
See 12-01-5. Assign: Tim
12-10-9 OK Minor differences in positions and character shapes - ADC
12-11-1 OK Minor differences in positions and character shapes - ADC
12-11-2 OK Color corrected by rev 7172.
12-11-3 OK Minor differences in positions and character shapes - ADC
12-11-4 OK reviewed by RJJ.
12-11-5 OK Minor differences in positions and character shapes - ADC
12-11-6 OK Minor differences visually reviewed by RJJ
12-11-7 DIFF Magenta in pattern at lower right in GS. Not present in CPSI.
See 12-01-5. Assign: Tim
12-11-8 OK Minor differences in positions and character shapes - ADC
12-12-1 OK Minor differences in positions and character shapes - ADC
12-12-10 OK
12-12-2 OK Minor differences in positions and character shapes - ADC
12-12-3 OK Minor differences in positions and character shapes - ADC
12-12-4 OK Color corrected by rev 7172.
12-12-5 OK Color corrected by rev 7172.
12-12-6 OK Minor differences in positions and character shapes - ADC
12-12-7 OK Color corrected by rev 7172.
12-12-8 OK
12-12-9 DIFF A pattern phase difference. assign: Tim
12-13-1 OK Minor differences in positions and character shapes - ADC
12-13-2 OK
12-13-3 OK Minor differences visually reviewed by RJJ
12-13-4 DIFF Magenta in pattern at lower right in GS. Not present in CPSI.
See 12-01-5. Assign: Tim
12-13-5 OK Works for me in rev. 7579. There's no differences around 30
and 31 in left column. The checksum matches too. - Alex.
12-14A-1 OK Test03: ok - fixed.
rangecheck error is fixed - Alex.
12-14A-10 OK Minor differences in positions and character shapes and minor
differences in shading gradients. - ADC
12-14A-11 DIFF Analyzed by Igor.
Test03: diff - same as 09-47B-11 Test05. Assign: Tim
-
Test04: aok. Similar to 09-47B-10 test11.
-
Test05, Test12: aok
A bug in the test : insufficient data.
The stream interpretes like this :
000000ff 000000ff black black
00ffffff 00000000 black-red white
CPSI renders :
black green
black black -- looks wrong.
GS renders :
black red white
black black black --- looks correct.
-
Test08, Test09: ?
same as 09-47B-11 Test05. Assign: Tim
Please pay attention to the vertical diagonal of the rhombus.
With CPSI it looks as a linear gradient, and does not with GS.
Looks that CMYK colors [v v v v] differently convert to RGB.
Not sure who is wrong, but CPSI looks more pleasant.
12-14A-12 DIFF Test03, 04, 10,11 :
same as 09-47B-11 Test05. Assign: Tim
-
Test07: aok - same as 12-14A-11 Test05.
12-14A-13 DIFF Test02:aok - same as 12-14A-11 Test05.
-
Test03: aok. same as 09-47B-10 Test11.
-
Test04:
GS problem with non-contigous function.
A fix is possible but isn't simple.
A suggestion was posted to 'teco' on November 8
and then dismissed. Now it arises anew.
.
First we need gs_function_evaluate to return a new bit.
which tells that the function has a discontinuty at
the point. Fortunately poly-argument PS finctions are
always contigous (otherwise the topology of discontinuty
would be a hell).
.
There are 20 calls to patch_interpolate_color,
which deal with it. It must return a bit,
which tells whether the function
isn't contiguous at the point.
If so, it must be called at second
time for computing another limit to this point.
.
The change should look like this :
.
+ typedef enum {left, right} function_limit_direction;
.
{
- patch_color_t c;
+ patch_color_t *C;
+ function_limit_direction discontinuty;
- patch_interpolate_color(&c, c0, c1, pfs, 0.5);
+ discontinuty = patch_interpolate_color(&c, c0, c1, pfs, 0.5, left);
+ if (discontinuty)
+ patch_interpolate_color(&C, c0, c1, pfs, 0.5, right);
+ else
+ C = &c;
// Then a half of occurances 'c' to be replaced with 'C'.
}
.
Assign: igor.
-
Test05, 06, 07, 08, 12 : same as 09-47B-11 Test05. Assign: Tim
-
Test09: aok - same as 09-47B-11 Test04.
Likely CPSI problem.
The test is buggy - insufficient data for function.
The data are interpreter like this :
000000ff000000 ff00ffffffxxxx
xxxxxxxxxxxxxx xxxxxxxxxxxxxx
where x is an insufficient data. GS pads it with zeros,
CPSI is indeterministic.
After tintTransform we get CMYK colors :
000000ff ff00ffff
xxxxxxxx xxxxxxxx
GS should paint : black black*green
white white
It's what we observe.
CPSI appears wrong.
12-14A-14 OK Minor differences in positions and character shapes and minor
differences in shading gradients. - ADC
12-14A-2 OK Minor differences in positions and character shapes and minor
differences in shading gradients. - ADC
12-14A-3 OK Minor differences in positions and character shapes and minor
differences in shading gradients. - ADC
12-14A-4 OK GS error reports are changed to match CPSI. - Alex
12-14A-5 AOK GS error reports are changed to match CPSI. - Alex
-
Test07, 10, 11, 12 : aok
CPSI problem : a non-monotonic color space is rendered as monotonic.
Same as 09-47B-7 Test10.
12-14A-6 OK Minor differences in positions and character shapes and minor
differences in shading gradients. GS gradients are smoother
in tests 3 and 4. - ADC
12-14A-7 AOK Test01, 02, 03, 04, 07, 08, 09, 10 : aok
CPSI problem : a non-monotonic color space is rendered as monotonic.
Same as 09-47B-7 Test10.
12-14A-8 AOK CPSI reports 7 typechecks in shfill. GS reports none. There
are 7 gradients missing in test 6 to 12 in CPSI. - ADC
-
Analyzed by Igor :
CPSI misses the gradients due to the typecheck mentioned above.
PLRM doesn't allow packedarray as 'names' element of the color space.
GS appears tolerant to it.
IMO it should be aok.
12-14A-9 OK Minor differences in positions and character shapes and minor
differences in shading gradients. - ADC
12-14B-1 OK All test report errors now match CPSI and the test. - Alex
12-14C-1 OK Fixed - Alex.
12-14D-1 OK Fixed in rev. 7209. - Alex
12-14E-1 OK All test report errors now match CPSI and the test. - Alex
12-14F-1 OK All test report errors now match CPSI and the test. - Alex
12-14G-1 OK All test report errors now match CPSI and the test. - Alex
12-14H-1 OK Minor differences visually reviewed by RJJ
12-14H-2 OK Minor differences in positions and character shapes - ADC
12-14H-3 OK Minor differences in positions and character shapes and minor
differences in shading gradients. - ADC
12-14I-1 OK Minor differences in positions and character shapes and minor
differences in shading gradients. - ADC
12-14I-2 OK Minor differences in positions and character shapes - ADC
12-14I-3 OK Minor differences in positions and character shapes and minor
differences in shading gradients. - ADC
12-14I-4 OK Minor differences visually reviewed by RJJ
12-14J-1 OK Minor differences in positions and character shapes and minor
differences in shading gradients. - ADC
12-14J-2 OK Minor differences in positions and character shapes - ADC
12-14J-3 OK Minor differences in positions and character shapes and minor
differences in shading gradients. - ADC
12-14K-1 AOK Test04: aok - CPSI is wrong, GS is correct.
The test is buggy : it contains a wrong comment about colors
and excessive data for the function.
There is nothing special about diagonals in the function,
so CPSI paints them inaccurately.
-
Test06: aok
The test is buggy : it contains a wrong comment about colors
and insufficient data for the function.
As we know CPSI is indeterministic with insufficient function data.
There is nothing special about diagonals in the function,
so CPSI paints them inaccurately.
The colors must be :
000000ff 000000ff
00ffffff xxxxxxxx (x = insufficient)
black black
blackred white
GS appears correct.
-
test07: aok
The test is buggy : it contains a wrong comment about colors
and insufficient data for the function.
As we know CPSI is indeterministic with insufficient function data.
There is nothing special about diagonals in the function,
so CPSI paints them inaccurately.
The colors must be :
000000ff000000 ff00ffffffxxxx
xxxxxxxxxxxxxx xxxxxxxxxxxxxx (x = insufficient)
After tintTransfer :
000000 ff00ff
xxxxxx xxxxxx (x = insufficient)
white magenta
black black
GS appears correct.
-
Test08, Test09: Test10: test1: aok - same as test04.
12-14K-2 OK Minor differences in positions and character shapes - ADC
12-14K-3 AOK Test06 : aok. The function cropping, same as 09-47G-2 Test10.
-
Test07 : aok. The function cropping
(a few different cropping than test06), same as 09-47G-2 Test10.
-
Test09 : aok. The function cropping
(a few different cropping than previous), same as 09-47G-2 Test10.
-
Test10 : aok. The function cropping
(a few different cropping than previous), same as 09-47G-2 Test10.
12-14L-1 OK Minor differences in positions and character shapes - ADC
12-14L-2 OK Minor differences in positions and character shapes - ADC
12-14L-3 OK fixed.
12-14L-4 OK Fixed by rev 7206
12-14M-1 OK fixed.
12-14M-2 OK fixed.
12-14M-3 OK fixed.
12-14M-4 DIFF Test05: Tek has top half of cone solid dark gray, GS has light
gray (as in Test06). Assign: Igor
-
Test06: ok - fixed.
-
Test04, Test07: ok - Fixed by rev 7206
Note: CPSI gets rangecheck errors on Test06, 07, 08.
12-14M-5 OK fixed.
12-14M-6 OK fixed.
12-14M-7 OK fixed.
12-14M-8 OK fixed.
