Reopen PDFBOX-483?

[steve poling]

Andreas, I am using the trunk version (like before). And I have 
commented out the W/W* lines from the PageDrawer.properties file (like 
before). But I am seeing the same phenomena observed for PDFBOX-483.


I am doing things a little diferently. I am running PrintPDF.java, and I 
have made a kludged hack to get past the PDFBOX-490 issue, but when I 
print, I get just the form's lines and the AcroField text like before. 
I've stepped through the code at PDFTrueTypeFont.java where my kludge 
occurs and nobody's throwing exceptions or anything. (I temporarily 
hard-coded a stream from the TTF file in place of the embedded stream.)


I recall that when we opened PDFBOX-490, I had commented out W/W* and I 
was getting some text, just the wrong typeface. Where should I look to 
see whether the clipping or whatnot responsible for the text going away 
when printed. I first saw this a while back, and text was rendering OK 
to the screen, just not the HP LaserJet.


smiles and cheers,

steve

Re: Reopen PDFBOX-483?

[Andreas Lehmkühler]

Hi,


Gesendet: Mi, 13. Jan 2010 Von: steve poling

> Andreas, I am using the trunk version (like before). And I have 
> commented out the W/W* lines from the PageDrawer.properties file (like 
> before). But I am seeing the same phenomena observed for PDFBOX-483.
I won't reopen PDFBOX-483 without knowing more details on your issue. I could 
be related, 
but probably it's just a sideeffect which only appears if clipping is 
actviated/deactivated.
Is it possible to provide us with a sample document?

> I am doing things a little diferently. I am running PrintPDF.java, and I 
> have made a kludged hack to get past the PDFBOX-490 issue, but when I 
> print, I get just the form's lines and the AcroField text like before. 
> I've stepped through the code at PDFTrueTypeFont.java where my kludge 
> occurs and nobody's throwing exceptions or anything. (I temporarily 
> hard-coded a stream from the TTF file in place of the embedded stream.)
Hmmm, I've made similar experiences in an other case [1]. Everything worked 
fine displaying the pdf
using PDFReader, but if I tried to print the pdf some of the content got lost.
Perhaps you're running in the same issue, or perhaps your hack is the problem? 

> I recall that when we opened PDFBOX-490, I had commented out W/W* and I 
> was getting some text, just the wrong typeface. Where should I look to 
> see whether the clipping or whatnot responsible for the text going away 
> when printed. I first saw this a while back, and text was rendering OK 
> to the screen, just not the HP LaserJet.
During my investigations concerning [1] I've learned that displaying and 
printing could be quite different things.
Both are using the (PDFBox) code for rendering but the used Graphics2d (there 
are different classes used for
displaying and printing within the JDK). Furthermore the result may depend on 
the used printer driver/OS/JDK.
What environment are you using?

> smiles and cheers,
> 
> steve

BR
Andreas Lehmkühler


[1] https://issues.apache.org/jira/browse/PDFBOX-552

Re: Reopen PDFBOX-483?

[steve poling]

Andreas,

You said:
I won't reopen PDFBOX-483 without knowing more details on your issue. I could be related, 
but probably it's just a sideeffect which only appears if clipping is actviated/deactivated.

Is it possible to provide us with a sample document?
  


That's only fair. You've got the document. I'm using the PDF that I 
attached to PDFBOX-490. (Sorry I didn't say so earlier.)


If memory serves, when I opened PDFBOX-490, I got the current behavior 
until I commented out W/W* whereupon I started getting text output, but 
in the wrong typeface.


Is there any place I should look in the code to follow up on your 
clipping activated/deactivated conjecture?




Hmmm, I've made similar experiences in an other case [1]. Everything worked 
fine displaying the pdf
using PDFReader, but if I tried to print the pdf some of the content got lost.
Perhaps you're running in the same issue, or perhaps your hack is the problem? 
  


The possibility that my hack is at fault weighs heavily on my mind. I'll 
drop back to the current trunk version and repeat my experiment. (I'm 
not seeing any NullPointerException as described in PDFBOX-552.)





During my investigations concerning [1] I've learned that displaying and 
printing could be quite different things.
Both are using the (PDFBox) code for rendering but the used Graphics2d (there 
are different classes used for
displaying and printing within the JDK). Furthermore the result may depend on 
the used printer driver/OS/JDK.
What environment are you using?
  


I feel you're onto something about the distinction between printer vs 
display draw code.


I'm running Windows XP, JDK 1.6.0_17, and printing to the HP LaserJet 4.

Re: Reopen PDFBOX-483?

[steve poling]

Hello again,

I got distracted by other work from this issue and I've returned to it 
today. Here are the experiments I've performed:


1) view PDF with PdfReader (it renders correctly)
2) print PDF to HP LaserJet 4P (it renders with many lines and text omitted)
3) comment-out W/W* in PageDrawer.properties
4) restart PdfReader and repeat #1 and #2 (with same result)

I archived all my code changes and retrieved the latest sources from 
svn. So, I should be running the latest and greatest code.


If you goto PDFBOX-490 
, you'll find attached 
file filled.pdf that manifests this error, but I've been seeing this 
with a lot of different PDFs: display looks good, print looks bad. I can 
attach another file to PDFBOX-483 
 if you'd like.


Can you point me to where I should look to see why text appears on 
screen, but not on paper?


Thanks in advance,

steve

Re: Reopen PDFBOX-483?

