[Freeciv-Dev] Re: do you really want to work on the ~ (was: registry)

["pasky@xxxxxxxxxxx" <dbsmith@xxxxxxxxxxx>]

Originally to: David Bryan Smith

  by provide.net (CommuniGate Pro SMTP 3.5b5)
  with ESMTP id 6258185 for dbsmith@xxxxxxxxxxx; Wed, 05 Dec 2001 14:25:42 -0500
Received: from pi (localhost [127.0.0.1])
        by pi.glockenspiel.complete.org (Postfix) with ESMTP
        id 9DBF03BA78; Wed,  5 Dec 2001 14:25:26 -0500 (EST)
Received: with LISTAR (v1.0.0; list freeciv-dev); Wed, 05 Dec 2001 14:25:26 
-0500 (EST)
Delivered-To: virtual-freeciv.org-freeciv-dev@xxxxxxxxxxxx
Received: from machine.sinus.cz (pasky.ji.cz [62.44.12.54])
        by pi.glockenspiel.complete.org (Postfix) with SMTP id D37FD3B854
        for <freeciv-dev@xxxxxxxxxxx>; Wed,  5 Dec 2001 14:25:24 -0500 (EST)
Received: (qmail 17593 invoked by uid 2001); 5 Dec 2001 19:25:23 -0000
Date: Wed, 5 Dec 2001 20:25:23 +0100
From: Petr Baudis <pasky@xxxxxxxxxxx>
To: freeciv-dev@xxxxxxxxxxx
Subject: [Freeciv-Dev] Re: do you really want to work on the ~ (was: registry)
Message-ID: <20011205192523.GN19852@xxxxxxxxxxx>
References: <E16BcSO-0000lK-00@xxxxxxxxxxxxxxxxxxxxxxxxx> 
<20011205170113.B19770@xxxxxxxxxx>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20011205170113.B19770@xxxxxxxxxx>
User-Agent: Mutt/1.3.23.2i
X-archive-position: 5093
X-listar-version: Listar v1.0.0
Sender: freeciv-dev-bounce@xxxxxxxxxxx
Errors-To: freeciv-dev-bounce@xxxxxxxxxxx
X-original-sender: pasky@xxxxxxxxxxx
Precedence: bulk
X-list: freeciv-dev

Dear diary, on Wed, Dec 05, 2001 at 05:01:13PM CET, I got a letter, where
Reinier Post <rp@xxxxxxxxxx> told me, that...
> Now the ruleset format is undocumented and inadequate - it doesn't really
> support nested structures.  So perhaps it's better to use libxml2?  libxml2
> depends on libm (which Freeciv doesn't depend on I think) and libz, which
> Freeciv depends on for compression, but nothing else.
I think we don't need XML at all. It at least looks much less readable and
comfortablee for me. I would be completely happy with C-like ruleset syntax -
i.e.:

building {
  id = "Airport";
  name = _("Airport");
  tech_req = "Radio";
  effect {
    type = "Unit_Veteran";
    range = "City";
    aff_unit = "Air";
  }
  effect {
    type = "Airlift";
    range = "City";
  }
  helptext = _("Blablabla\
blabla\n\nbla\
blah!");
}

It would be IMHO much cleaner than current one, and actually not so much
different. As a parser, I don't see anything particulary wrong on flex+bison.
Easy to use, portable, effecient.

--

                                Petr "Pasky" Baudis

UN*X programmer, UN*X administrator, hobbies = IPv6, IRC, FreeCiv hacking
.
  "A common mistake that people make, when trying to design
   something completely foolproof is to underestimate the
   ingenuity of complete fools."
     -- Douglas Adams in Mostly Harmless
.
Public PGP key, geekcode and stuff: http://pasky.ji.cz/~pasky/