12-14N-1 OK Minor differences visually reviewed by RJJ
12-14N-10 OK Minor differences in positions and character shapes - ADC
12-14N-2 OK Minor differences in positions and character shapes - ADC
12-14N-3 OK Minor differences in positions and character shapes - ADC
12-14N-4 OK Minor differences in positions and character shapes - ADC
12-14N-5 OK Fixed by rev 7206
12-14N-6 OK Fixed by rev 7206
12-14N-7 OK Minor differences visually reviewed by RJJ
12-14N-8 OK Minor differences in positions and character shapes - ADC
12-14N-9 OK fixed.
12-14O-1 OK Minor differences in positions and character shapes - ADC
12-14O-2 OK
12-14O-3 OK Fixed with rev 7211.
12-14O-4 OK Fixed as of rev. 7174
Note: Tek tests 9-12 blank.
12-14O-5 OK Fixed by rev 7206
12-14O-6 OK Minor differences in positions and character shapes - ADC
12-14O-7 OK Minor differences in positions and character shapes - ADC
12-14O-8 OK Minor differences in positions and character shapes - ADC
12-14P-1 OK Minor differences in positions and character shapes - ADC
12-14P-2 OK Minor differences visually reviewed by RJJ
12-14P-3 OK Minor differences in positions and character shapes - ADC
12-14P-4 AOK Fixed by rev 7206
CPSI has a worse shading precision with a non-contigous color.
12-14P-5 AOK CPSI problem with self-intersecting tensor shadings - Analyzed by Igor.
GS matches Distiller/Acrobat. - ADC
12-14Q-1 OK Minor differences in positions and character shapes - ADC
12-14Q-2 OK Minor differences in positions and character shapes - ADC
12-14Q-3 OK Minor differences in positions and character shapes - ADC
12-14Q-4 OK Fixed by rev 7206
12-14Q-5 OK Minor differences in positions and character shapes - ADC
12-14R-1 AOK Test02, 03, 06, 07, 08: aok. The function cropping, same as 09-47G-2 Test10.
CPSI is wrong.
12-14R-2 AOK Worse smoothing with CPSI.
12-14R-3 DIFF Analyzed by Igor.
-
Test07:
Analyzed by Igor.
The test looks buggy : duplicate DataSource.
CPSI uses the first DataSource (padded with zeros at end),
and GS uses the second one while constructing the shading dictionary.
GS interpretes the stream like this :
000000ff 000000ff
00ffffff 00000000
so the shading is not completely black - it is 1/256 gray in one corner.
Note : lib/gs_*.ps files assume the usage of the last definition.
I recall that at least gs_ttf.ps depends on that.
Passing to Ray to decide what to do with it.
assign: Ray.
-
Test08: aok - CPSI ingores Encode.
It maps a haff of shading Domain outside (to negative values)
of the function domain, so that shading must look constant at
the riht half. That's what we see with GS.
-
Test09, 10, 11, 12: aok, same as Test07 except a different Encode.
Note : the vertical 'spike' in Test10 is a false human visual effect.
12-14R-4 AOK Test1: aok - A worse smoothing with CPSI.
-
Test12: aok :
The function itself :
000000 0000ff black blue
ff0000 00ff00 red green
ff00ff 00ff00 magenta green
ffff00 ffffff yellow white
It is linear, but the green component has Decode 0 3,
so it maps outside the color domain and
therefore it is cropped by the max color value 1.
Due to the color croppig it has a plattue
around the point where the green peak would apper.
Then the function domain is cropped
with 300/400 = 0.75 by X
and 500/600 = 0.83 by Y.
Therefore we don't see yellow and white.
GS paint it fine.
CPSI has a bad smoothing and a linear gradient instead the plattue.
12-14S-1 OK Minor differences in positions and character shapes - ADC
12-14S-2 OK Minor differences in positions and character shapes - ADC
12-14S-3 OK Minor differences in positions and character shapes - ADC
12-14T-1 OK fixed
12-14T-2 OK fixed
12-14U-1 OK
12-14U-2 OK
12-14U-3 OK Minor differences visually reviewed by RJJ
13-01-1 OK Minor differences visually reviewed by RJJ
13-01-2 OK
13-01-3 OK Minor differences visually reviewed by RJJ
13-01-4 OK Minor differences visually reviewed by RJJ
13-01-5 OK
13-01-6 OK Minor differences visually reviewed by HAS
13-01-7 OK rev 7617.
13-02-1 OK Minor differences visually reviewed by RJJ
13-02-10 OK rev 7617.
13-02-2 OK Minor differences visually reviewed by RJJ
13-02-3 OK
13-02-4 OK
13-02-5 OK Fixed as of rev. 7583.
13-02-6 OK Minor differences visually reviewed by RJJ
13-02-7 OK Minor differences visually reviewed by RJJ
13-02-8 OK
13-02-9 OK Minor differences visually reviewed by HAS
13-03-1 OK
13-03-2 DIFF show error incomformity.
{{/a /b add } (abc) cshow} stopped
returns only true on the stack. assign: Peter
13-03-3 OK Minor differences visually reviewed by RJJ
13-03-4 OK Matches Tek hardcopy. PhotoPRINT appication we use for testing
redefines halftone operations as no-op, which is wrong and
cannot be considered a reference implementation. - Alex
13-03-5 OK Minor differences visually reviewed by RJJ
13-03-6 OK
13-03-7 OK Color corrected by rev 7172.
13-03-8 AOK Same as cash21, look30 : ustrokepath ending point.
13-04-1 AOK Analyzed by Igor.
CPSI prints an incomplete page.
GS raster matches Tek except (1) different order
of keys in a dictionary forall (implementation dependent),
and (2) tek has an incorrect value FontType -1.
13-04-2 OK
13-04-3 OK
13-05-1 OK
13-05-2 OK
13-05-3 AOK Adobe's interpreters accept a read-only font dictionary as a
valid argument of definefont and modify it anyway. This is a
bug that we will not emulate. (r7165 by Alex + comment by
Peter)
13-05-4 OK Minor differences visually reviewed by HAS
13-06-1 OK Fixed in rev. 7163. - Alex
13-06-2 OK
13-07-1 OK
13-07-2 OK Adjust error reporting from findfont procedure to match
Distiller 5 and CET 13-06.PS.tek360 . The PhotoPrint product
is clearly wrong here leaving a few objects on the stack
besides the arguments when findfont fails. rev. 7164 - Alex
13-07-3 OK
13-08-1 OK Fixed by r7642.
13-09-1 OK Fixed by r7642.
13-10-1 OK
13-10-2 OK
13-10-3 OK
13-10-4 OK
13-10-5 OK Works for me in rev. 7543 - Alex.
13-10-6 OK Minor differences visually reviewed by RJJ
13-10-7 OK
13-10-8 OK Minor differences visually reviewed by HAS. Color corrected by rev 7172.
13-10-9 OK rev 7617.
13-11-1 OK
13-12-1 OK
13-12-10 OK rev 7617.
13-12-2 OK
13-12-3 OK
13-12-4 OK
13-12-5 OK Minor differences visually reviewed by RJJ
13-12-6 OK Matches Tek hardcopy, same as 13-03-4
13-12-7 OK Minor differences visually reviewed by RJJ
13-12-8 OK
13-12-9 OK Color corrected by rev 7172.
13-13-1 OK
13-13-2 OK
13-13-3 OK Minor differences visually reviewed by HAS
(large diff count because of minor color diffs)
13-13-4 OK Color corrected by rev 7172.
13-13-5 OK Minor differences visually reviewed by HAS
13-13-6 OK
13-13-7 OK
13-13-8 OK Color corrected by rev 7172. Fixed by rev 7617.
13-14-1 AOK same as 13-04-1
13-14-2 OK
13-14-3 OK
13-15-1 OK
13-15-2 OK Color corrected by rev 7172.
13-15-3 OK
13-15-4 OK
13-15-5 OK Minor differences visually reviewed by HAS
13-16-1 OK
13-16-2 OK
13-16-3 OK Minor differences visually reviewed by RJJ
13-16-4 OK
13-16-5 OK Minor differences visually reviewed by RJJ
13-16-6 OK
13-16-7 OK Minor differences visually reviewed by RJJ
13-16-8 OK
13-16-9 OK rev 7617.
13-17-1 OK Corrected by 7152 and 7162. ADC
13-17-2 OK Corrected by 7152 and 7162. ADC
13-17-3 OK Corrected by 7152 and 7162. ADC
13-17-4 OK Corrected by 7152 and 7162. ADC
13-17-5 OK Minor differences visually reviewed by RJJ
13-17-6 OK Corrected by 7152 and 7162. ADC
13-17-7 OK Minor differences in positions and character shapes - ADC
13-17-8 OK rev 7617.
13-18-1 OK
13-18-10 OK Minor differences in positions and character shapes - ADC
13-18-11 OK rev 7617.
13-18-2 OK
13-18-3 OK
13-18-4 OK
13-18-5 OK
13-18-6 OK
13-18-7 OK
13-18-8 OK
13-18-9 OK
13-19-1 AOK GS has a better stroke adjustment. not impotant. Analyzed by Igor.
13-19-2 AOK Same as 13-19-1. Analyzed by Igor.
13-19-3 AOK Same as 13-19-1. Analyzed by Igor.
13-19-4 AOK Same as 13-19-1. Analyzed by Igor.
13-19-5 OK Minor differences visually reviewed by RJJ
13-19-6 OK Minor differences in positions and character shapes - ADC
13-19-7 OK rev 7617.
13-20-1 OK Minor differences in positions and character shapes - ADC
13-20-2 OK
13-20-3 OK Minor differences visually reviewed by RJJ
13-20-4 OK
13-20-5 OK Corrected by 7152 and 7162. ADC
13-20-6 OK rev 7617.
13-21-1 OK Fixed by r7642.
13-22-1 OK
13-22-2 OK Minor differences visually reviewed by RJJ
13-22-3 OK Minor differences visually reviewed by RJJ
13-22-4 OK Minor differences visually reviewed by RJJ
13-22-5 OK Color corrected by rev 7172.