[Andreas Lehmkuehler]

Hi,


steve poling schrieb:

Hello again,

I got distracted by other work from this issue and I've returned to it 
today. Here are the experiments I've performed:


1) view PDF with PdfReader (it renders correctly)
2) print PDF to HP LaserJet 4P (it renders with many lines and text 
omitted)

3) comment-out W/W* in PageDrawer.properties
4) restart PdfReader and repeat #1 and #2 (with same result)

OK, so the clipping has nothing to do with that issue.

I archived all my code changes and retrieved the latest sources from 
svn. So, I should be running the latest and greatest code.


If you goto PDFBOX-490 
, you'll find attached 
file filled.pdf that manifests this error, but I've been seeing this 
with a lot of different PDFs: display looks good, print looks bad. I can 
attach another file to PDFBOX-483 
 if you'd like.
I've tried that pdf and it works like a charm except for some misplaced 
characters. I'm using ubuntu linux, java 1.6.0_15 32bit and a HP Laserjet 2550N.


Correct me if I'm wrong, but your Laserjet 4P is a quite old model, isn't it?
Did you update the driver? Has it a postscript option or is it a pcl only model?
How much memory is installed to the printer? Probably that could be a potential
bottleneck.

Can you point me to where I should look to see why text appears on 
screen, but not on paper?
Hmmm, good question. Perhaps you should try to find differences between working 
documents and those which don't work. Or you will probably find something that 
documents have in common which don't work, e.g. font tpye (true type, type1, 
etc.), fonts are embedded or not, font encoding (ansi, CID, identity-H, etc.), 
acroform included or not 


BR
Andreas Lehmkühler

Re: Reopen PDFBOX-483?

[Andreas Lehmkuehler]

Hi,

Andreas Lehmkuehler schrieb:
If you goto PDFBOX-490 
, you'll find 
attached file filled.pdf that manifests this error, but I've been 
seeing this with a lot of different PDFs: display looks good, print 
looks bad. I can attach another file to PDFBOX-483 
 if you'd like.
I've tried that pdf and it works like a charm except for some misplaced 
characters. I'm using ubuntu linux, java 1.6.0_15 32bit and a HP 
Laserjet 2550N.
I've made another test on my MacBook (MacOSX 10.6., jdk 1.6.0_17 64bit, same 
printer) and it works well too.



BR
Andreas Lehmkühler

Re: Reopen PDFBOX-483?

[steve poling]

Andreas,
I archived all my code changes and retrieved the latest sources from 
svn. So, I should be running the latest and greatest code.


If you goto PDFBOX-490 
, you'll find 
attached file filled.pdf that manifests this error, but I've been 
seeing this with a lot of different PDFs: display looks good, print 
looks bad. I can attach another file to PDFBOX-483 
 if you'd like.
I've tried that pdf and it works like a charm except for some 
misplaced characters. I'm using ubuntu linux, java 1.6.0_15 32bit and 
a HP Laserjet 2550N.


Correct me if I'm wrong, but your Laserjet 4P is a quite old model, 
isn't it?
Did you update the driver? Has it a postscript option or is it a pcl 
only model?
How much memory is installed to the printer? Probably that could be a 
potential

bottleneck.


My printer is as old as the hills.  However, I've been seeing this 
issue when I print to other devices, such as the Canon iR 5570. I'll 
perform my experiment on all the different printers I can find. If 
memory serves (and sometimes it doesn't) I ran the experiment on the HP 
990cxi, too (with the same effect). I am using Windows XP. So, that's a 
difference (between working for you and breaking for me). Keep in mind, 
the copy on the display looks grand, it's just the printed output that 
looks bad. I hate working bugs that require killing trees to reproduce.


Can you point me to where I should look to see why text appears on 
screen, but not on paper?
Hmmm, good question. Perhaps you should try to find differences 
between working documents and those which don't work. Or you will 
probably find something that documents have in common which don't 
work, e.g. font tpye (true type, type1, etc.), fonts are embedded or 
not, font encoding (ansi, CID, identity-H, etc.), acroform included or 
not 


I'll give that a go. Can you point me to a test repository of PDF files 
of ascending difficulty? Does such a thing exist or would it be 
worthwhile to create one?


Thanks again,

steve

Re: Reopen PDFBOX-483?

[Daniel Wilson]

These are not in ascending difficulty ... but the files we currently test
for rendering are at trunk\src\test\resources\input\rendering

Daniel

On Sun, Mar 7, 2010 at 11:11 PM, steve poling  wrote:

> Andreas,
>
>  I archived all my code changes and retrieved the latest sources from svn.
>>> So, I should be running the latest and greatest code.
>>>
>>> If you goto PDFBOX-490 ,
>>> you'll find attached file filled.pdf that manifests this error, but I've
>>> been seeing this with a lot of different PDFs: display looks good, print
>>> looks bad. I can attach another file to PDFBOX-483 <
>>> https://issues.apache.org/jira/browse/PDFBOX-483> if you'd like.
>>>
>> I've tried that pdf and it works like a charm except for some misplaced
>> characters. I'm using ubuntu linux, java 1.6.0_15 32bit and a HP Laserjet
>> 2550N.
>>
>> Correct me if I'm wrong, but your Laserjet 4P is a quite old model, isn't
>> it?
>> Did you update the driver? Has it a postscript option or is it a pcl only
>> model?
>> How much memory is installed to the printer? Probably that could be a
>> potential
>> bottleneck.
>>
>
> My printer is as old as the hills.  However, I've been seeing this
> issue when I print to other devices, such as the Canon iR 5570. I'll perform
> my experiment on all the different printers I can find. If memory serves
> (and sometimes it doesn't) I ran the experiment on the HP 990cxi, too (with
> the same effect). I am using Windows XP. So, that's a difference (between
> working for you and breaking for me). Keep in mind, the copy on the display
> looks grand, it's just the printed output that looks bad. I hate working
> bugs that require killing trees to reproduce.
>
>
>  Can you point me to where I should look to see why text appears on screen,
>>> but not on paper?
>>>
>> Hmmm, good question. Perhaps you should try to find differences between
>> working documents and those which don't work. Or you will probably find
>> something that documents have in common which don't work, e.g. font tpye
>> (true type, type1, etc.), fonts are embedded or not, font encoding (ansi,
>> CID, identity-H, etc.), acroform included or not 
>>
>
> I'll give that a go. Can you point me to a test repository of PDF files of
> ascending difficulty? Does such a thing exist or would it be worthwhile to
> create one?
>
> Thanks again,
>
> steve
>

