Summary: | no output, only an error message for a certain test case | ||
---|---|---|---|
Product: | GhostPCL | Reporter: | Horiana Costea <horiana.costea> |
Component: | PCL interpreter | Assignee: | Default assignee <ghostpdl-bugs> |
Status: | NOTIFIED FIXED | ||
Severity: | normal | ||
Priority: | P2 | ||
Version: | 1.53 | ||
Hardware: | PC | ||
OS: | Windows XP | ||
Customer: | 460 | Word Size: | --- |
Attachments: | pcl6.zip |
Description
Horiana Costea
2009-03-26 08:14:08 UTC
Created attachment 4868 [details]
testcase.pcl
2550x3300 is not a pjl paper size name, use letter. I am sure this command will not select letter, it may be ignored and letter paper is the default. If it does select letter please reopen the bug, I am closing it because the behavior is implementation dependent and we prefer an error. Just for reference, the set of 'known' paper sizes (and the sizes in 300 dpi units) from pcl/pcpage.c is: "executive", 2175, 3150 "letter", 2550, 3300 "legal", 2550, 4200 "ledger", 3300, 5100 "a4", 2480, 3507 "a3", 3507, 4960 "monarch", 1162, 2250 "com_10", 1237, 2850 "dl", 1299, 2598 "c5", 1913, 2704 "b5", 2078, 2952 We tried your suggestion with GhostPCL 1.53 and we used -J"@PJL SET PAPER =letter" instead of -J"@PJL SET PAPER =2550x3300" and there are no more messages regarding of not supported paper, but still we don't get any output, this is in fact the problem. the following message occurs: "Warning interpreter exited with error code -1 Flushing to end of job" Since the document prints ok on an HP laserjet 9000 printer we think that the problem might be in Ghost PCL 1.53 Also we tested with a previous version of Ghost PCL (1.40) and the behavior is different: With Ghost PCL 1.40 we get the output with a problem on it (some text is shifted down). It produces output without error in the current code, I am traveling and don't have access to an HP printer to check correctness. Reassigning to support. Marcos will test on his HP printer (I don't have one) I printed the file on an HP LaserJet 4100dn and an HP Color LaserJet CP3505dn and both produced ~26 pages of output. Both printers were loaded with US letter sized paper. Output from the current GhostPCL head (r9614) is identical to the HP printer output, except for the page size. The HP printers print onto letter size paper and cutoff the bottom of each page; GhostPCL output is 8.27 x 11.69 (inches) (with and without -J"@PJL SET PAPER=LETTER"). We are approaching the CVT (component verification test) start date for our component calling Artifex GhostPCL 1.53, so if this problem is fixed already in 1.54, we would need a patch for 1.53. Thank you. I'm not able to reproduce this with 1.53. The command line: ghostpdl-1.53/main/obj/pcl6 -dNOPAUSE -sDEVICE=pcxmono -r300 -J"@PJL SET PAPER=LETTER" - sOutputFile=%iout.pcx testcase.pcl produces 27 pcx files and no error messages. We are having some difficulties in trying to transform the testcase using Ghost PCL. Since we tried both GhostPCL 1.53 and 1.54 and we've got the same error message we suspect that it could be an issue related with the compiler we are using: Intel Compiler 10.1.021 . Could you please let us know what is the compiler that Artifex team uses when they build on Windows? And also could you please send us output you obtained for the first two pages? Thanks. Created attachment 4924 [details]
out1-2.zip
Output of first two pages
Created attachment 4925 [details]
pcl6.zip
pcl6.exe for Windows created with Microsoft Visual Studio 8.
Note that the "Express Edition" can be downloaded free of charge from
Microsoft.
Could you please tell me how can I download the private file? When I am trying to see the attachment I am getting: "Access to this attachment is restricted." message. Thanks Ahh, I was just about to work on this and realized Ray already did it. testcase.pcl and out1-2.zip are private (restricted) but pcl6.zip is not as far as I can tell. It sounds to me like you should download the free microsoft express edition, like Ray said, then you could do a bit compare of all output files, if you think there are compiler problems. I am reassigning to support if an actual pcl bug unfolds please assign it back to me. We eventually succeeded to perform the transformation. And the output looks ok on both versions 1.53 and 1.54. You can change the resolution of this bug to FIXED. Thanks. |