13-22-6 OK rev 7617.
13-23-1 OK GS matches CPSI and Tek when Adobe fonts are used.
With URW or UFST fonts, an extra line under /ordfeminine and
/ordmasculine characters shows as well as slightly different
appearance of other characters. This is a font issue and must
be resolved by the font vendor chosen for the implementation.
13-24-1 OK
13-24-2 AOK An unimportant difference in the character cache logics. Analyzed by Igor.
13-24-3 OK Minor differences visually reviewed by RJJ
13-24-4 OK Minor differences in positions and character shapes - ADC
13-24-5 OK Minor differences visually reviewed by Igor.
13-24-6 AOK Same as cash21, look30 : ustrokepath ending point.
13-25-1 OK
13-25-2 OK Fixed as of r7610 (when run with -dLOCALFONTS).
13-25-3 OK
13-26-1 OK Minor differences visually reviewed by RJJ
13-26-2 OK
13-26-3 OK
13-26-4 OK Fixed in rev. 7562. - Alex
13-26-5 OK Minor differences visually reviewed by RJJ
13-26-6 OK Minor differences visually reviewed by RJJ
13-26-7 OK Minor differences visually reviewed by RJJ
13-26-8 AOK Minor differences in colors - ADC
13-26-9 OK rev 7617.
13-27-1 OK Minor differences visually reviewed by RJJ
13-27-2 OK
13-27-3 OK rev 7542.
13-27-4 OK Minor differences visually reviewed by RJJ
13-27-5 OK Minor differences visually reviewed by RJJ
13-27-6 OK
13-27-7 AOK Minor differences in colors - ADC
13-27-8 OK rev 7617.
13-28-1 OK Minor differences visually reviewed by RJJ
13-28-2 OK Minor differences visually reviewed by RJJ
13-28-3 OK
13-28-4 OK rev 7542.
13-28-5 OK Minor differences visually reviewed by RJJ
13-28-6 OK Minor differences visually reviewed by RJJ
13-28-7 OK Minor differences visually reviewed by RJJ
13-28-8 AOK Minor differences in colors - ADC
13-28-9 OK rev 7617.
13-29-1 OK Minor differences visually reviewed by RJJ
13-29-2 OK
13-29-3 OK rev 7542.
13-29-4 OK Minor differences visually reviewed by RJJ
13-29-5 OK Minor differences visually reviewed by RJJ
13-29-6 OK Minor differences visually reviewed by RJJ
13-29-7 OK Minor differences visually reviewed by Igor.
13-29-8 OK rev 7617.
14-01-1 OK Minor differences visually reviewed by RJJ
14-01-10 OK Minor differences visually reviewed by RJJ
14-01-11 OK Minor differences visually reviewed by RJJ
14-01-12 OK Minor differences visually reviewed by RJJ
14-01-13 OK Minor differences visually reviewed by RJJ
14-01-14 OK Minor differences visually reviewed by RJJ
14-01-15 OK Minor differences visually reviewed by RJJ
14-01-2 OK Minor differences visually reviewed by RJJ
14-01-3 OK Minor differences visually reviewed by RJJ
14-01-4 OK Minor differences visually reviewed by RJJ
14-01-5 OK Minor differences visually reviewed by RJJ
14-01-6 OK Minor differences visually reviewed by RJJ
14-01-7 OK Minor differences visually reviewed by RJJ
14-01-8 OK Minor differences visually reviewed by RJJ
14-01-9 OK Minor differences visually reviewed by RJJ
14-02-1 OK Minor differences visually reviewed by RJJ
14-02-2 OK
14-03-1 OK Minor differences visually reviewed by RJJ
14-03-2 OK
14-03-3 OK Minor differences visually reviewed by RJJ
14-04-1 OK Minor differences visually reviewed by RJJ
14-04-2 OK Minor differences visually reviewed by RJJ
14-05-1 OK
14-05-2 OK
14-06-1 OK
14-07-1 OK Minor differences visually reviewed by RJJ
14-08-1 OK Color corrected by rev 7172.
14-08-10 OK Corrected by 7152 and 7162. ADC
14-08-11 OK Corrected by 7152 and 7162. ADC
14-08-12 OK Corrected by 7152 and 7162. ADC
14-08-13 OK Corrected by 7152 and 7162. ADC
14-08-14 OK Corrected by 7152 and 7162. ADC
14-08-15 OK Corrected by 7152 and 7162. ADC
14-08-16 OK Corrected by 7152 and 7162. ADC
14-08-17 OK A minor difference in color. Analyzed by Igor.
14-08-18 OK Color corrected by rev 7172.
14-08-19 OK A minor difference in color. Analyzed by Igor.
14-08-2 OK Color corrected by rev 7172.
14-08-20 OK A minor difference in color. Analyzed by Igor.
14-08-21 OK A minor difference in color. Analyzed by Igor.
14-08-22 OK A minor difference in color. Analyzed by Igor.
14-08-23 OK Color corrected by rev 7172.
14-08-24 OK A minor difference in color. Analyzed by Igor.
14-08-3 OK A minor difference in color. Analyzed by Igor.
14-08-4 OK Color corrected by rev 7172.
14-08-5 OK A minor difference in color. Analyzed by Igor.
14-08-6 OK Color corrected by rev 7172.
14-08-7 OK A minor difference in color. Analyzed by Igor.
14-08-8 OK Colors corrected by rev 7172.
14-08-9 OK Corrected by 7152 and 7162. ADC
14-09-1 OK
14-09-2 OK
14-09-3 OK
14-09-4 OK
14-10-1 OK
14-10-2 OK
14-10-3 OK
14-10-4 OK
14-11-1 OK Minor differences visually reviewed by RJJ
14-11-2 OK Minor differences visually reviewed by RJJ
14-12-1 OK
14-12-2 OK
14-12-3 AOK A glyph positioning difference.
Analyzed by Igor.
I believe it is CPSI problem.
A simplified test is placed to
peeves:/home/igor/pscet/mytests/14-12-3-Igor01.PS
With this test CPSI prints :
.
aaaa1:
Helvetica
has WMode
def:/Helvetica-Oblique
1
defined:1
aaaa2:
Helvetica
nas no WMode
.
Thus the redefinition of the font Helvetica
is dropped when Helvetica-Oblique is redefined.
It is an incorrect behavior.
Due to that the cshow procedure prints
currentfont /WMode like this :
[1 1 /U /U /U /U 1 1 0 0 0 0]
The right result must be :
[1 1 1 1 0 0 1 1 0 0 0 0]
Using the wrong WMode, glyph placements
are wrong with CPSI.
Tek paints something different,
which is closer to Ghostscript,
but looks wrong again.
.
Some tips for those who wants to analyze it :
1. Font_Dict is a FMap 2 font with 8 descendents :
1.0. Times-Roman
1.1. Type3WMode0 (paints rectangles with size depending on char code);
1.2. Type3WMode1 (same as above but with WMode 1)
1.3. Type0Font1 2 scalefont - the test doesn't use it.
1.4. FMAP2 - a FMap 2 font [ Helvetica, Helvetica-Bold, Helvetica-Oblique, Helvetica-BoldOblique]
1.5. FMAP4 [Symbol, Symbol 2 scalefont] - the test doesn't use it.
1.6. FMAP5 [Courier, , , ] - the test doesn't use it.
1.7. FMap6 [Helvetica, ...] - the test doesn't use it.
2. All 1st level descendent fonts above has WMode = DescendentWMode = 1
3. Helvetica, Helvetica-Bold, Helvetica-Oblique, Helvetica-BoldOblique
are built into the font above aftyer redefinition with find&mod,
which changes WMode like this :
Helvetica WMode 1
Helvetica-Bold WMode 0
Helvetica-Oblique WMode 1
Helvetica-BoldOblique WMode 0
4. The text string is decomposed like this :
2A 2a 40A 40a 41A 41a 42A 42a 43A 43a 1A 1A
The 1st byte selects the 1st level descendent,
then it tripples the 2nd byte selects the 2nd level descendent.
5. We believe that WMode of the leaf font only have sence.
14-12-4 OK
14-12-5 AOK Same as 14-12-3.
14-12-6 AOK Same as 14-12-3.
14-13-1 OK Minor differences visually reviewed by RJJ
14-13-2 OK Minor differences visually reviewed by RJJ
14-14-1 OK
15-01-1 OK Fixed as of Nov 4 2005
15-01-2 OK Minor rasterizing differences. -Tor
15-02-1 OK Fixed as of rev 7598.
15-02-2 OK Minor differences visually reviewed by RJJ
15-03-1 OK Fixed as of Nov 4 2005
15-04-1 AOK GS has -12dict- and CPSI has -11dict- in debug text. Different checksum. The
font dictionary debug for GS has one more key: /OrigFont.
I think these differences are harmless, and are not particularly easy to fix. If we do decide it needs fixing, Peter is a good choice. - Raph
15-04-2 AOK Different checksum. The font dictionary debug for GS has more keys: /OrigFont and /FontInfo.
See 15-04-1 - Raph
15-04-3 AOK Different checksum. The font dictionary debug for GS has one more key: /OrigFont.
See 15-04-1 - Raph
15-05-1 OK Fixed when tested with 7210
15-05-2 OK Fixed when tested with 7210
15-05-3 OK Color corrected by rev 7172.
15-06-1 OK Corrected by rev 7172.
15-07-1 OK Minor rasterizing differences. -Tor
15-08-1 OK Fixed in r7156 - Raph
15-09-1 OK Minor rasterizing differences. Stroke width difference. -Tor
15-09-2 OK Minor rasterizing differences. -Tor
15-09-3 OK Minor rasterizing differences. -Tor
15-10-1 OK Minor differences visually reviewed by RJJ
15-10-2 OK Minor differences visually reviewed by RJJ
15-10-3 OK Minor rasterizing differences. -Tor
15-10-4 OK Minor rasterizing differences. -Tor
15-11-1 OK Fixed by rev. 7806. - Alex
15-14-1 OK Minor differences visually reviewed by RJJ
15-15-1 OK Minor rasterizing differences. -Tor
16-01-1 AOK glyphs 19-22 : Tiny gaps between almost adjoining black rectangles;
output will vary from device to device to device, and our rendering is valid.