Re: Reopen PDFBOX-483?

[steve poling]

Andreas Lehmkuehler schrieb:
If you goto PDFBOX-490 
, you'll find 
attached file filled.pdf that manifests this error, but I've been 
seeing this with a lot of different PDFs: display looks good, print 
looks bad. I can attach another file to PDFBOX-483 
 if you'd like.
I've tried that pdf and it works like a charm except for some 
misplaced characters. I'm using ubuntu linux, java 1.6.0_15 32bit and 
a HP Laserjet 2550N.
I've made another test on my MacBook (MacOSX 10.6., jdk 1.6.0_17 
64bit, same printer) and it works well too.


I'd like to know if anyone has repeated the experiment on any 
Windows-based platform, since Ubuntu and OSX are both Linux-based. If 
someone else can reproduce the failure on Windows, I'll start trusting 
my sanity again.

Re: Reopen PDFBOX-483?

[Andreas Lehmkuehler]

Hi,

steve poling schrieb:

Andreas Lehmkuehler schrieb:
If you goto PDFBOX-490 
, you'll find 
attached file filled.pdf that manifests this error, but I've been 
seeing this with a lot of different PDFs: display looks good, print 
looks bad. I can attach another file to PDFBOX-483 
 if you'd like.
I've tried that pdf and it works like a charm except for some 
misplaced characters. I'm using ubuntu linux, java 1.6.0_15 32bit and 
a HP Laserjet 2550N.
I've made another test on my MacBook (MacOSX 10.6., jdk 1.6.0_17 
64bit, same printer) and it works well too.


I'd like to know if anyone has repeated the experiment on any 
Windows-based platform, since Ubuntu and OSX are both Linux-based. If 
someone else can reproduce the failure on Windows, I'll start trusting 
my sanity again.

I'm a software development for a lot of years and sometimes it leads to
insanity, but we all have to do our best not to end in the programmers
nuthouse ;-))

I'll see if I can find some time to run that test on my rarely used windows box.

BR
Andreas Lehmkühler

Re: Reopen PDFBOX-483?

[Maruan Sahyoun]

Hi Andreas,

I can do a test on our Windows test server (Windows 2003, 32bit) and let you 
know the results around lunch time (german time) if that helps

Maruan Sahyoun

Am 09.03.2010 um 08:11 schrieb Andreas Lehmkuehler:

> Hi,
> 
> steve poling schrieb:
>> Andreas Lehmkuehler schrieb:
> If you goto PDFBOX-490 
> , you'll find attached 
> file filled.pdf that manifests this error, but I've been seeing this with 
> a lot of different PDFs: display looks good, print looks bad. I can 
> attach another file to PDFBOX-483 
>  if you'd like.
 I've tried that pdf and it works like a charm except for some misplaced 
 characters. I'm using ubuntu linux, java 1.6.0_15 32bit and a HP Laserjet 
 2550N.
>>> I've made another test on my MacBook (MacOSX 10.6., jdk 1.6.0_17 64bit, 
>>> same printer) and it works well too.
>> I'd like to know if anyone has repeated the experiment on any Windows-based 
>> platform, since Ubuntu and OSX are both Linux-based. If someone else can 
>> reproduce the failure on Windows, I'll start trusting my sanity again.
> I'm a software development for a lot of years and sometimes it leads to
> insanity, but we all have to do our best not to end in the programmers
> nuthouse ;-))
> 
> I'll see if I can find some time to run that test on my rarely used windows 
> box.
> 
> BR
> Andreas Lehmkühler
> 

Re: Reopen PDFBOX-483?

[Maruan Sahyoun]

Hi,

please find enclosed the result of the printing test conducted on 

Windows 2003 Server SP2 32 bit, Java 1.5 using a fresh built from trunk. The 
test was done using the Adobe PDF printer driver as well as Apple and HP 
Postscript printers with similar results.

kind regards

Maruan Sahyoun





Am 09.03.2010 um 09:49 schrieb Andreas Lehmkühler:

