1563678 - Blocking web fonts prevent Firefox front displaying PDF documents correctly

Status: NEW

(Firefox :: PDF Viewer, defect, P3)

Description

[Raphaël Jakse]

User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Firefox/60.0

Steps to reproduce:

• I blocked Web fonts in the Firefox settings.
• I loaded https://www.cdss.ca.gov/cdssweb/entres/forms/English/LIC500.pdf

Actual results:

PDF do not display correctly, like this one: https://www.cdss.ca.gov/cdssweb/entres/forms/English/LIC500.pdf

Expected results:

Fonts should still be loaded from PDFs. Blocking them in PDF seems useless and does not help accessibility / usability either.

(I block web fonts to save bandwidth, fonts included in PDFs are downloaded anyway)

Comment 1

[Alin Ilea, Desktop QA]

Mozilla/5.0 (X11; Ubuntu; Linux x86_64; rv:67.0) Gecko/20100101 Firefox/67.0

Hi,

I have managed to reproduce this issue on latest FF release (67.0.4) and latest Nightly build 70.0a1 (2019-07-08) using Ubuntu 18.04 and Windows 10.
I will move this over to a component so developers can take a look over it. If is not the correct component please feel free to change it to an appropriate one.

Thanks for the report.

Comment 2

[BugBot [:suhaib / :marco/ :calixte]]

Change the status for beta to have the same as nightly and release.
For more information, please visit auto_nag documentation.

Comment 3

[Brendan Dahl [:bdahl]]

When web fonts are disabled we use a different font renderer in pdf.js. Seems like there's a problem with it for the above PDFs.

Comment 5

[Vincent Lefevre]

With FF 99.0, most of the text on https://medias.unifrance.org/medias/127/140/35967/presse/the-duchess-dossier-de-presse-francais.pdf appears as blank when font download is blocked. The text can be copied, though.

Comment 6

[Vincent Lefevre]

This issue is still present in FF 107.0. Another example: https://docs.oracle.com/cd/E19957-01/800-7895/800-7895.pdf where the text is invisible, except for tt text and math symbols.

Comment 8

[Marco Castelluccio [:marco]]

I think we should follow what Jonathan suggested in bug 1796997 comment 11.