We received a test file from a customer which indicates a problem related to the signature inside the document. I will attach the test file to this defect. The customer is currently at the 1.40 version of Ghost PCL. We tested this file with the 1.53 to see if the problem occurs but we found that on 1.53 the output looks different than the 1.40 Output. The signature does not appear because we think the requested symbol set (included in the pcl file at the beginning) is not picked up by ghost PCL, so instead of the signature there is the default symbol set selected and some alphabetical characters appear . I'll also attach to this defect a page obtained with 1.53 where the problem can be seen. Could you take a look at this problem and let us know what is wrong with this pcl file? Thanks
Created attachment 5000 [details] test.pcl
Created attachment 5001 [details] out3.bmp
There are 2 problems with this file: ESC&l#V which sets the line spacing is not supported resulting in lines running together on page 1 and the resident courier font is selected instead of the downloaded font for the signature. The first problem we won't fix without a separate contract arranged with Miles Jones. The command code ESC&l#V is documented as obsolete in the pcl implementor's guide and is not even mentioned in modern pcl manuals. The second problem I am working on now. Oddly the font request parameters does match the courier more closely than the downloaded fonts, providing a closer pitch attribute, so I suspect this is a bug in the HP pcl code which we'll try to emulate. These problems are typically more difficult than fixing an incorrect implementation.
I have another question related to this testcase. On Ghost PCL 1.40 output we obtain blank pages one between each page of the testcase. I noticed it contains the page length command; I did not find a detailed description for this command in the PCL Reference. In the testcase the command is l70P which as I understand sets the number of lines per page to 70 (correct me if I'm wrong); removing the command from the testcase causes the blank pages to disappear and the output seems OK My question is since the testcase printed OK with the old 1.38 Ghost PCL and in 1.40 this problem occurs...is there a change regarding this command too? Probably obsolete.....? Because it seems to have an undesired effect in 1.40 Thank you
Could you tell us when you estimate to have a fix for the signature issue? And also could you answer us to the page length command related question? This is a customer testcase and we have to give him a feedback on the problem he reported. Thanks
> Could you tell us when you estimate to have a fix for the signature issue? I haven't studied it carefully enough to give an estimate. The nature of the problem is such that once I have an estimate the fix will probably be hours later, so an estimate is not useful, sorry. > also could you answer us to the page length command related question? I don't know of any implementation of that escape sequence in our code. The failure mode for an unimplemented escape sequence could cause the command to be printed as text - in which case there would be a heuristic to see if the page is dirty, I know that heuristic has changed quite a bit. Sorry to take so long on these issues, your company's priority is relatively low so your problems are bumped down as other customers come in. I don't really have control of that, you should talk to Miles Jones if this is unsatisfactory.
Hello. I would need the signature problem to be fixed by 8/21, since the product being integrated in is going to exit from test. Is this achievable? Thanks
No I don't think that date is possible, even if we started today. I haven't been given the go ahead to start working on this issue at customer priority so I am reassigning to support until that is resolved.
The signature problem is fixed in the current release. There was a residual problem with an extra blank page which is fixed in rev 10259. There is also a separate blank page problem that is only reproducible with the urw fonts because the space character is present in the courier font. We have a contrary request to keep space characters in the urw fonts, so this is being handled with bug #690293, the problem has no impact on the customer who uses Monotype fonts and scaler. We don't have plans to fix the obsolete pcl commands noted in comment #3 which result in the first page (banner) printing incorrectly. I am not sure the original customer wanted this to be fixed, it was never reported as a problem. I am closing this family of bugs since they seem to be addressed in the current code. Please open new bugs such that there is one bug per report or contact Artifex support and we will create the bug reports. Thanks.
Could you please give us a patch for this signature problem? We want to try to apply it on top of GhostPCL 1.53 (this is the version we are using). Thank you
Assigning to me to attempt to find a patch.
Created attachment 5625 [details] patch The attached patch, applicable to GhostPDL 1.53, fixes the missing signature issue.
Created attachment 5626 [details] patch2 Patch referred to in Comment #9 from r10259 that fixes an extra blank page problem.
I believe this issue is now resolved, please reopen if I'm mistaken.
We took the first patch, ported it to GhostPCL 1.53 code and the signature is displayed well. Then, we took the second patch (that was supposed to fix an extra blank page problem), ported it to GhostPCL 1.53 code but the extra blank page is still visible in the output. That's why I am reopening this defect. Maybe the second patch was not correctly identified. Thank you
The output of the test file when processed with GhostPCL does not match the printed output of various HP laserprinters I tested. Reassigning to Henry.
(In reply to comment #16) > The output of the test file when processed with GhostPCL does not match the > printed output of various HP laserprinters I tested. Reassigning to Henry. Testing "head" pcl against a color laserjet 4700 the output is the same modulo a "wontfix" blank page. Do you have a description of the HP output, was there something obviously wrong?
The actual problem from the customer was fixed, the header of this file seems to print differently on different HP printers. A single overprinted line is rendered by Artifex and the HP 4700, other hp printers print a banner page (CP3505 ?). Also we note an extra blank page in the Artifex output.
Hello, The fact that the header page is not displayed right is not a problem for us anymore. But the problem with the extra blank page still persists. We are now at the GhostPCL 8.70 code and it's still there. We are not using urw fonts, instead we use Monotype fco files. Page 5 from the output file is a blank page (page 1 is the header, page 2-4 contain info, page 5 is blank, page 6-7 contain info too). Thanks
Okay I will increase the priority, I have looked at this (extra blank page with orientation changes) in the past and didn't feel I could fix it without high risk of breaking other tests file in the field.
We haven't heard from any of our customers complaining about this problem. We found it during our internal tests. So, I cannot say this problem has a high priority for us. But it might have at a particular moment in time.
The residual blank page problem found in this bug is not a good candidate for the bounty program, making unbountiable.
(In reply to comment #19) > Hello, > > The fact that the header page is not displayed right is not a problem for us > anymore. But the problem with the extra blank page still persists. We are now > at the GhostPCL 8.70 code and it's still there. We are not using urw fonts, > instead we use Monotype fco files. Page 5 from the output file is a blank page > (page 1 is the header, page 2-4 contain info, page 5 is blank, page 6-7 contain > info too). > > Thanks There was some sort of mistake with this problem. The HP4600 also prints a blank page for page 5 when the file is sent to a printer with A4 paper, which is the page size requested on the second page immediately after the job. Please check and reopen if you find different behavior on other HP devices.