> Hi,
> 
> Betreff: Re: Reopen PDFBOX-483?
> Gesendet: Di, 09. Mrz 2010
> Von: Maruan Sahyoun
> 
>> Hi Andreas,
>> 
>> I can do a test on our Windows test server (Windows 2003, 32bit) and let you
>> know the results around lunch time (german time) if that helps
> Yeah, that would be great.
> 
> BR
> Andreas Lehmkühler
> 
>> Maruan Sahyoun
>> 
>> Am 09.03.2010 um 08:11 schrieb Andreas Lehmkuehler:
>> 
>>> Hi,
>>> 
>>> steve poling schrieb:
>>>> Andreas Lehmkuehler schrieb:
>>>>>>> If you goto PDFBOX-490
>> <https://issues.apache.org/jira/browse/PDFBOX-490>, you'll find attached
>> file filled.pdf that manifests this error, but I've been seeing this with a
>> lot of different PDFs: display looks good, print looks bad. I can attach
>> another file to PDFBOX-483
>> <https://issues.apache.org/jira/browse/PDFBOX-483> if you'd like.
>>>>>> I've tried that pdf and it works like a charm except for some misplaced
>> characters. I'm using ubuntu linux, java 1.6.0_15 32bit and a HP Laserjet
>> 2550N.
>>>>> I've made another test on my MacBook (MacOSX 10.6., jdk 1.6.0_17 64bit,
>> same printer) and it works well too.
>>>> I'd like to know if anyone has repeated the experiment on any
>> Windows-based platform, since Ubuntu and OSX are both Linux-based. If
>> someone else can reproduce the failure on Windows, I'll start trusting my
>> sanity again.
>>> I'm a software development for a lot of years and sometimes it leads to
>>> insanity, but we all have to do our best not to end in the programmers
>>> nuthouse ;-))
>>> 
>>> I'll see if I can find some time to run that test on my rarely used
>> windows box.
>>> 
>>> BR
>>> Andreas Lehmkühler
>>> 
>> 
>> 
> 
> --- original Nachricht Ende 
> 

Re: Reopen PDFBOX-483?

[Maruan Sahyoun]

Hi Andreas,

yes, the results are similar BUT most of the text and some of the lines are 
missing. Converting to Image output using PDFToImage provides a different and 
much better result where all text and lines are included and only some 
misplacement occurs. Is there a way to submit the attachment so you can see for 
yourself?

Maruan Sahyoun

Am 09.03.2010 um 13:38 schrieb Andreas Lehmkühler:

> Hi,
> 
> Betreff: Re: Reopen PDFBOX-483?
> Gesendet: Di, 09. Mrz 2010
> Von: Maruan Sahyoun
> 
>> Hi,
>> 
>> please find enclosed the result of the printing test conducted on 
>> 
>> Windows 2003 Server SP2 32 bit, Java 1.5 using a fresh built from trunk. The
>> test was done using the Adobe PDF printer driver as well as Apple and HP
>> Postscript printers with similar results.
> Thanks for testing. Your attachments didn't make it due to some restrictions 
> of the mailing list.
> Probably it would be sufficient to describe the results. Let me guess, they 
> are all similar. All
> contain text, some characters are misplaced and a wrong font is used.
> 
> BR
> Andreas Lehmkühler
> 

RE: Reopen PDFBOX-483?

[Martinez, Mel - 1004 - MITLL]

Minor correction - OSX is based on Darwin which is based on BSD, not Linux.
Your basic point is not about that, so I know this is just a nit.  :-)

-Original Message-
From: steve poling [mailto:s...@i2k.com] 
Sent: Monday, March 08, 2010 11:57 PM
To: dev@pdfbox.apache.org
Subject: Re: Reopen PDFBOX-483?

Andreas Lehmkuehler schrieb:
>>> If you goto PDFBOX-490 
>>> <https://issues.apache.org/jira/browse/PDFBOX-490>, you'll find 
>>> attached file filled.pdf that manifests this error, but I've been 
>>> seeing this with a lot of different PDFs: display looks good, print 
>>> looks bad. I can attach another file to PDFBOX-483 
>>> <https://issues.apache.org/jira/browse/PDFBOX-483> if you'd like.
>> I've tried that pdf and it works like a charm except for some 
>> misplaced characters. I'm using ubuntu linux, java 1.6.0_15 32bit and 
>> a HP Laserjet 2550N.
> I've made another test on my MacBook (MacOSX 10.6., jdk 1.6.0_17 
> 64bit, same printer) and it works well too.

I'd like to know if anyone has repeated the experiment on any 
Windows-based platform, since Ubuntu and OSX are both Linux-based. If 
someone else can reproduce the failure on Windows, I'll start trusting 
my sanity again.


smime.p7s
Description: S/MIME cryptographic signature

Re: Reopen PDFBOX-483?

[Maruan Sahyoun]

Hi,

FYI - using PDFReader the PDF is displayed OK but when printed the same results 
are produced as with PrintPDF. The printed output contains the variable data 
only (and some lines), Boilerplate text is not printed.  

Maruan Sahyoun

Am 09.03.2010 um 13:58 schrieb Andreas Lehmkühler:

