Stop using corrupted, incorrect embedded fonts in PDF reports.

[admaecc211151]

The word "Open", "Side", "；"( ; in Full-width) have wrong shapes.

It's been more than three months and haven't fix yet.

These are three question.

1. Why using Noto Sans CJK JP for Traditional Chinese, Not TC?

2. There already exist Noto Sans CJK JP in Windows Font, why not use same one but create a new one?

3. Why using embed font?

I don't care about using what fonts, Proportional MingLiU is fine, too.

But being readable is the minimum requirement.
It's very hard (and waste time) to edit the text fonts again in PDF.

[ryan_warsaw]

Hello @admaecc211151, I recognize the example provided and was the engineer that worked on your support case. Thank you for bringing this to our attention, and sorry that you had a poor experience with our Chinese language support. Let me address your questions: 

Q: Why using Noto Sans CJK JP for Traditional Chinese, Not TC? 

A: This was a bug. For certain Kanji characters, they can share the same identifier (unicode code point) but visually look different depending on the language. For the browser, this isn't much of an issue because it can see what your system's language preference is and render the correct version. In our case, we weren't providing the locale you specified in your report settings when rendering the PDF on the server to help the system figure out the correct font to use so it would try to "guess" and often got it wrong for Chinese / Korean / Japanese fonts. 

Q: There already exist Noto Sans CJK JP in Windows Font, why not use the same one but create a new one?

A: We don't modify these fonts in anyway. The issue of certain characters appearing to be "corrupt" was an issue in an older version of these fonts.  I was able to reproduce this in Noto Sans v2.002 and it was fixed in Noto Sans v2.004. We worked with our vendor who provides the operating system to get the newer fonts. This is why it took a while to get resolved. We are currently using the latest version.

Q: Why using embed font? 

A: All reports generated in ACC / BIM360 are PDF/A compliant. PDF/A is a variant of a PDF that is optimized for long term preservation. The PDF/A standard dictates that all resources (including fonts) are embedded within the PDF itself. This is to ensure the document looks the same to all readers, regardless of the PDF viewing tool they are using. It also means that 5-10 years down the road the report will look the exact same as it does today, which is important to some of our customers. 

[admaecc211151]

Thank you for your reply, but the font is still incorrect for some characters, such as "開"(open) and "議"(issue).

[ryan_warsaw]

Hey @admaecc211151 just acknowledging that I saw this, and was able to reproduce the issue. The characters on themselves seem to render fine, but used in conjunction with English characters (for example an Issue title) they appear to not be rendering properly. Is this your experience also? I will take a closer look and get back to you when I have further information.

[admaecc211151]

Yes, some words only broked in title, and some only in body text. 
I suspect it’s caused by bold text using a different font.

[ryan_warsaw]

Hey @admaecc211151, thank you for your patience. It ended up being a bug with our font subsetting logic. A fix was released this Wednesday (Oct. 1st). Future exports should no longer experience this issue. I see you have an open support case with us, please let them know if you run into any more issues.