fenics team mailing list archive

[Anders Logg <logg@xxxxxxxxx>]

There will be one once we have documented the demos. The idea is to
throw all our undocumented (but well commented) demos in one big
directory and put them back in the right category once they are
documented.

Here's what we have now:

http://www.fenics.org/newdoc/demos/index-python.html

This is the level of commenting we want for all our demos.

--
Anders


On Tue, Aug 31, 2010 at 02:35:18PM -0700, Ridgway Scott wrote:
> What about an ODE directory?
>
> On Aug 31, 2010, at 2:26 PM, Anders Logg wrote:
>
> >On Tue, Aug 31, 2010 at 11:22:10PM +0200, Kristian Ølgaard wrote:
> >>On 31 August 2010 22:10, Anders Logg <logg@xxxxxxxxx> wrote:
> >>>Garth and I have been discussing the location of the demos and it
> >>>seems clear that they should be part of DOLFIN (not fenics-doc).
> >>>
> >>>In the same way as for the C++ documentation, it is not practical to
> >>>split the demos between two different repositories (doc in one, and
> >>>code in the other). So we put them in the same repository.
> >>>
> >>>Then the question is where. For the C++ documentation, it is obvious
> >>>that the code+documentation should be part of DOLFIN (since the code
> >>>needs to be part of DOLFIN).
> >>>
> >>>For the same reason, the demos need to be part of DOLFIN since
> >>>we need
> >>>to be able to run and update them frequently while developing.
> >>>
> >>>It's also good to have the same kind of dependency both for C++ code
> >>>and demos, with changes propagating from DOLFIN to fenics-doc,
> >>>not the
> >>>opposite (or both).
> >>>
> >>>I therefore suggest that we put all the subdirectories inside
> >>>source/demos into DOLFIN_DIR/demo/.
> >>>
> >>>This directory would then contain three subdirectories:
> >>>
> >>> pde
> >>> la
> >>> undocumented
> >>>
> >>>The last directory would contain all demos (uncategorized)
> >>>which have
> >>>not yet been documented. A simple script can be run from
> >>>fenics-doc to
> >>>update.
> >>
> >>So we pull copies of source code files (*.py, *.ufl, *.cpp) from
> >>DOLFIN to fenics-doc, to be uploaded with the rest of the
> >>documentation on the web server?
> >>Sounds reasonable to me, and the test environment for DOLFIN
> >>will stay
> >>as it is now.
> >
> >Exactly. Let's not touch the files in dolfin-doc/source/demos for a
> >little while until we have sorted out the copying to DOLFIN and then
> >back again.
> >
> >We can setup a simple script (essentially calling rsync) to do the
> >copying. Perhaps the following scripts would be useful:
> >
> > utils/extract-dolfin-programmers-reference  # same as today but
> >renamed
> > utils/extract-dolfin-demos                  # new simple script
> >calling rsync
> > utils/extract-all                           # calling the above
> >two scripts
> >
>

--
Anders