Re: Wheezy printing problem: 2 identical (?) machines and 1 does not print PDFs
On Sat, 26 May 2012 13:01:16 +0100, Brian wrote:
> On Sat 26 May 2012 at 09:58:31 +0000, Camaleón wrote:
>
>> On Fri, 25 May 2012 18:51:16 +0100, Brian wrote:
>> > Not quite. The PPD file/driver combination has to produce a file
>> > which, when sent to the printer, is understandable by the printer.
>> > The Gutenprint driver. for example, will always send PCL to an HP2200
>> > because, even though the printer can understand PostScript, the
>> > driver is designed to output only in the native language of the
>> > printer.
>>
>> But that only happens when the printer does not natively support PS
>> which does not seem to be the case; according to the printer specs¹ it
>> features HP PCL 6, HP PCL 5e and HP Postscript Level 2 emulation.
>
> No, it happens because there is a line like this
>
> *cupsFilter: "application/vnd.cups-raster 100
> rastertogutenprint.5.0"
>
> in the PPD file.
There's no "cupsFilter:" line in any of PPD files so maybe the printer is
not a true PostScript printer or uses a slightly different implementation
developed by HP.
>> > CUPS in Debian turns any input file into PDF, apart from a PDF file,
>> > of course. A PostScript file is filtered first by pstopdf, then by
>> > pdftopdf. A PDF file would only be filtered by pdftopdf. If the PPD
>> > file tells CUPS the printer wants PostScript it uses pdftops.
>> > Otherwise, there are other filters for converting the PDF coming from
>> > pdftopdf into the printer specific language.
>>
>> The problem still remains: the PDF file is not able to be printed
>> properly, it can be either because of a corrupted file (badly-crafted),
>> a problem with the driver (that is not able to interpret the code of
>> the file) an error in CUPS, a problem within the application used to
>> display the file... to solve the problem we have to start discarding
>> all these things.
>
> The OP has a PCL error on one machine. It is legitimate to use a PPD
> file producing PostScript to see if it resolves the problem. Converting
> the input file to PostScript before sending it to CUPS is something
> entirely different and of doubtful merit.
Converting the file to PS with a success result when printing could be of
interest for people in the know, that is, developers. It also can be a
workaround to get the job done until a proper solution applies. I don't
care about "merits" (dude, this is not a race!) but getting a problem
resolved with a solution that can fit the user needings.
>> > I wasn't concerned with speed as such but with your giving this as a
>> > reason for saying '. . . no conversion is needed between the doc and
>> > the printer' for a PS file.
>>
>> When you are using a PPD which defines the capabilities of the printer
>> and how it has to behave and both (printer harwdare and printer driver)
>> "speak" a standard language (like PostScript) you're reducing the
>> chances for an error than when using any other emulated language such
>> as PCL, simply because PS is not device-dependant, but PCL is, meaning
>> -for the latter- that you completely rely on the printer's capablities
>> and how your device interprets the job you are sending to it.
>
> The standard language used for printing in Debian has been PDF for over
> three years. There is no inherent benefit in inputting a PostScript file
> to CUPS.
I assure there is.
Again, I'm tired of seeing these kind problems with PDF, TIFF and PS
files (here at the office we do fax a lot and it's very common to see a
user with problems when printing hylafax produced files). We also have a
set of different printers (HP Laserjet, Ricoh and Oki) so we're very
accustomed to these problems, even from windows machines...
> You are also confusing the input file with the file sent by a CUPS
> backend to the printer. In the latter case, you reply completely on the
> capabilities of the printer's interpreters whether the file contains
> PostScript, PCL or any other language.
If the printer does support PostScript and uses a PS driver, the above is
completely irrelevant.
Greetings,
--
Camaleón
Reply to: