What do you think might be causing this problem?

Ramiro Espinoza
Ramiro Espinoza Posts: 839
edited October 2019 in Font Technology
Hi there,
I have a client using Quark Xpress blaming one of my fonts for spacing problems in the last paragraphs lines of a document. In my opinion this is probably just a H&J issue and not related to the fonts. However, as he claims to be an experienced Quark user I would like to know what you guys think.
Any idea about what might be causing it?
TIA




Comments

  • I am with you that it might be H&J, but there’s a key potential bug related to that: in one of my own fonts, when I didn’t include either the non-breaking space as a glyph, or one of the hyphen types, several rendering engines choked (Safari, Chrome, but also InDesign). Similar problems also showed in other fonts produced with a similar workflow, so it’s worth looking into the glyphs I thought were synthesised in their absence (soft hyphen, non-breaking space, things like that).
  • Ramiro Espinoza
    Ramiro Espinoza Posts: 839
    edited October 2019
    Thanks Robin for your advice. I've always included all necessary hyphens plus the non-breaking space U+00A0<br>

  • A font is a passive component; Quark is at fault in this instance.
  • It looks like the advance width of each character has been doubled.
  • Has this been resolved?

    It's likely neither your font nor QXP explicitly.

    I've never seen the problem myself, and I use QXP versions 9 forward nearly every day. It may be an issue of forced justification + some heavy use of tracking and or the kerning tables. Or?

    If you have the user copy a paragraph or two thus affected into a new document, send it to you, I can take a look. All settings should travel with the new document. I would simply need the Q file and the single font for the affected text, which I would delete following reporting to you.

    Alternatively, you can have them contact me directly or Quark itself. I can send you my email address via a private message.

    Mike
  • Given that the character advance widths are doubled means this is probably not a tracking-related issue. Tracking adds the same amount between characters, not an amount proportional to the individual character widths.

    I'm not sure how this could happen, but it seems almost like Quark is interpreting the character width data as if the UPM were 2048 instead of 1000, but not the outlines. Or something like that.
  • I have tried to get a text block into the same state and cannot. I used 5 versions of Q and a dozen different fonts from various makers since I posted.
  • Could be a corrupted document, I suppose. Copying affected text blocks out of the affected document into a new document could/should then display properly.
  • Feedback from the client:
    I've just re-opened a chapter in this book (the long document I was referring to) and found two more lines that have gone wrong.
    There's an endnote reference number in both cases -- and now that I see that, I remember that in most of the cases I have encountered there was also a superior figure.
    I tried changing the font for a few lines, including the faulty one, and the problem went away at once. When I change back to your font, the lines display correctly. But when I re-open the file, the problems come back.
    The display corrects itself if I click on "Normal" in Quark's style menu, but returns on re-opening the file.


  • Thomas Phinney
    Thomas Phinney Posts: 2,892
    edited October 2019
    Feedback from the client:
    I tried changing the font for a few lines, including the faulty one, and the problem went away at once. When I change back to your font, the lines display correctly. But when I re-open the file, the problems come back.
    The display corrects itself if I click on "Normal" in Quark's style menu, but returns on re-opening the file.
    They have not actually proven that the problem doesn’t happen with another font, as switching fonts back and forth made the problem go away, even WITH your font.

    I suggest the user try saving the file (under a new/temporary name, of course) AFTER changing the font to something else. Then re-open that file. If the problem returns for the other font (as I suspect it might!) the same way it did for your font, then that absolves your font of any responsibility/involvement.
  • notdef
    notdef Posts: 168
    edited October 2019
    I made a note ages ago that Quark assumes the UPM is a multiple of 1000 when drawing the advance width. @Mark Simonson ran into this here: https://typedrawers.com/discussion/comment/1529/#Comment_1529