Mailing List Archive

Re: [STATUS]
Ralf,

I suspect that a lot of the people on this list aren't familiar with
the way we run the developer list, and that the whole business of
subject lines and voting is rather mystical. I'm not at all sure
that the same conventions - voting, at least - are really
appropriate here. Let's not go too fast, eh?

If there *are* any here that fall into that category, may I suggest
a peek at <http://dev.apache.org/>, which includes some background
that may or may not be useful.

That said..

I don't much care what the base format of the documentation is as
long as

o it can be maintained with a text editor on a dumb terminal
o the (or a) back-end produces HTML with which I'm happy

I'd rather not install another ten tons of cruft.. er,
single-purpose tools on my system. I'd be particularly happy with
something Perl-based (which I think your recommended tools already
are, Ralf). Perl rocks.

The current pile of documentation - our basic foundation here - is
the pages at <http://www.apache.org/>, including those under
/library/ and /docs/. What you see there are files with SSI
instructions in them, so comparable boilerplatibility would be nice.
If you want to get your hands on the source HTML files, see the
above-named URL and read the part about working with CVS snapshots.
The manual pages themselves are in CVS module apache/htdocs, and the
Apache site files are in CVS module apache-site.

One nice thing about the current arrangement is that the HTML files
at the above location are *very* easy to maintain - they're just
checked out from a CVS library. Any replacement system can only be
more complicated, but it *must* allow easy regeneration of the site
documents in HTML.

Also, I'm already working on the FAQ and making it
news.answers-ready, so that's another data-point. However,
conversion to that format, if not automatic, should be a simple
matter of Perl. I'd appreciate it if anyone wanting to do a
wholesale reorg of the FAQ would coordinate with me first. (I'm
including you, Ralf! ;-)

#ken :-)}
Re: [STATUS] [ In reply to ]
On Thu, 7 Aug 1997, Rodent of Unusual Size wrote:

> I suspect that a lot of the people on this list aren't familiar with
> the way we run the developer list, and that the whole business of
> subject lines and voting is rather mystical. I'm not at all sure
> that the same conventions - voting, at least - are really
> appropriate here. Let's not go too fast, eh?

I gave a bunch of people a brief introduction to how apache coding works.
I can also give them a hand getting the tools going... but yeah you'll
have to start a bit slowly. I'm hoping that the people I invited will
give us some hints as to how they normally manage documents and such,
because I'm sure the coding approach to docs is not the best.

Dean
Re: [STATUS] [ In reply to ]
In article <97080717493881@decus.org> you wrote:

> I suspect that a lot of the people on this list aren't familiar with
> the way we run the developer list, and that the whole business of
> subject lines and voting is rather mystical. I'm not at all sure
> that the same conventions - voting, at least - are really
> appropriate here. Let's not go too fast, eh?

> If there *are* any here that fall into that category, may I suggest
> a peek at <http://dev.apache.org/>, which includes some background
> that may or may not be useful.

I don't want to run the same voting procedures, but STATUS messages are ok
because we need to know what the current state is. At least initially when we
have to make some important decisions. Here infos are needed on alternatives
to make good decisions. But the hint to dev.apache.org is good. Thanks Ken.
Please members on apache-docs read also the ideas expressed there.

> That said..

> I don't much care what the base format of the documentation is as
> long as

> o it can be maintained with a text editor on a dumb terminal
> o the (or a) back-end produces HTML with which I'm happy

Provided both by SGML-Tools and SDF. These two points are also one of my
important wishes. I really want to enhance point two:

o the (or a) freely-available backend produces HTML with which I'm happy

Because FrameMaker or WinWord for the Postscript format is not an option for
us, I think. That's why the author of SDF currently tries to provide us with a
better solution.

> I'd rather not install another ten tons of cruft.. er,
> single-purpose tools on my system. I'd be particularly happy with
> something Perl-based (which I think your recommended tools already
> are, Ralf). Perl rocks.

Correct, Perl rocks. And this is why I also like both SGML-Tools and SDF. Both
are written mainly in Perl and can be easily ported. And both are a little bit
like "closed world approach", i.e. mainly we just need this package for HTML
generation. Only TeX is needed for Postscript, etc.

>[...]
> One nice thing about the current arrangement is that the HTML files
> at the above location are *very* easy to maintain - they're just
> checked out from a CVS library. Any replacement system can only be
> more complicated, but it *must* allow easy regeneration of the site
> documents in HTML.

Correct, currently we login to taz and do just a "cvs update ...." to bring
the webserver areas up-to-date. Ok with a meta-language we need one step more,
but that should be no problem.

> Also, I'm already working on the FAQ and making it
> news.answers-ready, so that's another data-point. However,
> conversion to that format, if not automatic, should be a simple
> matter of Perl. I'd appreciate it if anyone wanting to do a
> wholesale reorg of the FAQ would coordinate with me first. (I'm
> including you, Ralf! ;-)

Yes, I'm currently only started a test for the Apache Handbook.
The same approach can be done for the Apache FAQ similar to the
FreeBSD FAQ. But this is not directly news.answers-ready.

Ralf S. Engelschall
rse@engelschall.com
www.engelschall.com