> Hi,
> 
> Betreff: Re: Reopen PDFBOX-483?
> Gesendet: Di, 09. Mrz 2010
> Von: Maruan Sahyoun
> 
>> Hi ,
>> 
>> please find enclosed the text extracted from the printed PDF. Extraction was
>> done using Adobe Acrobat 8.
>> 
>> X0X0X0 X0X0X05
>> X0X0X0 X0X0X05
>> X0X0X0 X0X0X05 
>> X0X0X05 MM/DD/ X0X2 
>> X0X2 
>> X0X0X0X X0X0X0X
>> X0X0X05 X0X0X05
>> X0X0X05 X0X0X05
>> X0X0X0X X0X0X05 
>> X0X0X05 
>> 
>> X0X0 
>> X0X0 
>> X05 X0X0X05 
>> MM/DD/ X0X0X05 X0X0 
>> 
> Hmm, that's odd. I'll run my own tests later when I'm at home. Finally that 
> seems to be a windows only issue. I'll also file an issue on JIRA
> 
> Thanks for the tests!
> 
> BR
> Andreas Lehmkühler
> 
>> 
>> Maruan Sahyoun
>> 
>> 
>> 
>> Geschäftsführer: Maruan Sahyoun
>> Handelsregister: AG Düsseldorf, HRB 53837
>> UST.-ID: DE248275827
>> 
>> Am 09.03.2010 um 13:45 schrieb Maruan Sahyoun:
>> 
>>> Hi Andreas,
>>> 
>>> yes, the results are similar BUT most of the text and some of the lines
>> are missing. Converting to Image output using PDFToImage provides a
>> different and much better result where all text and lines are included and
>> only some misplacement occurs. Is there a way to submit the attachment so
>> you can see for yourself?
>>> 
>>> Maruan Sahyoun
>>> 
>>> Am 09.03.2010 um 13:38 schrieb Andreas Lehmkühler:
>>> 
>>>> Hi,
>>>> 
>>>> Betreff: Re: Reopen PDFBOX-483?
>>>> Gesendet: Di, 09. Mrz 2010
>>>> Von: Maruan Sahyoun
>>>> 
>>>>> Hi,
>>>>> 
>>>>> please find enclosed the result of the printing test conducted on 
>>>>> 
>>>>> Windows 2003 Server SP2 32 bit, Java 1.5 using a fresh built from trunk.
>> The
>>>>> test was done using the Adobe PDF printer driver as well as Apple and
>> HP
>>>>> Postscript printers with similar results.
>>>> Thanks for testing. Your attachments didn't make it due to some
>> restrictions of the mailing list.
>>>> Probably it would be sufficient to describe the results. Let me guess,
>> they are all similar. All
>>>> contain text, some characters are misplaced and a wrong font is used.
>>>> 
>>>> BR
>>>> Andreas Lehmkühler
>>>> 
>>> 
>> 
>> 
> 
> --- original Nachricht Ende 
> 

Re: Reopen PDFBOX-483?

[Maruan Sahyoun]

Hi ,

please find enclosed the text extracted from the printed PDF. Extraction was 
done using Adobe Acrobat 8.

X0X0X0 X0X0X05 
X0X0X0 X0X0X05 
X0X0X0 X0X0X05 
X0X0X05 MM/DD/ X0X2 
X0X2 
X0X0X0X X0X0X0X 
X0X0X05 X0X0X05 
X0X0X05 X0X0X05 
X0X0X0X X0X0X05 
X0X0X05 

X0X0 
X0X0 
X05 X0X0X05 
MM/DD/ X0X0X05 X0X0 


Maruan Sahyoun



Geschäftsführer: Maruan Sahyoun
Handelsregister: AG Düsseldorf, HRB 53837
UST.-ID: DE248275827

Am 09.03.2010 um 13:45 schrieb Maruan Sahyoun:

> Hi Andreas,
> 
> yes, the results are similar BUT most of the text and some of the lines are 
> missing. Converting to Image output using PDFToImage provides a different and 
> much better result where all text and lines are included and only some 
> misplacement occurs. Is there a way to submit the attachment so you can see 
> for yourself?
> 
> Maruan Sahyoun
> 
> Am 09.03.2010 um 13:38 schrieb Andreas Lehmkühler:
> 
>> Hi,
>> 
>> Betreff: Re: Reopen PDFBOX-483?
>> Gesendet: Di, 09. Mrz 2010
>> Von: Maruan Sahyoun
>> 
>>> Hi,
>>> 
>>> please find enclosed the result of the printing test conducted on 
>>> 
>>> Windows 2003 Server SP2 32 bit, Java 1.5 using a fresh built from trunk. The
>>> test was done using the Adobe PDF printer driver as well as Apple and HP
>>> Postscript printers with similar results.
>> Thanks for testing. Your attachments didn't make it due to some restrictions 
>> of the mailing list.
>> Probably it would be sufficient to describe the results. Let me guess, they 
>> are all similar. All
>> contain text, some characters are misplaced and a wrong font is used.
>> 
>> BR
>> Andreas Lehmkühler
>> 
> 

Re: Reopen PDFBOX-483?

[Andreas Lehmkuehler]

Hi,

steve poling schrieb:

Andreas Lehmkuehler schrieb:
If you goto PDFBOX-490 
, you'll find 
attached file filled.pdf that manifests this error, but I've been 
seeing this with a lot of different PDFs: display looks good, print 
looks bad. I can attach another file to PDFBOX-483 
 if you'd like.
I've tried that pdf and it works like a charm except for some 
misplaced characters. I'm using ubuntu linux, java 1.6.0_15 32bit and 
a HP Laserjet 2550N.
I've made another test on my MacBook (MacOSX 10.6., jdk 1.6.0_17 
64bit, same printer) and it works well too.


I'd like to know if anyone has repeated the experiment on any 
Windows-based platform, since Ubuntu and OSX are both Linux-based. If 
someone else can reproduce the failure on Windows, I'll start trusting 
my sanity again.

Good news Steve you're obviously not insane. ;-) Maruan confirmed your issue on
W2K and I've tested it on my WinXP with jdk 1.6.0_13 with the same result. The
print looks bad. I have no explanation yet, except that it seems to be windows
only. For now I don't have a clue where to look. Perhaps I will have an idea in
a few days ...

BR
Andreas Lehmkühler

Re: Reopen PDFBOX-483?

[steve poling]

I said "Linux" but I was thinking "Unix." 

we work in a field where nits can bite you in the butt.

Martinez, Mel - 1004 - MITLL wrote:

Minor correction - OSX is based on Darwin which is based on BSD, not Linux.
Your basic point is not about that, so I know this is just a nit.  :-)
  

Re: Reopen PDFBOX-483?

