Re: broken lines with Ghostscript (pdf)

In article <1bedcf4e-086e-41ba-8743-7fe380522825
@e23g2000prf.googlegroups.com>, estufa8@xxxxxxxxx says...

My questions are:
Why 51 characters?

Could be coincidence, could be a consequence of the (Word) style, could
be a result of the page size and the line length. Hard to say really.

Is there a string length limitation using Ghostscript?

Not that you are likely to run across, and certainly not that small.


Or may be it is rounding up little rests of all glyph positioning
(with TJ)?

According to the source, its to work around a bug in at least some
versions of Acrobat.

From the pdfwrite source code:

/*
* Acrobat Reader limits the magnitude of user-space
* coordinates. Also, AR apparently doesn't handle large
* positive movement values (negative X displacements), even
* though the PDF Reference says this bug was fixed in AR3.
*
* Old revisions used the upper threshold 1000 for tdw,
* but it appears too big when a font sets a too big
* character width in setcachedevice. Particularly this
happens
* with a Type 3 font generated by Aldus Freehand 4.0
* to represent a texture - see bug #687051.
* The problem is that when the Widths is multiplied
* to the font size, the viewer represents the result
* with insufficient fraction bits to represent the precise
width.
* We work around that problem here restricting tdw
* with a smaller threshold 990. Our intention is to
* disable Tj when the real glyph width appears smaller
* than 1% of the width specified in setcachedevice.
* A Td instruction will be generated instead.
* Note that the value 990 is arbitrary and may need a
* further adjustment.
*/

Can I avoid this from the command line when I convert it?

No, it seems not, this is embedded into the source. Since it generates
valid PDF files, its unlikely to change.


Ken
.

Relevant Pages

  • Re: "change case" not working for Unicode characters
    ... I think that's not a bug (well, it may be a bug, but in the font, not ... I believe that if the font you are using does not contain upper-case ... characters for those characters, or if the font is not correctly encoded to ...
    (microsoft.public.word.docmanagement)
  • Re: Strange oe character on plain console
    ... >> probably a bug, since the font contains these characters. ... Which package is this? ...
    (Debian-User)
  • Re: Bug in Graphics.DrawString ?
    ... characters is more than that before 20th character. ... font "Microsoft Sans Serif" size 8.25 ... Although a bug is possible, I wouldn't expect one in this case. ...
    (microsoft.public.dotnet.framework)
  • Re: Looking for tk internals information
    ... >>User A runs the app, and sees the document just fine. ... > characters in question, any font that has the characters should do. ... B has the "bug" that they don't have a proper font installed. ...
    (comp.lang.tcl)
  • Bash-3.0 available for FTP
    ... This is the third major release of bash. ... Bash-3.0 provides a number of bug fixes and cleanups to the features ... pathname longer than PATH_MAX characters. ... with multibyte characters better. ...
    (comp.os.linux.misc)