There was a stray space before the "Returns:" heading so it wasn't
processed as intended. Also, there was an empty line above it leading
to an empty paragraph in the DocBook funcref.xml.
section. Both comedi_to_phys and comedi_to_physical complement
each other. Comedi_to_phys is linear and very fast whereas
comedi_to_physical is more precise but requires polynomials
and calibration data.
Change DocBook XSL parameters for HTML output to use numbered sections
(like it was before we switched from SGML to XML) and to use just the
numbers in cross references, rather than number and/or title.
Add extra detail to the section on acquisition functions about the
range and aref parameters, and describe the comedi_range type.
The whole section is still a bit higgledy-piggledy and could do with
some rearrangement one day!
This makes sure the HTML output for
<function>foo<parameter>bar</parameter><parameter>baz</parameter></function>
includes parentheses, although it doesn't seem to add empty parentheses
after a plain <function>foo</function> which perhaps it should.
Some DocBook mark-up changes. Also avoided absolute path-names for
sources as people tend to put them in different places! Also changed
instructions for submitting drivers for review.
Mention that the stuff that relies on the COMEDI_TRIG ioctl, including
the slowly varying input (comedi_sv_...) functions no longer work with
current versions of the Comedi kernel layer.
Removed subdevice_struct as it is internal to comedilib. Removed
struct comedi_t_struct, leaving only the comedi_t typedef as
applications only see it as a pointer to an opaque, incomplete type.
Add comedilib_dblatex_config.xsl stylesheet and pass it to dblatex when
producing PDF output. The current version of the stylesheet doesn't
make much difference to the output, but it's somewhere we can put stuff
later.
We probably want to do something similar for the other PDF backends, but
those are currently broken anyway.
info in the demo directory. I've made it a bit more verbose
and also took into account that now the generic_timed command
takes more than one channel so that a user can experiment which
sampling rates are possible with different channel numbers.
I've also added an option to set the sampling rate for the generic_timed
command. That sould actually not be needed but I'm sure there
are drivers out there which do not correct the sampling rate
downwards automatically.
Changed handling of --with-pdf-backend[=yes] to avoid picking the default
xmlto backend for PDF generation because it's broken.
The default xmlto backend for PDF generation can still be selected using
--with-pdf-backend=default if you really want to try it.