[Maruan Sahyoun]

Hi,

I did some initial "debugging" and it seems that the content of the form fields 
(date part) is being printed but the form template itself being held in 
Pages:Kids:Resources:XObject are not printed. Unfortunately as I'm currently in 
the stage of learning about the PDFBox code at that point in time I can't 
provide more help.

Kind regards

Maruan



Am 09.03.2010 um 21:01 schrieb Andreas Lehmkuehler:

> Hi,
> 
> steve poling schrieb:
>> Andreas Lehmkuehler schrieb:
> If you goto PDFBOX-490 
> , you'll find attached 
> file filled.pdf that manifests this error, but I've been seeing this with 
> a lot of different PDFs: display looks good, print looks bad. I can 
> attach another file to PDFBOX-483 
>  if you'd like.
 I've tried that pdf and it works like a charm except for some misplaced 
 characters. I'm using ubuntu linux, java 1.6.0_15 32bit and a HP Laserjet 
 2550N.
>>> I've made another test on my MacBook (MacOSX 10.6., jdk 1.6.0_17 64bit, 
>>> same printer) and it works well too.
>> I'd like to know if anyone has repeated the experiment on any Windows-based 
>> platform, since Ubuntu and OSX are both Linux-based. If someone else can 
>> reproduce the failure on Windows, I'll start trusting my sanity again.
> Good news Steve you're obviously not insane. ;-) Maruan confirmed your issue 
> on
> W2K and I've tested it on my WinXP with jdk 1.6.0_13 with the same result. The
> print looks bad. I have no explanation yet, except that it seems to be windows
> only. For now I don't have a clue where to look. Perhaps I will have an idea 
> in
> a few days ...
> 
> BR
> Andreas Lehmkühler
> 

Re: Reopen PDFBOX-483?

[Andreas Lehmkuehler]

Hi,


Maruan Sahyoun schrieb:

Hi,

FYI - using PDFReader the PDF is displayed OK but when printed the same results are produced as with PrintPDF. The printed output contains the variable data only (and some lines), Boilerplate text is not printed.  

That was a hard nut to crack, but I guess it's done. With resolving PDFBOX-632
it works for me on WINDOWS.!!

BR
Andreas Lehmkühler



Maruan Sahyoun

Am 09.03.2010 um 13:58 schrieb Andreas Lehmkühler:


Hi,

Betreff: Re: Reopen PDFBOX-483?
Gesendet: Di, 09. Mrz 2010
Von: Maruan Sahyoun


Hi ,

please find enclosed the text extracted from the printed PDF. Extraction was
done using Adobe Acrobat 8.

X0X0X0 X0X0X05
X0X0X0 X0X0X05
X0X0X0 X0X0X05 
X0X0X05 MM/DD/ X0X2 
X0X2 
X0X0X0X X0X0X0X

X0X0X05 X0X0X05
X0X0X05 X0X0X05
X0X0X0X X0X0X05 
X0X0X05 

X0X0 
X0X0 
X05 X0X0X05 
MM/DD/ X0X0X05 X0X0 


Hmm, that's odd. I'll run my own tests later when I'm at home. Finally that 
seems to be a windows only issue. I'll also file an issue on JIRA

Thanks for the tests!

BR
Andreas Lehmkühler


Maruan Sahyoun



Geschäftsführer: Maruan Sahyoun
Handelsregister: AG Düsseldorf, HRB 53837
UST.-ID: DE248275827

Am 09.03.2010 um 13:45 schrieb Maruan Sahyoun:


Hi Andreas,

yes, the results are similar BUT most of the text and some of the lines

are missing. Converting to Image output using PDFToImage provides a
different and much better result where all text and lines are included and
only some misplacement occurs. Is there a way to submit the attachment so
you can see for yourself?

Maruan Sahyoun

Am 09.03.2010 um 13:38 schrieb Andreas Lehmkühler:


Hi,

Betreff: Re: Reopen PDFBOX-483?
Gesendet: Di, 09. Mrz 2010
Von: Maruan Sahyoun


Hi,

please find enclosed the result of the printing test conducted on 


Windows 2003 Server SP2 32 bit, Java 1.5 using a fresh built from trunk.

The

test was done using the Adobe PDF printer driver as well as Apple and

HP

Postscript printers with similar results.

Thanks for testing. Your attachments didn't make it due to some

restrictions of the mailing list.

Probably it would be sufficient to describe the results. Let me guess,

they are all similar. All

contain text, some characters are misplaced and a wrong font is used.

BR
Andreas Lehmkühler




--- original Nachricht Ende 

Re: Reopen PDFBOX-483?

[Maruan Sahyoun]

Hi Andreas,

that's good news. Congrats that you found the issue.

