This change should produce only the .svg files: diff --git a/src/sage_docbuild/conf.py b/src/sage_docbuild/conf.py index 9b6b37480b..b5d6815c76 100644 --- a/src/sage_docbuild/conf.py +++ b/src/sage_docbuild/conf.py @@ -128,7 +128,7 @@ from sage.all_cmdline import * """ plot_html_show_formats = False -plot_formats = ['svg', 'pdf', 'png'] +plot_formats = ['svg'] # We do *not* fully initialize intersphinx since we call it by hand # in find_sage_dangling_links.
(I'm not suggesting this as a general change, just for your particular situation.) On Wednesday, June 28, 2023 at 12:07:46 PM UTC-7 Marc Culler wrote: > Note: I had already compressed the .svg files, hence the .gz extension. > > Before compression I get: > > % find local/share/doc/sage/html/en -name 'chart-12*' > local/share/doc/sage/html/en/reference/manifolds/_images/chart-12.svg > > local/share/doc/sage/html/en/reference/manifolds/sage/manifolds/chart-12.svg > > local/share/doc/sage/html/en/reference/manifolds/sage/manifolds/chart-12.png > > local/share/doc/sage/html/en/reference/manifolds/sage/manifolds/chart-12.pdf > > - Marc > > On Wednesday, June 28, 2023 at 2:02:47 PM UTC-5 Marc Culler wrote: > >> I think I may have found a resolution of this problem, that involves >> doing almost nothing. It appears that sage_docbuild oftern builds image in >> all three formats - svg, png and pdf. All three types get stored in the >> html directory and the svg is duplicated: >> >> find local/share/doc/sage/html/en -name 'chart-12*' >> local/share/doc/sage/html/en/reference/manifolds/_images/chart-12.svg.gz >> >> local/share/doc/sage/html/en/reference/manifolds/sage/manifolds/chart-12.svg.gz >> >> local/share/doc/sage/html/en/reference/manifolds/sage/manifolds/chart-12.png >> >> local/share/doc/sage/html/en/reference/manifolds/sage/manifolds/chart-12.pdf >> >> The actual html code seems to only uses the svg files stored in _images. >> Hopefully I can just remove the others without affecting the html >> documentation. >> >> grep -r chart-12 local/share/doc/sage/html/en/reference/manifolds >> local/share/doc/sage/html/en/reference/manifolds/sage/manifolds/chart.html:<img >> >> alt="../../_images/chart-12.svg" class="plot-directive" >> src="../../_images/chart-12.svg" /></figure> >> >> Since this seems to be a change between 10.1beta3 and 10.1beta4, I would >> have to guess that it might be considered a bug. So maybe I should move >> this back to sage-release? >> >> - Marc >> >> >> >> On Wednesday, June 28, 2023 at 1:13:20 PM UTC-5 Marc Culler wrote: >> >>> I am not proposing using gzipped svg for a normal sage build. I only >>> want to do it for the Sage_macOS app, which will have a simple >>> self-contained server running on the loopback interface that provides >>> gzipped content (with the Content-Encoding header set to "gzip"). Of >>> course I would be happy to work on incorporating the same mechanism in a >>> standard build of Sage, and that would reduce the footprint of the >>> documentation considerably. But in the short term, for the macOS binary, I >>> need to be able to find the magic switch which makes sage_docbuild use svg >>> instead of png. >>> >>> I don't know how much of a problem the image size will pose. There are >>> some svg images already in the documentation, which seem to work, but I >>> hadn't gotten to the stage of worrying about that yet. >>> >>> - Marc >>> On Wednesday, June 28, 2023 at 12:50:31 PM UTC-5 Michael Orlitzky wrote: >>> >>>> On Wed, 2023-06-28 at 10:07 -0700, Marc Culler wrote: >>>> > >>>> > BOTTOM LINE: we get more than a 90% reduction in size simply by >>>> choosing to >>>> > use the .svg extension when saving the plot instead of the .png >>>> extension. >>>> > >>>> >>>> SVG is the right choice for most graphics, but there are some practical >>>> problems: >>>> >>>> * The documentation needs to know how big to display an SVG graphic. >>>> With a PNG, the default is to use the image's height/width in >>>> pixels, but with SVG, there's no such obvious default. >>>> >>>> * gzipped SVG doesn't work over the file:/// protocol in my firefox. >>>> This hasn't been a big enough problem for me yet to diagnose it, >>>> so I can't say how serious a problem it is. (I'll play around later >>>> today.) >>>> >>>> * Browser support in firefox/chrome alternatives still isn't great, >>>> although I think webkit is getting a new SVG renderer "soon." This >>>> is actually relevant more today than it was ten years ago, because >>>> adding rust to firefox made it less portable, meaning you're >>>> more likely to be stuck with one of those alternatives. >>>> >>>> * Somebody's got to go through and look at 100MB of images to make >>>> sure they still look right if we change 'em. >>>> >>>> -- You received this message because you are subscribed to the Google Groups "sage-devel" group. To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/sage-devel/d838a05a-ab87-49ca-a09b-c64d5c86e037n%40googlegroups.com.