Glyphs 207, 208: reverse video in Adobe output, likely Adobe bug.
Simplified tests are placed at
peeves:/home/igor/pscet/mytests/16-01/
A patch for emulating such behavior is placed at
peeves:/home/igor/pscet/mypatches/16-01-Igor-01.txt
We won't commit that patch.
Suggesting to give the patch to the customer as a separate product.
16-02-1 AOK Same as 16-01-1.
16-03-1 OK Fixed by rev 7172.
16-03-2 OK
16-03-3 OK
16-03-4 OK Now appears correct - Raph
16-04-1 OK Fixed by rev. 7686 - Alex
16-05-1 OK Color corrected by rev 7172
16-05-2 DIFF Three smallest sizes are slanted squares in gs, same shape as larger ones in Adobe.
Also USE_ADOBE_CMYK_RGB issue - Raph
assign: Ray
16-05-3 OK Color corrected by rev 7172.
16-06-1 OK trivial font rendering diff - Raph
16-07-1 OK Checked by Raph
16-08-1 OK Fixed in r7156 - Raph
16-09-1 OK
16-10-1 AOK Same as 16-01-1.
16-10-2 AOK Same as 16-01-1.
16-10-3 AOK Same as 16-01-1.
16-11-1 OK Fixed by rev 7172.
16-14-1 OK
16-15-1 OK
16-16-1 OK
17-01-1 OK
17-01-2 OK Fixed as of rev 7598.
17-01-3 OK Fixed in rev 7630 (Peter)
17-01-4 OK Fixed as of rev 7598.
17-01-5 OK
17-01-6 AOK Matches Tek. CPSI different.
17-01-7 AOK Matches Tek and Distiller 5. CPSI is different and wrong,
The test exercises ineofill operator and can be easily checked
visually by modifying the test to fill the current path.
17-01-8 OK
17-02-1 OK
17-02-2 OK Fixed as of rev. 7598.
17-02-3 OK Same as 17-01-3
17-02-4 OK Fixed as of rev 7598.
17-02-5 OK
17-02-6 AOK Same as 17-01-6 but tests infill instead of ineofill.
Matches Tek and Distiller 5. CPSI is different and wrong,
17-02-7 AOK Same as 17-01-6 but tests infill instead of ineofill.
Matches Tek and Distiller 5. CPSI is different and wrong,
17-02-8 OK
17-03-1 OK
17-03-2 DIFF GS has curves where CPSI has straight lines. Also GS differs in marker symbols.
Note Tek has _different_ straight line pattern and marker symbols.
This is not identified as "Device Dependent" in the test.
curve flatness used to draw lines. Assign: ray
17-03-3 OK Same as 17-01-3
17-03-4 OK Minor differences visually reviewed by RJJ
17-03-5 DIFF Similar to 17-01-6 but tests instroke. The insideness testing is OK
but the pattern phase is different from Tek and CPSI. assign: Ray
17-03-6 OK
17-04-1 OK
17-04-2 AOK The test is unreasonable. The test checks whether a point is in
the user path (using inueofill operator) but places many of
the points on the boundary of the path where insideness is
affected by rounding errors. When the point is inside the path,
a small black circle is drawn around the point. The test is
performed for different CTM.
-
Ghostscript skips some of the points at low resolution but
draws all expected points at 300 dpi and above.
-
CPSI, Tek, and Distiller 5 are all different and skip different
points depending on the current CTM. - Alex
17-04-3 OK same as 17-02-3
17-04-4 OK
17-04-5 OK
17-05-1 OK
17-05-2 AOK Similar to 17-04-2 but testing inufill. - Alex
17-05-3 OK Same as 17-01-3
17-05-4 OK
17-05-5 OK
17-06-1 OK
17-06-2 AOK Similar to 17-04-2 but testing inustroke. Placing the dots in
middle of the stroke is reasonable but some of the dots are
placed at the end of the line where their relative position
depends on the rounding errors. Ghostscript generates
correct results above 300 dpi. CPSI and Tek miss some
points, - Alex
17-06-3 OK Fixed as of rev 7636.
17-06-4 OK
18-01A-1 AOK Minor difference in the font size.
Note: patterns are resolution dependent,
18-01A-2 OK Minor differences visually reviewed by HAS.
18-01A-3 OK Minor differences visually reviewed by RJJ
18-01A-4 OK
18-01A-5 OK Minor differences visually reviewed by RJJ
18-01A-6 OK Minor differences visually reviewed by RJJ
18-01A-7 OK Minor differences visually reviewed by RJJ
18-01B-1 DIFF The word "FORM" in the magenta oval throughout the test
is white with cpsi and gray with gs. (repeats). - color saturation problem.
assign Tim
18-01B-2 DIFF (18-01B-1) lower right graphic cpsi has red lines where gs has white. (assign Raph).
The locations of squares are shifted in this pattern. Distiller/Acrobat
7.0 reports error 'incorrect number of values for setcolor operator'.
This is probaly related to the problem for 12-01-5. See Jan's
email analysis of that problem. (He reports a problem with pattern
shifts with patterns.)
18-01C-1 OK Minor differences visually reviewed by RJJ
18-02A-1 DIFF pattern size differs first row of graphics. assign Tim
18-02A-2 OK Minor differences visually reviewed by HAS. Color corrected by rev 7172.
18-02A-3 OK Minor differences visually reviewed by HAS.
18-02A-4 OK Reviewed by Raph. Probably fixed by r7114.
18-02B-1 OK Minor differences visually reviewed by HAS.
18-02B-10 AOK 18-02A-4 repeat
Fill rule is ok, shadings sharper in GS than Adobe. - Raph
Per conversation with Ray, we have decided that our rendering
of these shadings (which have discontinuities) is more correct
than Adobe. Also note that for test 9, Tek and GS match, but
CPSI doesn't.
18-02B-11 DIFF 18-02A-4 repeat
Fill rule is ok. Test 1, 4 shadings sharper in GS than Adobe.
Test 9-12 colors darker in Adobe than GS. - Assign Tim
18-02B-12 DIFF 18-02A-4 repeat
Fill rule is ok. Test 1-3, 11 colors darker in Adobe.
Test 12 colors lighter in Adobe. - Raph
assign Tim
18-02B-13 DIFF 18-02A-4 repeat
Fill rule is ok. Test 1-6 colors different as above. Raph
assign Tim
18-02B-2 OK
18-02B-3 OK Minor differences visually reviewed by HAS.
18-02B-4 OK Minor differences visually reviewed by HAS.
18-02B-5 OK Minor differences visually reviewed by RJJ
18-02B-6 OK Minor differences visually reviewed by RJJ
18-02B-7 DIFF 18-02A-4 repeat
Fill rule is ok.
Test 8-12 colors darker in Adobe than GS. - Raph
assign Tim
18-02B-8 OK Probably fixed by r7114.
18-02B-9 AOK 18-02A-4 repeat
Fill rule is ok, shadings sharper in GS than Adobe. - Raph
18-02C-1 DIFF Partialy fixed by rev. 7738.
Adobe doesn't restore operands of rectfill operator when
an error happen in the PaintProc function but
Ghostscript does. assign: Alex
18-02D-1 OK Fixed by rev. 7738. - Alex
18-02E-1 DIFF much of postscript dump missing in gs. assign Ray
Ghostscript doesn't detect color operators in uncolored
pattern.
18-02E-2 DIFF much of postscript dump missing in gs. assign Ray
Ghostscript doesn't detect color operators in uncolored
pattern.
18-02F-1 OK Minor differences visually reviewed by HAS.
18-02F-2 DIFF pattern in third row of graphics out of phase. assign Tim
18-02F-3 DIFF 18-02A-4 repeat
repeat of AOK problem assign Tim
18-02F-4 OK Fixed before Nov 3 (after Oct 2).
18-02F-5 DIFF 18-02A-4 repeat
repeat of AOK problem assign Tim
18-02G-1 DIFF strokepath different - assign: Igor.
-
pattern differs in 4th column of stars in the last box.
Analyzed by Igor.
See also 12-01-5.
A simplified test case is placed to
peeves:/home/igor/pscet/mytests/18-02G-1-Igor01.PS .
This is a general pattern implementation problem.
The test defines a pattern with TESTPROC,
and then applyes this pattern for creating another pattern
Inf71ColorPattern. Doing the 2nd step, Ghostscript
instantiates the 1st pattern in the 2nd pattern's space.
Contrary to that, CPSI instantiates the 1st pattern
in the device space when instantiating the 2st pattern.
Thus the overall pattern period with CPSI is a multiple
of periods of the 1st and the 2nd patterns.
This problem may be fixed for Ghostscript
with implementing high level patterns for
clist and raster devices.
It would be a significant development,
which we currently have no time for.
Passing to Ray for schedulling.
assign: Ray.
18-02G-2 OK Fixed by r7123.
18-02G-3 AOK Minor differences visually reviewed by HAS.
