|
From: Robert P R. <ri...@cs...> - 2005-12-14 22:11:31
|
Okay, here are a few patches. I've broken them up into small pieces
so that it's easy for you to just take some of them if you want.
pointless-svg.patch - Adds a new command, =svg, for including SVG
graphics (note - I've patched Makefile.am and configure.in, so it
requires re-running automake and autoconf.)
pointless-coloreps.patch - Adds a new commands, =coloreps, for including
eps graphics in color (w/ transparent background) instead of
grayscale
pointless-cache.patch - Makes the svg and eps modules perform 'real'
caching: if a cached version of the .png exists, they won't
regenerate it unless it's older than the .eps or .png. I had some
figures that took a long time to render on my laptop, so this sped
things up considerably. In order for this to work, I had to change
the name of the cache directory to include the resolution. The SVG
patch must be applied before this one
pointless-plp.patch - If the input filename ends in '.plp', runs it
through the C pre-processor before opening it. I did this for a few
reasons:
* It allows me to use familiar #include and #define syntax
* The =include command was doing strange things for me, like
putting the included file into a strange order. I know, I
should have just fixed it, but I wanted the full 'power', such
as it is, of cpp
* It lets me do 'header guard' type stuff for files included by
more than one file. So, when I broke my presentation up into
small files, I could have them all include 'defs.plp' so that
I could use any of them standalone while I was working of
them.
* It theoretically allows the user to pass -DFOO type arguments
to cpp, so that you could have the presentation behave
differently depending on what arguments you passed to it. I
don't have support for these yet, though
Note - the processed file goes into $TMPDIR or /tmp - I wanted to
put it into the cache dir, but the pll object doesn't yet exist at
the time the .pll/.plp file is opened.
pointless-frame_duration.patch - There's an assert in the code that the
duration of a frame is > 0. I found that if I left pointless sitting
at one slide for too long, this would fire when I tried to leave the
slide. Not too pretty in the middle of a talk. :) So I just changed
the code to make the duration tiny but positive if this happens.
pointless-item_nostep.patch - I don't like the default behavior of
having every item start a new step. I figured you guys might not
take this patch, but some people might find it useful.
A tarball of the patches is available here:
http://www.flux.utah.edu/users/ricci/software/pointless-patches.tar.gz
Here are some slides that I made with pointless and these patches:
HTML/PNG:
http://www.flux.utah.edu/users/ricci/slides/assign-informs/
Source:
http://www.flux.utah.edu/users/ricci/slides/assign-informs.tar.gz
--
/-----------------------------------------------------------
| Robert P Ricci <ri...@cs...> | <ri...@fl...>
| Research Associate, University of Utah Flux Group
| www.flux.utah.edu | www.emulab.net
\-----------------------------------------------------------
|