Re: future of RTFM *.answers archives

---------

Jari Aalto+mail.emacs (jari.aalto@poboxes.com)
11 Mar 1999 11:46:12 +0200


[Sorry about broken Message-Id (no thread), but somehow I no longer
get faq-l postings, so I had to dig these from
http://lists.consensus.com/scripts/lyris.pl?visit=faq-maintainers]

>>clewis@ferret.ocunix.on.ca

IIRC www.landfield.com generates a fully HTMLized archive of FAQ
postings as gleaned from Usenet. Along with RFCs and other things. I'm
not sure of his selection heuristics. I suspect he looks for the
appropriate Approved line.

His HTML-izer is pretty good if you stick to digest format for FAQs.

Perhaps instead of the FAQ group maintaining HTML versions, just
acknowledge/adopt www.landfield.com, and provide a set of recommendations
to FAQ writers to ensure that his HTML-izer will work on their FAQs. I know
I abandoned thoughts of HTMLizing my FAQs because of that site.

About HTML conversions

The email2html converters are not the best tools to generate good HTML version
out of the document. They are adequate for people, that do not intend to
write HTML or .ps or other formats as well, but who keep simple text files.

Some of us actually use SGML, DocBook DTD, texinfo, etc which is base format
from which the .man .txt .html and .info versions of tyhe documents are generated.

I use my own plain text (called Techinical Format
ftp://cs.uta.fi/pub/ssjaaa/t2html.html), which gives me WysiWyg and
properties:

o Text files can be kept up to date with any editor
o The text can be mailed/grepped/indexed "as is"
o CVS and RCS diffs are best with plain text
o A bonus: An Emacs minor mode takes care of the editing, renumbering,
paragraph fixing, colouring etc.

So, each author us base text as they see fit and use a tool to convert
the text into HTML or other formats. A simple email2html conversion would
only approximates the results.

Author directories

As RTFM simply slurps the posted messages from the newsfeed and archives
the messages according to Archive-name: we would need to develop whole new
concept for a archive that can be accessed and updated directly.

Is anyone of you a CPAN author (Perl achive)? They have a nice Author https
account where files can be uploaded and the system takes care of distribution
to the correct directories.

For rtfm: if we had an FTP directory where the files could be
uploaded and a script would then place those files to correct directories.

Since documents can be in variety of formats, I suggest that we use a
control file, that guides placement of the uploaded files. Here is
initial design how I would see the update to happen:

1. log in the rtfm with my Author ID and PASSWORD
2. Upload the _doc_ files to my rtfm upload directory
3. Upload a _CONTROL_ file that tells where to place those files

Let's suppose I upload

pm-tips-frame.html # Framed version of the Doc
pm-tips-body.html
pm-tips-toc.html

pm-tips.ps
pm-tips.txt

The I upload the control (or config) file

# upload.conf
pm-tips-frame.html: http://www.faq.org/dir/structure/where/to/put/
pm-tips-body.html: http://www.faq.org/dir/structure/where/to/put/
pm-tips-toc.html: http://www.faq.org/dir/structure/where/to/put/
pm-tips.ps: ftp://rtfm.mit.edu/dir/structure/where/to/put
pm-tips.txt: ftp://rtfm.mit.edu/dir/structure/where/to/put

The script would check the uploaded files and move them to correct
locations acording to control file. Alternatively, this control file
could be assigned for each author by the RTFM staff, so that
whenever we ask "new" submission, they add entry to this control file.

jari



[ Usenet Hypertext FAQ Archive | Search Mail Archive | Authors | Usenet ]
[ 1993 | 1994 | 1995 | 1996 | 1997 ]

---------

faq-admin@faqs.org

© Copyright The Internet FAQ Consortium, 1997
All rights reserved