19-01-1 OK
19-01-2 OK
19-02-1 OK
19-02-2 OK
19-02-3 OK
19-03-1 OK
19-03-2 OK
19-03-3 OK
19-04-1 OK
19-04-2 OK
19-05-1 OK Minor differences visually reviewed by RJJ
19-05-2 OK Minor differences visually reviewed by RJJ
19-05-3 OK
19-06-1 OK
19-06-2 OK
19-07-1 OK
19-07-2 OK
19-08-1 OK
19-08-2 OK
19-09-1 OK
19-09-2 OK
19-09-3 OK
19-10-1 OK
19-10-2 OK
19-11-1 OK Minor differences visually reviewed by RJJ
19-11-2 OK
19-12-1 OK Minor differences visually reviewed by RJJ
19-12-2 OK
19-13-1 OK Minor differences visually reviewed by RJJ
19-13-2 OK Minor differences visually reviewed by RJJ
19-13-3 OK
19-14-1 OK Minor differences visually reviewed by RJJ
19-14-2 OK
20-01-1 OK Fixed in rev. 7712. - Alex
20-01-2 AOK GS prints { 1 2 3 } [ 1 2 3 ] (1 2 3) where CPSI prints { 1 2 3 } [mark /1 -3array- ] and ( )
i.e. CPSI modifies the content of the array arguments when operator >> fails.
Ghostscript matches Tek and Distiller 5. This is a bug in CPSI we don't
want to reproduce. - Alex
20-01-3 OK
20-02-1 OK Looks okay. -Tor
20-02-2 OK Fixed by rev. 7829. - Alex
20-02-3 OK Color corrected by rev 7172.
20-03-1 OK
20-03-2 OK
20-04-1 OK Minor differences visually reviewed by RJJ
20-04-2 OK Minor differences visually reviewed by RJJ
20-04-3 OK Color corrected by rev 7172.
20-05-1 OK
20-05-2 OK
20-06-1 OK
20-06-2 OK
20-07-1 OK
20-07-2 OK
20-08-1 OK
20-08-2 OK
20-08-3 OK
20-09-1 OK
20-09-2 OK
20-09-3 OK
20-10-1 OK
20-10-2 OK
20-11-1 OK fixed incidentally.
20-11-2 OK fixed incidentally.
20-11-3 OK
20-12-1 OK Looks okay. -Tor
20-12-2 OK
20-13-1 OK Looks okay. -Tor
20-13-2 OK Color corrected by rev 7172.
20-14-1 OK GS matches Tek hardcopy. (Peter->Ray)
20-14-2 OK Color corrected by rev 7172.
20-15-1 OK Fixed by rev. 7713. - Alex
20-15-2 OK
20-16-1 OK Looks okay. -Tor
20-16-2 OK Color corrected by rev 7172.
20-17-1 OK
20-17-2 OK
20-18-1 OK Looks okay. -Tor
20-18-2 OK Color corrected by rev 7172.
20-19-1 OK Minor differences visually reviewed by RJJ
20-19-2 OK Color corrected by rev 7172.
20-20-1 OK GS matches Tek and CPSI as of rev. 7741.
20-20-2 OK
20-21-1 OK
20-21-2 OK
20-22-1 OK
20-22-2 OK
20-22-3 OK
20-23-1 OK Tek and CPSI don't have /deviceinfo. Availability of this operator
on Ghostscript deprnds on gs_dps device. Rev. 7723 undefines it in
gs_cet.ps to pass the test regardless of the configuration. - Alex
20-23-2 OK
20-24-1 OK
20-24-2 OK
20-25-1 OK
20-25-2 OK
20-25-3 OK
20-26-1 OK
20-26-2 OK
20-27-1 AOK This test check the content of the $error dictionary.
The content varies depending on the code executed before.
There's no error on CPSI, as it was stated before, The
value of /errorname is an error name, which is always
printed as white on black by the test program. - Alex
20-27-2 OK
21-01-1 OK
21-01-2 OK
21-02-1 OK Minor differences visually reviewed by RJJ
21-02-2 OK Minor differences visually reviewed by RJJ
21-02-3 OK Minor differences in positions and character shapes - ADC
21-03-1 OK Minor differences in positions and character shapes - ADC
21-03-2 OK Minor differences in positions and character shapes - ADC
21-03-3 OK
21-04-1 OK Minor differences in positions and character shapes - ADC
21-04-2 OK Minor differences in positions and character shapes - ADC
21-04-3 OK Minor differences in positions and character shapes - ADC
21-05-1 OK Minor differences in positions and character shapes - ADC
21-05-2 OK Minor differences in positions and character shapes - ADC
21-05-3 OK Minor differences in positions and character shapes - ADC
21-06-1 OK Minor differences in positions and character shapes - ADC
21-06-2 OK Minor differences in positions and character shapes - ADC2
21-06-3 OK Minor differences in positions and character shapes - ADC
21-07-1 OK Minor differences visually reviewed by RJJ
21-07-2 OK Minor differences in positions and character shapes - ADC
21-07-3 OK Minor differences in positions and character shapes - ADC
21-08-1 OK Minor differences visually reviewed by RJJ
21-08-2 OK Minor differences visually reviewed by RJJ
21-08-3 OK Minor differences in positions and character shapes - ADC
21-09-1 OK
21-09-2 OK
21-10-1 OK Minor differences in positions and character shapes - ADC
21-10-2 OK Minor differences in positions and character shapes - ADC
21-11-1 OK
21-11-2 OK Fixed as of rev 7598.
21-11-3 OK Fixed as of rev 7616.
21-11-4 OK
21-11-5 OK
21-11-6 DIFF Differences in error messages for operator token - ADC
assign Peter
21-11-7 OK Corrected by 7152 and 7162. ADC
22-01-1 OK
22-01-2 OK
22-02-1 OK
22-02-2 OK
22-03-1 OK
22-03-2 OK
22-04-1 OK
22-04-2 OK
22-05-1 OK Minor differences visually reviewed by RJJ
22-05-2 OK
22-06-1 OK
22-06-2 OK
22-07-1 OK
22-07-2 OK
22-07-3 OK
22-08-1 OK Minor differences visually reviewed by RJJ
22-08-2 OK Minor differences visually reviewed by RJJ
22-08-3 OK Visually OK SK small variations in tiny colour cube.
22-09-1 OK
22-09-2 OK
22-10-1 OK Minor differences visually reviewed by HAS
22-10-2 OK
22-10-3 OK
22-11-1 OK
22-11-2 OK
22-12-1 OK Minor differences visually reviewed by RJJ
22-12-2 OK Minor differences visually reviewed by RJJ
22-13-1 OK
22-13-2 OK
22-13-3 OK
22-13-4 OK
23-01-1 OK
23-01-2 OK
23-02-1 OK
23-02-2 OK
23-03-1 AOK Device Dependent: Test checks for 'bytesavailable' on 'currentfile'.
This is generally not supported from %stdin.
23-03-2 OK
23-04-1 OK
23-04-2 OK
23-05-1 OK
23-05-2 OK
23-06-1 OK
23-06-2 OK
23-07-1 OK Fixed by rev. 7442
23-07-2 AOK There is still a diff in 23-07-2 in that the contents of our resource
category dictionaries (in this case, the Generic dictionary) are different
from the Adobe interpreters: we and Adobe have different implementation-specific
(non-PLRM) entries, and the PLRM-documented procedures (FindResource, etc.)
are different. This is inherently implementation-dependent and does not
affect any real program, so no further changes are needed. (Peter)
23-07-3 OK Fixed by rev. 7442
23-07-4 OK
23-08-1 OK
23-09-1 DIFF The first bytesavailable diff is identified as "highly device dependent".
As such, I recommend discounting it. No real program will rely on the
result.
-
The two invalidaccess diffs occur because the reference implementation does
not allow PostScript programs to open files for writing. This is not a gs
bug, but it does suggest that gs should provide some way to indicate/detect
that it is running in an embedded environment with no (writable) filesystem.
Does this exist already?
-
The remaining 3 diffs occur because gs implements %null as a special file
name and the reference implementation apparently doesn't, per a previous
e-mail.
-
assign: ray
23-10-1 AOK Device Dependent: GS looks like it is working similar to CPSI ot
Tek, but all three differ. RJJ
23-10-2 OK
23-11-1 OK Minor differences visually reviewed by RJJ
23-12A-1 OK
23-12A-2 OK
23-12A-3 AOK The discrepancies are in the length of the encoded
string produced by the ASCII85Encode and
RunLengthEncode filter respectively. These filters
have some discretion in their output -- ASCII85Encode
must insert a \n at least every 80 characters, but may
insert one more often, and RunLengthEncode is not
required to produce maximum-length runs -- and it
appears that gs's implementation of these filters
produces output that is 1 byte shorter than Adobe's in
some cases. gs's output is (also) correct -- the CET
verifies it by decoding it and comparing against the
original. (peter)
23-12A-4 AOK same resolution as 23-12A-3.
23-12B-1 DIFF GS gets "-file- ( ioerror --readstring--" on FAX_6 test. Last
image of top row missing, others may be shifted. assign: Peter
23-12C-1 OK
23-12C-2 OK
23-12C-3 OK
23-12D-1 OK Fixed incidentally
23-12D-2 OK Fixed incidentally
23-12D-3 OK Fixed incidentally
23-12D-4 OK Fixed incidentally
23-12E-1 AOK Device Dependent: Test says "length can vary due to multiple valid
methods". GS differs by 1 to 5 in results (hope that's close enough)
RJJ
23-12E-2 AOK (Peter) 7605 fixes ioerror and incorrect output problems.
-
"There are still discrepancies in the length of the
compressed data, for the same reason as 23-12A-3 et
al. In this case, gs's implementation is slightly
inferior to Adobe's (about 6% larger compressed data
in the worst case on the CET page), probably because
in the mixed 1- and 2-D case, we don't bother to
compress each line both ways and use whichever method
produces the smaller output."
23-12E-3 OK Device Dependent: minor position/stairstep differences, but visually
equivalent to Tek. RJJ.
23-12E-4 OK Device Dependent: minor position/stairstep differences, but visually
equivalent to Tek. RJJ.
