Re: what can I do to help?
On 5/28/05, Bill Allombert <ballombe@master.debian.org> wrote:
> Debian use SGML. HTML is a subset of SGML, so there is not
> a lot of difference between Gentoo and Debian approach.
> I am sure Gentoo could write a xsl transformer to generate
> plain-text output.
Gentoo writes their docs in XML, and then transform.
Their web site runs the transformation on the fly
for some of the linked documentation. I guess the
same sort of thing happens in Debian Doc, but I have
not yet seen the inside of the tool debiandoc2html.
> Plain-text files are ten times more convenient to read that HTML or PDF
> because you can process them arbitrarily with standard UNIX tool and
> your favorite editor. How many browsers support searching inside a
> HTML document made of several files and links ?
Can you give an example of something an end user
would need to do like this? I can grep on text or HTML,
unless there was a tag that was in common with something
I was searching for. I can't imagine why anyone would be
running something like sed or awk, etc. on the documentation,
aside from documentation maintainers.
All graphical browsers I've used in recent years have
search functions that work on the single open document.
> Furthermore a teletype terminal can display HTML, using tools like
> 'lynx -dump', but actually, Debian supporting S390 mainframes that have
> no screen nor keyboard, I would not be surprised if someone read the
> documentation on a dumb terminal.
If there are users who have no access to something
with a screen, I guess that is a real need.
> Yes, but it is not a limitation of their design, they just do not
> realise the benefit of plain text documents.
Well, like you said, it would likely be possible to
make a transformation to plain text with the same tools.
What I'm seeking to discover is how things can be trimmed
down to eliminate extra steps and required compatibilities.
Regards,
Donald
Reply to: