Re: Latest texi/info to www

[Harry Putnam]

Eli Zaretskii <address@hidden> writes:

> On 18 Jul 2001, Harry Putnam wrote:
> 
> > Thanks Eli.  I think I may have expressed myself poorly.  I knew about
> > html converting capability and about texi2html.  However I understood
> > these methods did a basic conversion and then you had an html file you
> > could put on the internet.  Not `on the fly'.
> 
> No, not on the fly.  I've read your citations, and I still don't see why 
> an on-the-fly conversion would be a good idea.  Keep in mind that large 
> parts of the markup information in the Texinfo source is gone in the Info 
> output, because Info is (amost) a pure-ASCII format.  So you can't 
> reproduce the same display from Info as you can from Texinfo.

As you have no doubt guessed, I know next to nothing about texinfo
mainly just the `info' end of it where someone else has done the hard
part and all I do is read.....

It hadn't occured to me that information would be lost in the
process. So I guess that could be a big stumbling block.  What kind of
things end up missing?  Do you mean someting like the Notes, cross refs
and so on?  Can you give a simple example of what it would mean in a
web browser?

> is the only reason for this the desire to avoid producing HTML output 
> from every Texinfo manual?  Or is there any other reason?

I prefer info style documentation to any of the other common ones, in
must cases.  Especially when accessed with emacs.  Although stand
alone info has some good functionality missing in emacs.  However,
that is not viable in remote settings.  At least not natively.

I had visions of browsing info files remotely and having all the
functionality of browsing them with info or emacs (so far as the info
functions, I mean).  If this were possible, it would be much nicer to
have only one data base to tend to.  Updating info or installing
things etc would be automatically available to the remote browsing.

I think that was the only consideration.

> No, AFAIK, this feature was never considered for inclusion, or even 
> discussed.

Is that because of the information loss you cite, or are there other
things that make it a bad idea?