23-12E-5 DIFF GS has lower three graphics 'squashed' horizontally. assign: Ray
23-12E-6 DIFF GS has lower three graphics as .step-pyramids'. assign: Ray
23-12E-7 OK
23-12F-1 OK
23-12G-1 AOK same as 23-12A-3 (peter)
23-12G-2 AOK same as 23-12A-3 (peter)
23-12G-3 OK
23-12H-1 OK
23-12I-1 OK. Fix from Peter
23-12I-2 OK. Fix from Peter
23-12J-1 OK
23-12J-2 OK
23-12J-3 OK
23-12J-4 OK
23-12K-1 OK Fixed as of r7500.
23-12K-2 OK Fixed by rev 7808. - Alex
23-12L-1 AOK same as 23-12A-3
23-12M-1 AOK same as 23-12A-3
23-12N-1 AOK same as 23-12A-3
23-12O-1 AOK same as 23-12A-3
23-12O-2 OK Fixed by rev 7489.
23-12P-1 AOK same as 23-12A-3
23-12P-2 OK Fixed by rev 7489.
23-12Q-1 OK
23-12R-1 OK
23-12S-1 OK
23-12T-1 OK
23-12U-1 OK Fixed as of r7500.
23-12V-1 OK Fixed as of r7500.
23-12W-1 OK Fixed as of r7500.
23-12W-2 OK Fixed as of rev 7636.
23-12W-3 OK Minor differences visually reviewed by RJJ
23-12W-4 OK Matches Tek hardcopy. CPSI-based PhotoPrint product ignores
sethalftone and cannot be a reference implementation here.
23-12W-5 OK Fixed by rev 7149.
23-12W-6 OK Fixed as of r7500.
23-13-1 OK Fixed as of r7610 (when run with -dLOCALFONTS).
23-13-2 AOK Device Dependent: GS has more entries for Filter, ColorSpace
page fills before ImageType and PatternType shown.
23-13-3 DIFF Device Dependent: GS has different Resource Categories and also
has different number of columns. Assign: Ray for further analysis
23-13-4 AOK Ghostscript has extra keys in the font dictionary. PLRM says that
implementations can add specific keys. Rev. 7724 - Alex
23-13-5 OK Fixed by rev 7442.
23-13-6 OK
23-14-1 OK
23-14-2 OK
23-15-1 OK
23-15-2 OK
23-16-1 OK Minor differences visually reviewed by RJJ
23-16-2 OK
23-17-1 OK Minor differences visually reviewed by RJJ
23-17-2 AOK Probably fixed with 23-17-3
23-17-3 AOK Fixed see 23-29-2, 23-36-3 and 23-36-4
23-18-1 OK Minor differences visually reviewed by RJJ
23-18-2 OK
23-19-1 OK
23-19-2 OK
23-20-1 OK Fixed by rev. 7809. - Alex
23-20-2 OK
23-21-1 OK Fixed in rev. 7140. - Alex.
23-21-2 OK
23-22-1 OK Fixed in rev. 7143. - Alex.
23-22-2 OK
23-23-1 OK Fixed in rev. 7144. - Alex.
23-24-1 OK
23-24-2 OK
23-25-1 DIFF GS gets 'undefinedfilename' error. Tek and CPSI do not. Assign: Ray
23-25-2 DIFF GS gets 'undefined' error during defineresource. (PLRM3 requires
CIDSystemInfo). Device Dependent: CPSI doesn't define Localization
category. Needs Tek hardcopy. Assign: Ray
23-25-3 DIFF GS close to Tek, but stack contents when 'stackunderflow' reported
differs. GS reports 'undefinedresource' vs. 'typecheck' of Tek.
Different error and stack from ((FNT1) {cvn} () resourceforall).
Needs Tek hardcopy. Assign: Ray
23-25-4 OK
23-25-5 OK
23-25-6 OK
23-25-7 OK
23-26-1 OK
23-26-2 OK Fixed before Oct 2.
23-26-3 AOK Device Dependent: GS has more entries for Filter, ColorSpace
page fills before ImageType and PatternType shown. Same as 23-13-2.
23-26-4 OK Adobe interpreters have resoursestatus as an operator but report errors
in the executable name. Since rev. 7619 we do the same. - Alex
23-26-5 OK
23-26-6 OK
23-28-1 OK Fixed in rev. 7146. - Alex.
23-29-1 OK
23-29-2 OK fixed in rev. 7232 (Peter)
23-29-3 OK
23-30-1 OK
23-30-2 OK
23-31-1 OK Matches Tek hardcopy. RJJ (CPSI differs)
23-31-2 OK
23-32-1 OK
23-32-2 DIFF 'token' error inconformity.
Environment dependent (a virtual address in the printing).
assign: Peter
23-32-3 OK Fixed as of rev 7598.
23-32-4 OK fixed in rev 7672 (Peter)
23-32-5 OK
23-32-6 OK
23-33-1 AOK Device Dependent: status and number of Font, Encoding, and
ColorSpaceFamily resources pre-loaded differs between CPSI and
Tek and Ghostscript. (this is not a problem).
23-33-2 OK GS matches Tek hardcopy. CPSI shows "0 -1 T" in last line.
23-33-3 OK Fixed in rev. 7620 - Alex
23-33-4 OK Fixed in rev. 7620 - Alex
23-33-5 OK Fixed in rev. 7619; See 23-26-4 for details. - Alex.
23-33-6 OK
23-34-1 OK
23-34-2 OK
23-35-1 OK
23-35-2 OK
23-35-3 OK
23-36-1 OK
23-36-2 OK Fixed in r7631 Peter
23-36-3 AOK fixed in rev. 7232 (Peter)
23-36-4 OK fixed in rev. 7232 (Peter)
23-37-1 OK
23-37-2 OK
23-37-3 OK
24-01-1 OK
24-01-2 OK
24-02-1 OK
24-02-2 OK
24-02-3 OK
24-03-1 OK
24-03-2 OK
24-04-1 OK
24-04-2 OK
24-05-1 OK Fixed by Alex, rev. 7131.
24-05-2 AOK Fixed by Alex, rev. 7131. except that Adobe interpreters have
a bug when "1234556789 10 ( ) cvrs" restores the stack as
"123456789 ( ) ( )" Ghostscript doesn't have this bug.
Since we don't want to put in a bug, this will be an anomaly.
24-05-3 OK
24-06-1 OK
24-06-2 OK Fixed by Alex, rev. 7131.
24-06-3 OK
24-07-1 OK
24-07-2 OK
24-08-1 OK Minor differences visually reviewed by RJJ
24-08-2 OK
24-08-3 OK
24-09-1 OK
24-09-2 OK
24-09-3 OK
24-10-1 OK
24-10-2 OK
24-10-3 OK
24-11-1 OK Minor differences visually reviewed by RJJ
24-11-2 OK
24-11-3 OK
24-12-1 OK Minor differences visually reviewed by RJJ
24-12-2 OK
24-13-1 OK
24-13-2 OK
24-13-3 OK Fixed by r7642.
24-14-1 OK
24-14-2 OK
25-01-1 OK
25-01-2 OK
25-02-1 OK
25-02-2 OK
25-03-1 OK Minor differences visually reviewed by RJJ
25-03-2 OK
25-04-1 OK Minor differences visually reviewed by RJJ
25-04-2 OK
25-05-1 OK GS matches Tek hardcopy. Fixed by rev 7140.
GS has more accurate (and correct) sin/cos functions than CPSI
25-05-2 OK
25-06-1 OK
25-06-2 OK
25-07-1 OK Fixed by rev 7139.
25-07-2 OK
25-08-1 OK Minor differences visually reviewed by RJJ
25-08-2 OK
25-09-1 OK Fixed by rev 7225.
25-09-2 OK Fixed by rev 7224.
25-10-1 OK GS matches Tek hardcopy. Fixed by rev 7140.
25-10-2 OK
25-11-1 OK GS matches Tek hardcopy. Fixed by rev 7140.
GS has more accurate (and correct) sin/cos functions than CPSI
25-11-2 OK
25-12-1 OK
25-12-2 OK
25-13-1 OK Fixed - Alex.
25-13-2 OK
25-14-1 OK
25-14-2 OK
25-15-1 OK
25-15-2 OK
25-16-1 OK
25-16-2 OK
25-16-3 OK
25-17-1 OK
25-17-2 OK
25-18-1 OK GS matches Tek hardcopy. Fixed by rev 7140.
GS has more accurate (and correct) sin/cos functions than CPSI
25-18-2 OK
25-19-1 OK Works for me in rev. 7149.
25-19-2 OK
25-20-1 OK
25-20-2 OK
25-21-1 OK
25-21-2 OK
25-22-1 OK
25-22-2 OK
26-01-1 OK
26-01-2 OK
26-01-3 OK
26-02-1 OK
26-02-2 OK
26-03-1 OK
26-03-2 OK Minor differences visually reviewed by RJJ
26-03-3 OK
26-04-1 OK
26-04-2 OK
26-05-1 OK
26-05-2 OK
26-05-3 OK
26-06-1 OK
26-06-2 OK
26-06-3 OK
26-07-1 OK
26-07-2 OK
26-07-3 OK
26-08-1 OK
26-08-2 OK
26-09-1 OK
26-09-2 OK
26-10-1 OK
26-10-2 OK
26-11-1 OK
26-11-2 OK
26-11-3 OK
26-12-1 OK
26-12-2 OK
26-13-1 OK
26-13-2 OK
26-13-3 OK
27-01-1 OK Minor differences visually reviewed by RJJ
27-01-2 OK
27-02-1 OK Adobe has upper cache limit of 1e8 bytes.
Rev. 7746 does the same in CPSI compatibility mode. - Alex
27-02-2 OK
27-03-1 OK Fixed as of rev 7598.
27-03-2 AOK test unreasonable:
This page enumerates all the IODevices, arbitrarily selects the last one,
and then runs the GIF test, checking whether the currentdevparams values for
that IODevice change as a result of running the test procedures. This is
dependent both on the set of IODevices (which is implementation-dependent,
although gs could be expected to include at least the non-hardware-dependent
ones documented in the PLRM3, which I believe it does) and the order in
which they are enumerated by /IODevice resourceforall (which is totally
implementation-dependent).