Maruan Sahyoun

Am 23.03.2010 um 20:00 schrieb Andreas Lehmkuehler:

> Hi,
> 
> 
> Maruan Sahyoun schrieb:
>> Hi,
>> FYI - using PDFReader the PDF is displayed OK but when printed the same 
>> results are produced as with PrintPDF. The printed output contains the 
>> variable data only (and some lines), Boilerplate text is not printed.  
> That was a hard nut to crack, but I guess it's done. With resolving PDFBOX-632
> it works for me on WINDOWS.!!
> 
> BR
> Andreas Lehmkühler
> 
>> Maruan Sahyoun
>> Am 09.03.2010 um 13:58 schrieb Andreas Lehmkühler:
>>> Hi,
>>> 
>>> Betreff: Re: Reopen PDFBOX-483?
>>> Gesendet: Di, 09. Mrz 2010
>>> Von: Maruan Sahyoun
>>> 
>>>> Hi ,
>>>> 
>>>> please find enclosed the text extracted from the printed PDF. Extraction 
>>>> was
>>>> done using Adobe Acrobat 8.
>>>> 
>>>> X0X0X0 X0X0X05
>>>> X0X0X0 X0X0X05
>>>> X0X0X0 X0X0X05 
>>>> X0X0X05 MM/DD/ X0X2 X0X2 
>>>> X0X0X0X X0X0X0X
>>>> X0X0X05 X0X0X05
>>>> X0X0X05 X0X0X05
>>>> X0X0X0X X0X0X05 
>>>> X0X0X05 
>>>> X0X0 X0X0 X05 
>>>> X0X0X05 MM/DD/ 
>>>> X0X0X05 X0X0 
>>> Hmm, that's odd. I'll run my own tests later when I'm at home. Finally that 
>>> seems to be a windows only issue. I'll also file an issue on JIRA
>>> 
>>> Thanks for the tests!
>>> 
>>> BR
>>> Andreas Lehmkühler
>>> 
>>>> Maruan Sahyoun
>>>> 
>>>> 
>>>> 
>>>> Geschäftsführer: Maruan Sahyoun
>>>> Handelsregister: AG Düsseldorf, HRB 53837
>>>> UST.-ID: DE248275827
>>>> 
>>>> Am 09.03.2010 um 13:45 schrieb Maruan Sahyoun:
>>>> 
>>>>> Hi Andreas,
>>>>> 
>>>>> yes, the results are similar BUT most of the text and some of the lines
>>>> are missing. Converting to Image output using PDFToImage provides a
>>>> different and much better result where all text and lines are included and
>>>> only some misplacement occurs. Is there a way to submit the attachment so
>>>> you can see for yourself?
>>>>> Maruan Sahyoun
>>>>> 
>>>>> Am 09.03.2010 um 13:38 schrieb Andreas Lehmkühler:
>>>>> 
>>>>>> Hi,
>>>>>> 
>>>>>> Betreff: Re: Reopen PDFBOX-483?
>>>>>> Gesendet: Di, 09. Mrz 2010
>>>>>> Von: Maruan Sahyoun
>>>>>> 
>>>>>>> Hi,
>>>>>>> 
>>>>>>> please find enclosed the result of the printing test conducted on 
>>>>>>> Windows 2003 Server SP2 32 bit, Java 1.5 using a fresh built from trunk.
>>>> The
>>>>>>> test was done using the Adobe PDF printer driver as well as Apple and
>>>> HP
>>>>>>> Postscript printers with similar results.
>>>>>> Thanks for testing. Your attachments didn't make it due to some
>>>> restrictions of the mailing list.
>>>>>> Probably it would be sufficient to describe the results. Let me guess,
>>>> they are all similar. All
>>>>>> contain text, some characters are misplaced and a wrong font is used.
>>>>>> 
>>>>>> BR
>>>>>> Andreas Lehmkühler
>>>>>> 
>>>> 
>>> --- original Nachricht Ende 
>>> 
> 

Re: Re: Reopen PDFBOX-483?

[Andreas Lehmkühler]

Hi,

Gesendet: Fr, 15. Jan 2010 Von: steve poling

> Andreas,
> 
> You said:
> > I won't reopen PDFBOX-483 without knowing more details on your issue. I
> could be related, 
> > but probably it's just a sideeffect which only appears if clipping is
> actviated/deactivated.
> > Is it possible to provide us with a sample document?
> >   
> 
> That's only fair. You've got the document. I'm using the PDF that I 
> attached to PDFBOX-490. (Sorry I didn't say so earlier.)
Ah, ok.

> If memory serves, when I opened PDFBOX-490, I got the current behavior 
> until I commented out W/W* whereupon I started getting text output, but 
> in the wrong typeface.
> 
> Is there any place I should look in the code to follow up on your 
> clipping activated/deactivated conjecture?
It's not that easy. You have to have a look at the pdf and find the part with 
the clipping 
operator to understand what part is clipped and what happens if you 
activate/deactivate it.

> > Hmmm, I've made similar experiences in an other case [1]. Everything
> worked fine displaying the pdf
> > using PDFReader, but if I tried to print the pdf some of the content got
> lost.
> > Perhaps you're running in the same issue, or perhaps your hack is the
> problem? 
> >   
> 
> The possibility that my hack is at fault weighs heavily on my mind. I'll 
> drop back to the current trunk version and repeat my experiment. (I'm 
> not seeing any NullPointerException as described in PDFBOX-552.)
Please run some test with the following environments:

- trunk version, clipping activated/deactivated
- trunk version including your hack, clipping activated/deactivated

Please describe the different results for displaying and printing the pdf.

> > During my investigations concerning [1] I've learned that displaying and
> printing could be quite different things.
> > Both are using the (PDFBox) code for rendering but the used Graphics2d
> (there are different classes used for
> > displaying and printing within the JDK). Furthermore the result may depend
> on the used printer driver/OS/JDK.
> > What environment are you using?
> >   
> 
> I feel you're onto something about the distinction between printer vs 
> display draw code.
> 
> I'm running Windows XP, JDK 1.6.0_17, and printing to the HP LaserJet 4.
Yes, the behaviour of the JDK-part could be different, the PDFBox part is 
inboth cases the same.
But before we dig deeper into that, I'd like to see the results of the tests 
described above.

As I'm on a 2 week vacancy in denmark, I can't run my own tests.

