[Starmap2-devteam] Fw: [rpg-tools] Re: Makeover for an old tool

[Trace K. <Tra...@ne...>]

----- Original Message -----=20
From: Robert Brix Kjelgaard=20
To: Trace Kern=20
Sent: Saturday, October 13, 2001 7:17 PM
Subject: Re: [rpg-tools] Re: Makeover for an old tool


Well, I agree with some of that- these are the main places I think I =
disagree:
=20
I think I already mentioned my concern about going multi-platform- I'd =
rather solve one or two problems at once, then add to the complexity =
once I understand the "domain" or "problem space" better.
=20
Coordinating processes (or apps, if you prefer) on Windows isn't all =
that hard- you do have to understand multitasking, inter-process =
communication and synchronization issues, but they teach that sort of =
thing these days- heck they did in the 70's when I was in college... =20
=20
Flat text files are very simple to work with- they're also very simple =
to screw up and crash your program.  Techies love them- real people who =
just want to play a game once in a while don't.  There are plenty of =
ways to make visually appealing editors for this info, and to store it =
in less space.  There are also plenty of ways to "modularize" the =
program to keep the interface clean.  Besides, I'm a professional- an =
interface that isn't clean and elegant isn't something I want my name =
on, anyway.
=20
Along the same lines, I agree that you want people to be able to link in =
pieces from other maps, etc.  But there are ways to do all of that, even =
without text files (or even with them, you just have to understand what =
the OS can actually do for you).
=20
As far as the structure and design of the software, I should look more =
closely before I say something too rash, but so far, I think it's a =
specious claim- the code's structured the way it is because of the =
limitations of QBasic and DOS, not because it's such a great thing.  =
It's good solid work, mind you- it's just not groundbreaking in any way. =
 Of course, I probably spend most of my time dealing with well-written =
stuff, and this person may spend a lot of their time with the kind of =
stuff you get out of "beginners".  (But I was a beginner once too- every =
experienced programmer knows the feeling of going back and looking at =
something they wrote, and thinking- "was I really that awful?")
=20
My last comment is that what he's talking about at the end is a =
distributed RPG played over the web.  That certainly isn't what this is =
at the moment, and I'm not quite ready to go there (that might be a good =
place to change programmers or seek to add new ones).
=20
For now, I'd like to start by doing essentially a straight "port" of the =
existing program.  The GUI may not be stellar (no pun intended), and the =
features may not have changed, but once I've done that, I'll have a =
better feel for how it all REALLY works.  At that point, I can talk =
about how to break down the design so it's easier to add features, =
figure out how to implement more of the features you mention, etc.  =
Windows has support for technologies (COM/ActiveX/etc) that we can build =
into this pretty easily, and they let you do things like add features to =
the existing program without screwing it all up.  In fact, one reason I =
chose C++ is that the language itself has features supporting this kind =
of thing.
=20
BTW, another option for going multiplayer on the web and using "ASCII =
files" is a technology (an internet standard) called SOAP.  I've used =
toolkits from Microsoft that support it, and (this is more common with =
them than you might think) they're free- probably in part because it's a =
key part of their .NET (Microsoft monopolizes the web) strategy.  We can =
do SOAP with Windows-based servers easily enough, and the neat thing is =
that it's "clean" enough that you can also make it work on the other =
platforms (there are some SF projects already related to all of that).
----- Original Message -----=20
  From: Trace Kern=20
  To: Robert Brix Kjelgaard=20
  Sent: Saturday, October 13, 2001 3:06 PM
  Subject: Fw: [rpg-tools] Re: Makeover for an old tool


  In one of my many cries for help, OF wasn't the only place I looked, =
it seems I got another valuable bite.

  Take a look over this email and tell me what you think, as a lot of =
what he's talking about is a bit over my head in technical detail.


