>> 3. Is it worth the effort to automate the process of updating this
>> site? (via CI, github actions.)
> I am not sure what trouble Ralf had.
I just updated
fricas.github.io. Well, my "trouble" is in an old setup
that still lies around and always confuses me. Releases are rare enough
to eventually get it right, but I wanted to check before I update.
There is, for example,
https://github.com/fricas/fricas/blob/master/src/doc/sphinx/source/citation.rst
it shows
@Misc{FriCAS,
key = {FriCAS},
author = {{FriCAS team}},
year = {2024},
title = {{FriCAS} 1.3.10---an advanced computer algebra system},
note = {Available at \url{
http://fricas.github.io}}
}
with the wrong number. I would be happy if Waldek puts it on his list
"to do before releas" to update the version also in that bib-entry.
I think that is the only piece left, otherwise
cd builddir/src/doc; make doc
build already everything to be put into the fricas.github.io.git repository.
>> 4. Is it worth the effort to also have the documentation site for
>> the nightly code (git HEAD)?
> I do not think so.
I also don't think so. Using
cd builddir/src/doc; make localdoc
Anyone can get the documentation that even needs no internet connection
to browse through the api description. The book is anyway locally built.
I builld specifically, for
fricas.org, I haven't investigated, but I
think it should be possible to do via "make doc" and setting a few
PACKAGE_...
variables before calling make. See here...
https://github.com/fricas/fricas/blob/master/src/doc/Makefile.in
I do not care much about the size of the fricas.github.io.git
repository. I try to keep it small, but I actually want to also record
the exact state as it was when I uploaded it (for historical reasons).
And 100MB is not something that I find too trouble making when it
contains already 14 releases.
> If we host the reference book in one location, then we can get a more
> accurate statistics on how many people have downloaded the book.
Well, I would also like to know how many people actually use FriCAS (it
would be a kind of motivation---maybe also for other developers), but I
think that the download statistics is only a very rough measure to get
such a number. Maybe we can encourage our users to send feedback to us
even if there are no problems to solve.
Ralf