-
It appears that in the reference output, the last-enumerated IODevice is one
whose devparams values change somewhat unpredictably -- probably %Calendar%.
In the gs output, it's one whose devparams values don't change -- i.e. any
of the others! (peter)
27-04-1 DIFF Installation dependent (FontResourceDir).
Inconforming currentsystemparams. Analyzed by Igor.
assign: Ray for further analyzis.
27-04-2 OK
27-05-1 DIFF Inconforming currentuserparams. Analyzed by Igor.
assign: Ray for further analyzis.
27-05-2 OK memo36 - setvmthreshold; cash18 ucachestatus;
assign: Ray for further analyzis.
more15 - same as currentstrokeadjust in 09-49D-3
peter writes:
These CET files will not run on my 64-bit build (they expect to be able to
'cvs' any integer into a 12-character string), and I don't seem to be able
to get gcc to produce 32-bit output, even with Ralph's help. Please
reassign these files.
27-06-1 OK Fixed in rev. 7157 - Alex.
27-06-2 OK
27-07-1 AOK Fixed by rev. 7733 - Alex
Alternate use of --setcacheparams-- and /setcacheparams in error
messages is an obvious bug in Adobe interpreters that we won't
reproduce.
27-07-2 OK
27-08-1 OK Fix error reporting for setdevparams operator, rev. 7161. - Alex
27-08-2 OK
27-09-1 DIFF Same as 27-04-1. Analyzed by Igor.
assign: Ray for further analyzis.
27-09-2 DIFF Same as 27-04-1. Analyzed by Igor.
assign: Ray for further analyzis.
27-09-3 OK Minor differences visually reviewed by RJJ
27-09-4 OK
27-10-1 OK Minor differences visually reviewed by RJJ
27-10-2 OK
27-11-1 DIFF Inconforming currentuserparams. See 27-05-1. Analyzed by Igor.
assign: Ray for further analyzis.
27-11-2 OK Minor differences visually reviewed by RJJ
27-11-3 OK
27-12-1 DIFF ucachestatus differs. Analyzed by Igor.
assign: Raph.
27-12-2 DIFF Same as 27-05-2. Analyzed by Igor. assign: Ray..
28-01-1 OK
28-01-2 OK
28-02-1 OK
28-02-2 OK Fixed before Nov 4 2005 (after Oct 2)
28-02-3 OK
28-03-1 AOK The following differences remain:
1. names of internal operators are different
2. execution stack od 49 elements doesn't fit into 40-element
arrays provided by the test causing different errors.
Rev. 7730 - Alex
28-03-2 OK
28-04-1 OK
28-04-2 OK
28-05-1 OK
28-05-2 OK Fixed by rev. 7725
28-05-3 OK
28-06-1 AOK Implementation Dependent: Section 6.3 - SYNTAX 2. The PLRM says the
the order of dict enumeration by 'forall' is "arbitrary".
28-06-2 OK
28-06-3 OK
28-07-1 OK
28-07-2 OK
28-07-3 OK
28-08-1 OK
28-08-2 OK
28-09-1 OK Minor differences visually reviewed by RJJ
28-09-2 OK Visually reviewed by RJJ
(previous report of '1' in more 15 is without gs_cet.ps).
28-10-1 OK Fixed by rev. 7727. - Alex
28-10-2 OK Minor differences in boxed black/white 'T' 'F' glyphs, font renderer ufst.
28-11-1 OK Minor differences visually reviewed by RJJ
28-11-2 OK Visually reviewed by RJJ
28-12-1 OK
28-13-1 OK Minor differences visually reviewed by RJJ
28-13-2 OK Visually reviewed by RJJ
28-14-1 OK Fixed by r7115.
28-14-2 OK Minor differences visually reviewed by RJJ
29-02-1 OK Minor differences visually reviewed by RJJ
29-02-2 OK Fixed by rev. 7716. Ghostscript matches Tek and CPSI except
that CPSI-based PhotoPrint doesn't implement copypage and skips
pages 3 and 4. - Alex
29-02-3 OK See 29-02-2.
29-02-4 OK See 29-02-2.
29-02-5 OK See 29-02-2.
29-02-6 OK See 29-02-2.
29-02-7 OK See 29-02-2.
29-02-8 OK See 29-02-2.
29-03-1 DIFF Device Dependent. Too many InputAttributed. Analyzed by Igor. assign: Ray.
29-03-2 DIFF Device Dependent. OutputPage not known. Analyzed by Igor. assign: Ray
29-03-3 DIFF Device Dependent. currentpagedevice different. Analyzed by Igor. assign: Ray
29-03-4 DIFF Many conformity tests fail. Analyzed by Igor.
The test expects currentpagedevice to return the same dictionaty each time.
assign: Alex.
29-05-1 OK Make the spot function in the high resolution halftone readable
to accomodate the test logic. Ghostscript matches Tek and
Distiller 5. CPSI-based PhotoPrint product redefines halftone
operations as no-op and cannot serve as a reference
implementation here. - Alex
29-05-2 OK
29-05-3 OK
29-07A-1 OK
29-07A-2 OK
29-07A-3 OK Device Dependent. CPSI ignores setcolortransfer to perform
image negation. Ghostscript matches Tektronix hardcopy.
29-07A-4 OK corrected with gs_cet.ps that only has 'letter' PageSize
in InputAttributes dictionary.
29-07A-5 OK corrected with gs_cet.ps that only has 'letter' PageSize
in InputAttributes dictionary.
29-07A-6 AOK Device Dependent: Error message differs. Ghostscript gives
"configurationerror in setpagedevice", CPSI gives bogus
"undefinedfilename in file" on this "Page 10 experiment"
29-07B-1 OK Verified visually by RJJ
29-07B-10 OK Verified visually by RJJ. CPSI ignores NumCopies.
cpsi/29-07B-10 missing
29-07B-11 OK Verified visually by RJJ. CPSI ignores NumCopies.
cpsi/29-07B-11 missing
29-07B-12 OK Verified visually by RJJ. CPSI ignores NumCopies.
cpsi/29-07B-12 missing
29-07B-2 DIFF GS shifts output up and left, CPSI and Tek ignore Margins [ -9 9 ]
assign ray.
29-07B-3 DIFF GS shifts output down and right, CPSI and Tek ignore Margins [ 9 -9 ]
assign ray.
29-07B-4 OK Verified visually by RJJ
29-07B-5 OK Verified visually by RJJ
29-07B-6 OK Verified visually by RJJ
29-07B-7 OK Verified visually by RJJ. CPSI ignores NumCopies.
29-07B-8 OK Verified visually by RJJ. CPSI ignores NumCopies.
29-07B-9 OK Verified visually by RJJ. CPSI ignores NumCopies.
29-07C-1 OK
29-07C-2 DIFF Device Dependent: CPSI shows offset page. GS shows 'undefined'
error. Tektronix shows 'configurationerror'. May be a default
Policy setting vs. CPSI. GS should issue 'configurationerror'
Assign: Ray.
29-07C-3 AOK Device Dependent: Ghostscript matches Tektronix hardcopy and
gives 'configurationerror'. CPSI output is offset.
29-07C-4 DIFF Device Dependent: CPSI shows offset page. GS shows 'undefined'
error. Tektronix shows 'configurationerror'. May be a default
Policy setting vs. CPSI. GS should issue 'configurationerror'
Assign: Ray.
29-07C-5 AOK Device Dependent: Ghostscript matches Tektronix hardcopy and
gives 'configurationerror'. CPSI output is offset.
29-07C-6 DIFF Device Dependent: CPSI shows offset page. GS shows several
parameters with value '1'. Tektronix shows only "/PageSize 1"
Assign: Ray.
29-07C-7 AOK Device Dependent: Ghostscript matches Tektronix hardcopy and
gives 'configurationerror'. CPSI output is offset.
29-07C-8 DIFF Device Dependent: CPSI shows offset page. GS offset differs.
Tektronix hardcopy shows << /PageSize 3 >> with offset that
might be the same as CPSI (clipping right edge after 's' of
'setpagedevice' in the lower right banner block).
assign ray
29-07C-9 DIFF Device Dependent: CPSI shows offset page. GS offset differs.
Tektronix hardcopy shows << /PageSize 3 >> with offset that
might be the same as CPSI (clipping right edge after 's' of
'setpagedevice' in the lower right banner block).
assign ray
29-07C-10 DIFF Device Dependent: CPSI shows offset page. GS has no offset
Tektronix hardcopy shows << /PageSize 3 >> with offset that
might be the same as GS (no clipping).
assign ray
29-07C-11 DIFF Device Dependent: CPSI shows offset page. GS has no offset
Tektronix hardcopy shows << /PageSize 3 >> with offset that
might be the same as GS (no clipping).
assign ray
29-07D-1 OK
29-07D-2 OK
29-07D-3 OK corrected with gs_cet.ps that only has 'letter' PageSize
in InputAttributes dictionary.
29-07D-4 OK
29-07D-5 DIFF Ghostscript is producing a page when OuputPage is 'false'.
cpsi/missing output for 'Page 5 experiment'.
assign ray
29-07D-6 DIFF Device Dependent: Ghostscript is honoring 'PageOffset'
Adobe ignores this. correct by modifying setpagedevice.
Assign: Ray
29-07D-7 DIFF Device Dependent: Ghostscript is honoring 'PageOffset'
Adobe ignores this. correct by modifying setpagedevice.
Assign: Ray
29-07E-1 OK
29-07E-2 AOK GS rotates the image to fit the page, same as CPSI when
rotated. PageSize Policy==7 of the PLRM says: "Positioning
of the page image on the medium is device-dependent and
unpredictable. This policy exists solely for use in the
emulations of certain LanguageLevel 1 compatibility operators
that perform media selection and page device setup separately.
...
