[cups] Printing to native XPS printers using CUPS

[Gernot Hassenpflug]

On Sat, Sep 9, 2017 at 4:47 AM, Helge Blischke <helgeblischke at web.de> wrote:
>
>> Am 08.09.2017 um 06:15 schrieb Gernot Hassenpflug <aikishugyo at gmail.com>:
>>
>> On Wed, Nov 4, 2015 at 6:50 PM, Helge Blischke <helgeblischke at web.de> wrote:
/../
>>> The gstoxps filter uses Ghostscript’s xpswrite device (notwithstanding the deficiencies of this device)
>>> and consumes both PostScript and PDF (implicitly using Ghostscript’s PDF interpreter).
>>
>> Hello all,
>> It has been a while, but I see that in the latest Ghostscript 9.21
>> released in March 2017, the following change is noted:
>>
>> "The GhostXPS interpreter now provides the pdfwrite device with the
>> data it requires to emit a ToUnicode CMap: thus allowing fully
>> searchable PDFs to be created from XPS input (in the vast majority of
>> cases)."
>>
>> References:
>> 9.21 release notes: https://www.ghostscript.com/doc/9.21/News.htm
>> 9.18 notes on XPS: https://www.ghostscript.com/doc/9.21/History9.htm#Version9.18
>>
>> I assume from the above that in general XPS input and output have been
>> much improved since 2015, and surmise that it might be possible to use
>> the filters to obtain a "better" output than before.
>> Since my system (Debian sid) has ghostscript 9.21 I will probably be
>> able to do a test on the weekend.
/../
>> I noticed there has been no update to Helge's experimental filter
>> since 2012, perhaps if the gsxps source has changed, more parameters
>> might be settable these days.
>> I will check if gsxps is available on my Debian machine at home (not
>> available on work Ubuntu machine with GS 9.18 installed) and see if
>> help shows any more parameters usable.
>> If anyone has information on this I'd be happy to hear about it.
>> (The problem, IIRC, was that the output files are bitmaps and thus huge.)

Hello Helge,

> As for the „experimental“ filter, I’m rewriting this thing to a more general
> gstofoo filter with foo specified by the name the filter is called as.
> Currently, I test using the Ghostscript devices
> pdfwrite, ps2write, xpswrite, pxlmono or pxlcolor,and txtwrite.
> As soon as I succeed with testing all this stuff, I’ll publish it at sourceforge

Thanks for the feedback, that sounds like an interesting work in progress.

> BTW, as far as I can tell up to now, the xpswrite device hasn’t changed very much since.

Ah, to bad. So changing XPS to PDF has improved a lot (ghostXPS
interpreter) but not the creation of XPS.
Regards,
Gernot