[crossfire] Documentation / handbook / playbook / spoilers

[Mark Wedel]

Nicolas Weeger wrote:
> Hello.
> 
> The doc subdirectory (apart hopefully Developers) is a slight mess, with many 
> many non working or obsolete things.
> 
> Currently, there following documentations are defined:
> * spoiler book
> * player book
> * spell list
> * handbook
> 
> The generation is broken for many things, so we should either fix or remove :)
> Ideally, we should almost be able to generate the spoilers at 
> http://crossfire.real-time.com/spoiler/index.html 
> And probably some handbooks or guides. Maybe not everything, but probably 
> things that depend on server settings/archetypes/artifacts.

  At one point the spoiler and handbook generation did work.

  That said, the non html version required other various tools, notably I think 
latex and dvi2ps or something.  IMO, the non html versions could probably go 
away - while the latex ones generate nicer versions to print out (table spanning 
pages is better done, etc), I really can't see many people printing that data 
out.  Or maybe to the point, I don't see the format being that much nicer to 
have the effort of two versions of basically the same thing (for the rare 
occasion someone does want to print it out, they can still do the html version).

  Taking a very quick link at the URL provided, I note that some of that data 
appears to be data from the spoiler, but better formatted.  It certainly 
wouldn't be hard to change the spoiler layout, but probably more important to 
get it to build.


> I'd be partisan to fix, because (game) documentation is important.
> 
> Here are different things that can be generated automatically:
> * standard item list (monsters, artifacts, alchemy formulae, ...)
> * basically, anything that can be extracted, from server or maps (mapper could 
> list items in the maps themselves if needed - of course, that would 
> documentation would be for "standard" maps)

  Extracting from maps gets to be more a problem - the current spoiler basically 
pulls data from the archetype or other data files, because that is easy/quick to 
do.  Pulling data from all maps is more a problem because it would require going 
through all the maps, a bit more time consuming (and something that the server 
can't do by default - the server obviously knows how to read the archetypes and 
other files, and thus spit out the data in a format that the generation scripts 
want - there isn't any logic right now for the server to traverse all the maps 
and trying to pull out useful information.

  It looks like some of the problem with the generation code is hardcoded server 
name (../../server/crossfire) - that probably also breaks in the case of 
compiling in seperate directory - don't know for sure.  Hopefully that isn't too 
hard to fix.