BR
Andreas Lehmkühler

Re: Re: Reopen PDFBOX-483?

[Andreas Lehmkühler]

Hi,

Betreff: Re: Reopen PDFBOX-483?
Gesendet: Di, 09. Mrz 2010
Von: Maruan Sahyoun

> Hi Andreas,
> 
> I can do a test on our Windows test server (Windows 2003, 32bit) and let you
> know the results around lunch time (german time) if that helps
Yeah, that would be great.

BR
Andreas Lehmkühler

> Maruan Sahyoun
> 
> Am 09.03.2010 um 08:11 schrieb Andreas Lehmkuehler:
> 
> > Hi,
> > 
> > steve poling schrieb:
> >> Andreas Lehmkuehler schrieb:
> >>>>> If you goto PDFBOX-490
> <https://issues.apache.org/jira/browse/PDFBOX-490>, you'll find attached
> file filled.pdf that manifests this error, but I've been seeing this with a
> lot of different PDFs: display looks good, print looks bad. I can attach
> another file to PDFBOX-483
> <https://issues.apache.org/jira/browse/PDFBOX-483> if you'd like.
> >>>> I've tried that pdf and it works like a charm except for some misplaced
> characters. I'm using ubuntu linux, java 1.6.0_15 32bit and a HP Laserjet
> 2550N.
> >>> I've made another test on my MacBook (MacOSX 10.6., jdk 1.6.0_17 64bit,
> same printer) and it works well too.
> >> I'd like to know if anyone has repeated the experiment on any
> Windows-based platform, since Ubuntu and OSX are both Linux-based. If
> someone else can reproduce the failure on Windows, I'll start trusting my
> sanity again.
> > I'm a software development for a lot of years and sometimes it leads to
> > insanity, but we all have to do our best not to end in the programmers
> > nuthouse ;-))
> > 
> > I'll see if I can find some time to run that test on my rarely used
> windows box.
> > 
> > BR
> > Andreas Lehmkühler
> > 
> 
> 

--- original Nachricht Ende 

Re: Re: Reopen PDFBOX-483?

[Andreas Lehmkühler]

Hi,

Betreff: Re: Reopen PDFBOX-483?
Gesendet: Di, 09. Mrz 2010
Von: Maruan Sahyoun

> Hi ,
> 
> please find enclosed the text extracted from the printed PDF. Extraction was
> done using Adobe Acrobat 8.
> 
> X0X0X0 X0X0X05
> X0X0X0 X0X0X05
> X0X0X0 X0X0X05 
> X0X0X05 MM/DD/ X0X2 
> X0X2 
> X0X0X0X X0X0X0X
> X0X0X05 X0X0X05
> X0X0X05 X0X0X05
> X0X0X0X X0X0X05 
> X0X0X05 
> 
> X0X0 
> X0X0 
> X05 X0X0X05 
> MM/DD/ X0X0X05 X0X0 
> 
Hmm, that's odd. I'll run my own tests later when I'm at home. Finally that 
seems to be a windows only issue. I'll also file an issue on JIRA

Thanks for the tests!

BR
Andreas Lehmkühler

> 
> Maruan Sahyoun
> 
> 
> 
> Geschäftsführer: Maruan Sahyoun
> Handelsregister: AG Düsseldorf, HRB 53837
> UST.-ID: DE248275827
> 
> Am 09.03.2010 um 13:45 schrieb Maruan Sahyoun:
> 
> > Hi Andreas,
> > 
> > yes, the results are similar BUT most of the text and some of the lines
> are missing. Converting to Image output using PDFToImage provides a
> different and much better result where all text and lines are included and
> only some misplacement occurs. Is there a way to submit the attachment so
> you can see for yourself?
> > 
> > Maruan Sahyoun
> > 
> > Am 09.03.2010 um 13:38 schrieb Andreas Lehmkühler:
> > 
> >> Hi,
> >> 
> >> Betreff: Re: Reopen PDFBOX-483?
> >> Gesendet: Di, 09. Mrz 2010
> >> Von: Maruan Sahyoun
> >> 
> >>> Hi,
> >>> 
> >>> please find enclosed the result of the printing test conducted on 
> >>> 
> >>> Windows 2003 Server SP2 32 bit, Java 1.5 using a fresh built from trunk.
> The
> >>> test was done using the Adobe PDF printer driver as well as Apple and
> HP
> >>> Postscript printers with similar results.
> >> Thanks for testing. Your attachments didn't make it due to some
> restrictions of the mailing list.
> >> Probably it would be sufficient to describe the results. Let me guess,
> they are all similar. All
> >> contain text, some characters are misplaced and a wrong font is used.
> >> 
> >> BR
> >> Andreas Lehmkühler
> >> 
> > 
> 
> 

--- original Nachricht Ende 

Re: Re: Reopen PDFBOX-483?

[Andreas Lehmkühler]

Hi,

Betreff: Re: Reopen PDFBOX-483?
Gesendet: Di, 09. Mrz 2010
Von: Maruan Sahyoun

> Hi,
> 
> please find enclosed the result of the printing test conducted on 
> 
> Windows 2003 Server SP2 32 bit, Java 1.5 using a fresh built from trunk. The
> test was done using the Adobe PDF printer driver as well as Apple and HP
> Postscript printers with similar results.
Thanks for testing. Your attachments didn't make it due to some restrictions of 
the mailing list.
Probably it would be sufficient to describe the results. Let me guess, they are 
all similar. All
contain text, some characters are misplaced and a wrong font is used.

BR
Andreas Lehmkühler