Hi Samuel
Samuel Thibault schrieb am 22.05.2020, 15:49 +0200:
>> I had the impression that it simply ignored settings that I have set in the
>> printer tab of my text editor (pluma), but I need a more reliable way to test
>> this. When I print from pluma, it always inserts the path to the file at the
>> top of the printed result, any clue why?
>
>Printing from an application is not recommended, because that'll
>basically always produce a pdf before giving it to cups. cups-filter
>will then have to extract the text again, thus getting top/bottom notes
>along the way, and have only a vague notion of what was meant to be a
>paragraph.
I see.
>> After a while, printing stopped working. I investigated the print command that
>> cups showed:
>>
>> ile2brl -p -Chyphenate=yes -CliteraryTextTable=en-us-brf.dis,en-GB-g2.ctb,de-g2.ctb,de-de-comp8.ctb,braille-patterns.cti -CinputTextEncoding=UTF8 -CbraillePages=yes -CbraillePageNumberAt=bottom -CpageNumberBottomSeparateLine=yes -CprintPages=no -CpageSeparator=no -CpageSeparatorNumber=yes -CcontinuePages=no -CcellsPerLine=27 -ClinesPerPage=26 | | addmargins
>>
>> First I figured out that liblouisxml-data was missing, should this be added as
>> a dependency?
>
>liblouisxml-data would only be needed when using the corresponding
>filters in liblouisxml-bin, which I don't think it is doing here?
Ah, this is a mistake of mine, the output doesn't change. Do you have a clue
about the last lines:
:2: invalid literaryTextTable
Finish read_configuration_file
Starting new document
Cannot open main table (null)
>> Second, there seems to be no addmargins in the Debian archive, is this an
>> oversight?
>
>See the cups-braille.sh helper script, sourced from texttobrf.
Ah, thanks.
>> Last but not least, there's a double pipe, I suppose dash doesn't like this.
>
>That would happen if TRANSLATE (between the two '|') is empty, but the
>texttobrf.in script should be already just dropping that part in that
>case, if it's not the case more detailed investigation is needed.
I don't know, this is what cups displayed and I thought I might run the shown
command as-is so that I can investigate the failure.
>> > I'd need to try it out myself. I'd think that curves are much more precise
>> >> with ~50 DPI,
>> >
>> >Yes, but I'm unsure one can feel the difference that much.
>>
>> I have tried to plot a diagram that has two close lines and it shows as one
>> line in the printed version.
>
>Do you also mean that with the lower DPI you are not feeling the same?
I don't have the comparison, sorry. I could test this at a later time.
What I found is that most images are scaled down to A4 + a bit. That means
that almost 90 % of the image is on the first page, a small remainder flows to
the second. Is this maybe a Index v3-v5 change?
Thanks
Sebastian
Attachment:
signature.asc
Description: PGP signature