Content Management and Capture

Hi,

I was just working on a ticket with IBM and was wondering what the general solution to the problem is.

Source of the ticket is a .msg file that has several PDF attachments. Wen an ICN user did a 'Download as PDF' the resultant PDF look garbled next to illegible.

The cause was missing fonts for the OIT conversion library on the ICN server.

Now the first question is, why would the OIT library take a perfect PDF (an attachment) and try a PDF->PDF conversion that can only worsen things????

A quick solution was to install the liberation fonts. They have fonts with the same metrics as the most widely used Windows fonts (e.g. Arial) but differ in detail.

The resultant PDF (when downloaded as PDF) looks much better as the layout is at least correct and it is legible, but surely not 100% identical which raises all sorts of concerns on complicance....

So the second question: What is the architecturally correct solution to this dilemma? mscorefonts project is discontinued so it is not allowed to just copy the *.ttf to Linux. And even if it were, surely there comes along a PDF with some weird google font.

Thanks,

/Gerold

PS: interestingly enough, if you download a PDF as PDF it is untouched, which I would expect.

The case for your post, ending in 9546 has been closed. Please work through the support channel if you have follow up questions.

Thank you, Gerold.

Hi,

I was just working on a ticket with IBM and was wondering what the general solution to the problem is.

Source of the ticket is a .msg file that has several PDF attachments. Wen an ICN user did a 'Download as PDF' the resultant PDF look garbled next to illegible.

The cause was missing fonts for the OIT conversion library on the ICN server.

Now the first question is, why would the OIT library take a perfect PDF (an attachment) and try a PDF->PDF conversion that can only worsen things????

A quick solution was to install the liberation fonts. They have fonts with the same metrics as the most widely used Windows fonts (e.g. Arial) but differ in detail.

The resultant PDF (when downloaded as PDF) looks much better as the layout is at least correct and it is legible, but surely not 100% identical which raises all sorts of concerns on complicance....

So the second question: What is the architecturally correct solution to this dilemma? mscorefonts project is discontinued so it is not allowed to just copy the *.ttf to Linux. And even if it were, surely there comes along a PDF with some weird google font.

Thanks,

/Gerold

PS: interestingly enough, if you download a PDF as PDF it is untouched, which I would expect.

Hi Nanda,

I'm not sure what you are aiming at. Support is working with problems that violate the 'integrity' of the software.

As a consequence of the ticket resolution I'm asking a very general architectural question and would like to know how others handle this kind of problems.

I would be surprised to get broad architectural questions answered by support, this is more a question for the consulting business. But if you still insist, I will certainly be glad to open another ticket referencing your comment.

Kind regards,

Gerold

PS: Given the lack of answers it does not seem a very broad issue :-(

The case for your post, ending in 9546 has been closed. Please work through the support channel if you have follow up questions.

Thank you, Gerold.

Hi,

I was just working on a ticket with IBM and was wondering what the general solution to the problem is.

Source of the ticket is a .msg file that has several PDF attachments. Wen an ICN user did a 'Download as PDF' the resultant PDF look garbled next to illegible.

The cause was missing fonts for the OIT conversion library on the ICN server.

Now the first question is, why would the OIT library take a perfect PDF (an attachment) and try a PDF->PDF conversion that can only worsen things????

A quick solution was to install the liberation fonts. They have fonts with the same metrics as the most widely used Windows fonts (e.g. Arial) but differ in detail.

The resultant PDF (when downloaded as PDF) looks much better as the layout is at least correct and it is legible, but surely not 100% identical which raises all sorts of concerns on complicance....

So the second question: What is the architecturally correct solution to this dilemma? mscorefonts project is discontinued so it is not allowed to just copy the *.ttf to Linux. And even if it were, surely there comes along a PDF with some weird google font.

Thanks,

/Gerold

PS: interestingly enough, if you download a PDF as PDF it is untouched, which I would expect.