Cool, I saw this at some point but forgot. How is it generated? I can't
find a corresponding script in that repo at a glance.
That is manually generated. I just update it whenever I add an SRFI or change its state. Note that it has much less information than the file you outlined. On the other hand, it has one extra bit of information: cross-reference information, so that I can generate "see also" links from one SRFI to another, e.g. for when one SRFI is intended to obsolete another, or is an alternative proposal to another.
Arthur, as the editor what is your opinion on what S-expression files to
have? It just occurred to me that if the files can be auto-generated
from HTML, they could just as well live in a central metadata repo
instead of the SRFI source repos. Then the source repos could have HTML
only, as they now do.
If we have symbol indices, I would prefer one metadata file per SRFI. If
not, we could have the data for all SRFIs in the same S-expression file.
But I'd really like the symbols :) And the abstract could be in the same
file.
I'm flexible. I just want to maintain editorial control over what appears on the SRFI web site, so I would want copies of the data to appear in the SRFI Git repos so that I could review changes before they were pushed to the site. But I defer, in general, to the volunteers who will generate the more elaborate files on whether there's a single file or many, and on whether they reside in a single repository (e.g.
srfi-common) or many (e.g.
SRFI-1).
My current file is only a single file because it contains so little data. That makes it easier for me to update it manually.