Because its behavior violates the PostScript page device model,
documents that use this policy are not portable; for this reason,
it should never be used directly in a page description."
29-07E-3 AOK See explanation under 29-07E-2.
29-07E-4 DIFF GS does not generate a configuration error. Tek does. PLRM says
> 5 units difference should generate configurationerror with
Policy == 7. GS rotates the page and generates a larger raster.
Assign: Ray (Peter can thank me that I didn't give this to him).
29-07E-5 DIFF Same as 29-07E-4. Assign: Ray
29-07E-6 DIFF Same as 29-07E-4. Assign: Ray
29-07E-7 OK
29-07F-1 OK
29-07F-2 OK Matches Tek hardcopy. RJJ (CPSI differs / wrong)
29-07F-3 OK
29-07F-4 OK
29-07F-5 OK
29-07F-6 OK
29-07F-7 OK Matches Tek hardcopy. RJJ (CPSI differs / wrong)
29-07F-8 OK Matches Tek hardcopy. RJJ (CPSI differs / wrong)
29-07G-1 OK
29-07G-2 OK
29-07G-3 OK
29-07G-4 OK
29-07G-5 OK
29-07G-6 OK
29-07G-7 OK
29-07G-8 OK
29-07H-1 OK
29-07H-2 OK
29-07H-3 OK
29-07H-4 OK
29-07H-5 OK
29-07H-6 OK
29-07H-7 OK
29-07H-8 OK
29-07I-1 OK
29-07I-2 OK
29-07I-3 OK
29-07I-4 OK
29-07I-5 OK
29-07I-6 OK
29-07I-7 OK
29-07I-8 OK
29-07I-9 OK
29-07J-1 OK
29-07J-2 OK
29-07J-3 OK
29-07J-4 OK
29-07J-5 OK
29-07J-6 OK
29-07J-7 OK
29-07J-8 OK
29-07J-9 OK
29-07K-1 OK
29-07K-2 OK
29-07K-3 OK
29-07K-4 OK
29-07K-5 OK
29-07K-6 OK
29-07K-7 OK
29-08-1 OK
29-08-2 OK
29-08-3 OK Minor differences visually reviewed by RJJ
29-08-4 OK Visually reviewd by RJJ. CPSI doesn't do #copies.
Ghostscript matches Tektronix hardcopy.
29-08-5 OK Visually reviewd by RJJ. CPSI doesn't do #copies.
Ghostscript matches Tektronix hardcopy.
29-08-6 OK Visually reviewd by RJJ. CPSI doesn't do #copies.
Ghostscript matches Tektronix hardcopy.
cpsi/29-08-6 missing
29-08-7 OK Visually reviewd by RJJ. CPSI doesn't do #copies.
Ghostscript matches Tektronix hardcopy.
cpsi/29-08-7 missing
30-01-1 OK
30-01-2 OK
30-02-1 OK
30-02-2 OK
30-03-1 AOK These discrepancies occur because
/show load gcheck
returns true in gs and false in the reference output. This is an unarguable
bug in the reference interpreters: see the documentation for gcheck in the
PLRM3. (peter)
30-03-2 OK
30-04-1 OK
30-04-2 OK Fixed by rev 7439.
30-04-3 OK
30-04-4 OK fixed in packed array colorspace fix (ray) don't recall the rev.
30-04-5 OK
30-05-1 OK
30-05-2 OK
30-06-1 AOK These discrepancies occur because
/show load gcheck
returns true in gs and false in the reference output. This is an unarguable
bug in the reference interpreters: see the documentation for gcheck in the
PLRM3. (peter)
30-06-2 OK
30-07-1 OK Fixed by r7642.
30-07-2 DIFF /GlobalFontDirectory /Type3Font1 known gs=true, cpsi=false. assign Peter.
30-07-3 DIFF gs missing '1' for gray33. Assign Peter.
30-08-1 OK Fixed by r7642.
30-08-2 DIFF same as 30-07-2 assign Peter.
30-08-3 DIFF same as 30-07-3 assign Peter
30-09-1 OK fixed in Revision 7233
30-09-2 OK
30-10-1 OK fixed in Revision 7233
30-10-2 OK
30-11-1 OK
30-11-2 AOK Environment dependent.
GS prints different numbers for every test in GLOBINT.
The number printed is the VM usage for the test.
31-01-1 DIFF Attributes differ (bottom of page filetype).
(%stdin)(r) file is global on GS and local on Adobe.
Also see bug 687682, where a similar problem was reported earlier.
Assign: Ray
31-01-10 OK
31-01-2 OK Works for me in rev. 7453. - Alex
31-01-3 OK Minor differences visually reviewed by RJJ
31-01-4 OK
31-01-5 OK MAX_DEPTH parameter is adjusted to match Adobe. - Alex
31-01-6 OK Fixed - Alex
31-01-7 OK Minor differences visually reviewed by RJJ
31-01-8 DIFF Test 03 and 07 wrong.
The test defines 2 or more idioms that match the target and expects
the last definition to be used. PLRM (p. 121) says that the order is
undefined. assign Alex.
31-01-9 DIFF same as 31-01-8 but in the global VM. assign Alex.
31-02-1 OK Fixed in rev. 7473 - Alex
31-02-2 OK
31-03-1 OK Fixed in rev 7474. - Alex
31-03-2 OK
31-04-1 OK Minor differences in positions and character shapes - ADC
31-04-2 OK Minor differences in positions and character shapes - ADC
31-05-1 OK
31-05-2 OK
31-06-1 OK Fixed by gs_cet.ps
31-06-2 OK Minor differences in positions and character shapes - ADC
31-07-1 OK Minor differences visually reviewed by Igor.
Note: This test draws different length line depending on timing
(number of loops depends on 'realtime' resolution and results).
31-07-2 OK
31-08-1 OK Fixed by gs_cet.ps rev. 7180 (Device Dependent).
Note: CPSI shows revision 0, Tek shows revision 5.
31-08-2 OK
31-09-1 OK Fixed by gs_cet.ps rev. 7180 (Device Dependent).
Note: CPSI shows serialnumber 233460, Tek shows 1401788461
31-09-2 OK
31-10-1 OK Minor differences in positions and character shapes - ADC
31-10-2 OK Fixed in rev. 7749. - Alex
31-10-3 OK
31-11-1 OK Fixed in rev 7474. - Alex
31-11-2 OK
31-12-1 OK Minor differences visually reviewed by RJJ
Note: This test draws different length line depending on timing
(number of loops depends on 'realtime' resolution and results).
31-12-2 OK
31-13-1 OK
31-13-2 OK Fixed as of rev 7583.
32_all-1 AOK CPSI inserts CDevProc while /CIDFont defineresource.
That is not required by PLRM3.
Inserting a CDevProc is unnecessary and would impackt performance.
32_all-10 OK
32_all-11 OK
32_all-12 OK
32_all-13 OK
32_all-14 OK Fixed as of rev 7583.
32_all-15 AOK Same as 34_all-6
32_all-16 OK
32_all-2 OK Fixed in rev. 7478. findfont, definefont, undefinefont also
handle /CIDFont resource instances now. - Alex
32_all-3 OK
32_all-4 OK
32_all-5 OK
32_all-6 OK
32_all-7 OK
32_all-8 OK
32_all-9 OK
33_all-1 OK Fixed as of rev 7583.
33_all-2 OK Fixed as of rev 7583.
33_all-3 OK Minor differences visually reviewed by RJJ
33_all-4 OK
33_all-5 OK
33_all-6 OK Fixed as of rev 7583.
33_all-7 AOK CMAP names differ implementation dependent analyzed by Ray.
33_all-8 OK Minor differences visually reviewed by RJJ
34_all-1 OK Fixed as of rev 7583.
34_all-2 OK Fixed as of rev 7583.
34_all-3 OK Fixed as of rev 7583.
34_all-4 DIFF Improved by rev 7583, but still has mismatch for glyph 222.
Assign: Ray
34_all-5 OK Fixed as of rev 7583.
34_all-6 AOK GS has more CMap resources. Not important. Analyzed by Igor.
34_all-7 OK Fixed as of rev 7583.
35_all-1 DIFF Even as of rev 7583, CID Font has /CharStrings array. Tek
and CPSI do not have this key. Tek has CDevProc that CPSI does
not. Assign: Ray
35_all-10 OK
35_all-11 DIFF GS paints glyphs when CPSI and Tek paint hollow boxes. It appears that
GS is not removing glyphs from previous page "currentfont removeall".
Assign: Ray
35_all-12 OK Fixed by rev. 7710 - Alex.
35_all-2 OK Fixed as of rev 7583.
35_all-3 OK Minor differences visually reviewed by RJJ
35_all-4 OK
35_all-5 OK
35_all-6 OK Fixed as of rev 7711 (detected during regression testing. RJJ)
35_all-7 AOK Same as 33_all-7. Analyzed by Igor.
35_all-8 OK Minor differences visually reviewed by RJJ
35_all-9 OK Minor differences visually reviewed by RJJ
36-01-1 OK
37-01-1 OK
38-01-1 DIFF CPSI has more fonts, particularly more Type 42 fonts.
Analyzed by Igor.
This test prints some data about installed Type 42 fonts :
Chicago, Geneva, HoeflerText-Black, HoeflerText-BlackItalic, HoeflerText-Italic.
Neither URW nor UFST PCLPS2 provide them.
UFST PCLPS3 does provide.
Passing to Ray to decide what to do with it.
Assign: Ray.
38-01-2 OK
99-01-1 AOK Implementation Dependent: Different default system parameters can be set
if customer desires.
Environment dependent (GenericResourceDir).
99-01-2 AOK Implementation Dependent: Different default device parameters can be set
if customer desires.
Environment dependent (OutputFile).
99-01-3 OK
99-02-1 OK GS OK visually reviewed by RJJ
cpsi/99-02-1 missing
Environment dependent (GenericResourceDir).
99-03-1 OK