[prev in list] [next in list] [prev in thread] [next in thread] 

List:       koffice-devel
Subject:    Re: CC: Re: "new" wp for linux
From:       Tomasz Grobelny <grotk () poczta ! onet ! pl>
Date:       2004-04-14 0:27:16
Message-ID: 200404140227.16094.grotk () poczta ! onet ! pl
[Download RAW message or body]

On Wednesday 14 of April 2004 01:45, James Richard Tyrer wrote:
> David Faure wrote:
> > On Wednesday 14 April 2004 00:09, Tomasz Grobelny wrote:
> >>On Tuesday 13 of April 2004 23:34, James Richard Tyrer wrote:
> >>>There is an additional issue with Printer rendering that KWord and/or Qt
> >>>must know the resolution of the printer.
> >>>My best guess is that KWord-1.3 is currently using Printer rendering so
> >>>there are still slight problems.
> >>
> >>I doubt it since Qt prints through postscript files and that means it has
> >> no knowledge about destination printer.
> >
> > There are case where you can find the resolution of the destination
> > printer (via QPrinter or KPrinter), but that doesn't always help anyway
> > (worst example is printing to a .ps file and then printing to any printer
> > you want).
>
> That is the way that Windows works.
AFAIK Windows API in this matter is as bad as Qt or even worse so each program 
implements high quality rendering on its own.

> What I would really like to know is how GhostScript
> renders fonts.  Does it hint them to the printer DPI or does it use the
> unhinted glyph size?
>
I think we should assume it makes the best thing it can for specified device. 
But there is certainly no device interdependency as in Qt.

> >>>That is, for KWord to work correctly, Qt needs to add what I have called
> >>>PostScript font rendering.
> >>And that means redesigning entire set of painting/font classes :-(. Why?
> >>Because:
> >>-Qt sets fixed width of strings in postscript files,
> > What else would it do?
> Other word processors do this differently.
> WordPerfect has this:
>
> /Times-RomanR 600 ff
> 17 0 32 (The quick brown fox jumps over the fat lazy dog.)W
>
> I don't know exactly what this does.  Some other fonts use: "S" rather
> than: "W" and don't have the three numbers at the start.
>
W and S are propably procedures defined in document prolog.

> So, my best guess is that WordPerfect simply lets GhostScript render the
> fonts based on the unhinted glyph metrics in the font file.
>
> OTOH, OpenOffice does this:
>
> (Times-Roman-iso1252) cvn findfont 50 -50 matrix scale makefont setfont
> <54686520717569636B2062726F776E20666F78206A756D7073206F76657220746865206661
>7420 6C617A7920646F672E>
> [31 25 22 12 25 25 14 22 25 13 25 16 25 36 25 13 16 25 25 13 14 25 39 25 19
> 13 25
>   25 22 16 13 14 25 22 12 17 22 14 12 14 22 22 25 13 25 25 25 0]
> xshow
>
> Which appears to be data to fudge the spacing, but no line length.
>
Looks weird as typographical point is quite a big unit for glyph positioning 
(but yes, that is what xshow does)...

Tomek
_______________________________________________
koffice-devel mailing list
koffice-devel@mail.kde.org
https://mail.kde.org/mailman/listinfo/koffice-devel
[prev in list] [next in list] [prev in thread] [next in thread]