GRASS 5.7 man pages - was: Re: [Pkg-grass-general] Neteler points finger at Debian for delays
On Thu, Nov 25, 2004 at 12:17:15PM +0100, Paolo Cavallini wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> Francesco, could you make specific raccomendations? Then we could forward them
> to grass developers, as I believe they'll be happy of improving their
> At 10:25, giovedì 25 novembre 2004, Francesco P. Lovergine has probably
> > BTW, I would know WHY grass developers insist in distributing formatted
> > man pages. It's against our policy and any common sense too.
> > Any decent man program can produce formatted pages on fly and this habit
> > prevents creating different formats (e.g. PS or PDF) from them.
I don't fully understand your comment above.
Currently (from today onwards) in GRASS 5.7 it works like this:
- first a GRASS command get's compiled
- then it is run immediately in a fake session in a local
"demo location" (a valid minimalistic GRASS database which
comes within the source code) to render the documentation
in HTML format
To be more details: the fake session renders the header of
the HTML page with parameters/flags of the command, then
the Makefile adds a local 'description.html' file, merges
both to create valid HTML code. Then this file and, if present,
local PNG/JPG graphs are copied over to the docs subdir.
- after compilation MAN pages are rendered from the HTML pages
No good like that? Better suggestions?
The background is that we wanted to semi-automate the procedure,
especially to keep parameters/flags descriptions always up to
> > I'm of the opinion to remove completely those crap pages and
> > creating a simple grass.1 to point the html docs. Then, forwarding
> > upstream every single complain about missing man pages...
> > On Thu, Nov 25, 2004 at 06:36:23AM +0800, Dan Jacobson wrote:
> > > While waiting for GRASS 5.7 on Debian sid, and noting that current
> > > GRASS lost even its man pages, I read
This comment I don't understand. Man pages are there.