  ----- Original Message -----=20
  From: ey...@ya...=20
  To: rpg...@ya...=20
  Sent: Saturday, October 13, 2001 1:44 PM
  Subject: [rpg-tools] Re: Makeover for an old tool


  --- In rpg-tools@y..., "Trace Kern" <TraceFox@t...> wrote:
  > Anyone wanna help me give a very promising game-system independant=20
  > Sci-Fi RPG-maker tool a facelift for the new millenium? Heh!

  --- In rpg-tools@y..., "Trace Kern" <TraceFox@t...> wrote:
  > Anyone wanna help me give a very promising game-system independant=20
  > Sci-Fi RPG-maker tool a facelift for the new millenium? Heh!

  This is something I've had on my mind for quite some time.  Jim's
  program is by far the best of its sort that I've ever seen, but no dos
  app has ever aged very gracefully.  The time has definitely come for a
  successor.

  The strength of Starmap is its complexity and scalability, especially
  the way it interlocks smaller programs, and allows for all sorts of
  flexibility.  All the data files are flat ascii dbs, and are well
  commented, making manipulation by hand or script extremely easy.

  The problem with Starmap is that there is far far far too much data,
  and not enough tools to handle it within the main interface.  There
  are no filters, no statistics.  Also, once the stars are generated,
  turning all those stars into a galactic empire is a near superhuman
  task (at least in my experience, but it might just be my attention
  span).  Also, the DOS thing is an issue, although the interface is
  remarkably clean and easy to use.

  The problems can I see with simply taking Starmap and improving it,
  but for Windows, are:

  - Environment.  I'm speaking outside of my expertise here, but I'm
  going to assume that it's hard to make separate programs interact
  within the windows environment, mainly because I've never really seen
  it happen outside of Microsoft apps.  Yet I already have argued that
  Starmap's strength is its open, sum-of-the-parts approach.

  - Centralization.  One reason why all starmapping programs are hard to
  work with is that only one person can work on the universe at a time,
  and transferring data files is a pain.

  - Platform.  I happen to use Linux for my desktop, and I know many
  people who use Macs.  While there are compatibility layers for Windows
  apps, they are usually a pain.  Dual-boots even more so.

  - Interface.  Perhaps I'm being cynical, but I haven't seen a Windows
  starmapping program yet that I liked for its interface.  A large part
  of replacing Starmap would be making sure the interface stays clean
  and facile.

  - Complexity.  Starmap used flat ascii files to store its data, and
  while this was easy to edit, it resulted in a /very/ unwieldly
  directory structure.  Also, Starmap's data files were relatively
  simple-- essentially coordinates and a few other vital statistics.=20
  Ideally, we would want to describe stars and planets more completely,
  but complex data and flat ascii files are poor company.  I've seen a
  starmapping program (I forget which one now) which, while its level of
  detail was unmatched, had data files which were, dare I say, worse
  than Gliese 3.0 raw text files in their arcanity and complexity.

  So what is my solution?  I think that the ideal Starmap would, at its
  heart, be an SQL database with a Perl/CGI or PHP web interface and
  perhaps even a multi-platform client.  The benefits?

  - All these tools are open-source and freely available, and promote a
  modular development approach. =20
  - An rpg group that was developing a new universe could all work on
  the project together.  Being networked, you could access it from home,
  from work, from school, from anywhere. =20
  - Mac hacks, Linux gurus, Microsoft junkies and anyone else could have
  equal access.
  - SQL servers are robust and would swallow up a universe of reasonable
  complexity without a burp.  No more corrupted data files!

  I've got some preliminary brainwork residing at
  "http://eykd.net/open/".  I call it "OpenSphere".  There's a very
  basic data structure in MySQL and some incomplete Perl script for
  generating planets.  I welcome anyone, especially programmers, who
  would be interesting in seeing this sort of thing become reality.  Let
  me know what you think.


        Yahoo! Groups Sponsor=20
              ADVERTISEMENT
               =20
      =20
      =20

  To unsubscribe from this group, send an email to:
  rpg...@eg...



  Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.=20