Getting bored with developing my RL, what to do?
personjerry
It's the continued version of the post from another thread:
google groups link:
http://groups.google.com/group/rec.games.roguelike.development/browse_thread/thread/4df5a52b6f32c8e6/2ca08445e95ee30b
Anyway, here's the zipped file
http://www.filefactory.com/file/a1bb0h8/n/evrl.zip
My code is really bad, I know.
Even when I see it I get disappointed in the bad code in there, so
that discourages me from working on it...
So I'd like to get your feedback, see how I could improve this bit by
bit, or even fix a little up for me? :D
Also, it's STILL my first roguelike, don't be too harsh please :)
purplearcanist
> Well guys here's my RL... only included a linux executable, sorry.
>
> It's the continued version of the post from another thread:
>
> Anyway, here's the zipped filehttp://www.filefactory.com/file/a1bb0h8/n/evrl.zip
>
> My code is really bad, I know.
> Even when I see it I get disappointed in the bad code in there, so
> that discourages me from working on it...
> So I'd like to get your feedback, see how I could improve this bit by
> bit, or even fix a little up for me? :D
>
> Also, it's STILL my first roguelike, don't be too harsh please :)
My reccomendation, if you are bored, is to halt work on it. It may
seem harsh, considering that you have already put a lot of work on
it. Instead, work on a 7DRL.
Krice
> My code is really bad, I know.
I don't know, the .zip seems to be corrupted.
If you know your code is bad, why you write bad code in
the first place?
Juho Julkunen
@k19g2000yqc.googlegroups.com>, Krice (pau...@mbnet.fi) says...
Because bad code is better than nonexistent code, perhaps.
--
Juho Julkunen
Ray
> If you know your code is bad, why you write bad code in
> the first place?
I think you have to write a lot of bad code when you set
out to become someone who can write good code.
Just like novelists who generally agree that the first
million words someone writes are going to be crap so you
should just pound them out and get it over with so you can
become a good writer.
Bear
skreeg
coding very tentatively and concentrating too much on doing things
right first time however that doesn't get production moving very fast.
Now I've hammered out so much code in the last few months and I'm
always going back and reworking the design. While I do this I always
come across bits that are "near' the bit I'm working on and I think
"What complete numpty wrote this?" and I get sidetracked while I tidy
that bit up. The good thing is I see myself developing as a go and the
mistakes I made back then are more obvious.
Write lots of code. Good or bad it doesn't matter. That's what a)
get's your game out the door and b) helps you develop as a programmer.
Gerry Quinn
3c444f...@x6g2000prc.googlegroups.com>, skr...@gmail.com says...
With C++ I found that initially I started making class hierarchies far
too deep, then I reacted against that and made them too shallow, and
only afterwards found some sort of a happy medium.
- Gerry Quinn
Krice
> I think you have to write a lot of bad code when you set
> out to become someone who can write good code.
Maybe true, but when I started writing Kaduria I had no
idea how bad my source code was. In this case the author
seems to know it, so one would assume that he can do
something about it.
personjerry
In general, I do not work on it a lot at a time, so it's been a bit of
code here and there, and as I've learned more things my code has
improved.
I think that reworking the old code is not as much "fun" as writing
new code, but the new code requires additions in the old code, so I'm
stuck. I wonder is it worth it to go through the code, or just start
anew?
Here's another upload, i can unzip my file just fine so it might be
the upload service.
BirdoPrey
> Well guys here's my RL... only included a linux executable, sorry.
>
> It's the continued version of the post from another thread:
>
> Anyway, here's the zipped filehttp://www.filefactory.com/file/a1bb0h8/n/evrl.zip
>
> My code is really bad, I know.
> Even when I see it I get disappointed in the bad code in there, so
> that discourages me from working on it...
> So I'd like to get your feedback, see how I could improve this bit by
> bit, or even fix a little up for me? :D
>
> Also, it's STILL my first roguelike, don't be too harsh please :)
If your code is amateurish enough to fill you with despair when you
think about working on it, there's a good chance you'll be happier
with a complete rewrite. When I first tried to write a roguelike my
code was abysmal. Scrapped that project and started anew, things are
going much better now. Don't delete the old code though... keep it
around and scavenge what you can from it.
Martin Read
>Maybe true, but when I started writing Kaduria I had no
>idea how bad my source code was. In this case the author
>seems to know it, so one would assume that he can do
>something about it.
With code as well as with prose, it's easier to *see* bad quality than to
*produce* good quality.
--
\_\/_/ turbulence is certainty turbulence is friction between you and me
\ / every time we try to impose order we create chaos
\/ -- Killing Joke, "Mathematics of Chaos"
Jotaf
> Don't delete the old code though... keep it
> around and scavenge what you can from it.
That's actually the best advice IMO. For instance, my C++ RL was
getting too complex, to the point when adding minor things required
big refactoring of the structure (we've all been there I suppose), so
I started rewriting in python, just to try it out -- I decided that at
the first signs of doubt about this rewrite I'd go right back to the
old codebase (it didn't help that I just read about one tutorial for
the language and mucked around with a couple of tiny test programs).
Some parts, like rendering and the object system, were direct ports: I
copied them wholesale and just changed the syntax. The other parts
were rewritten at a much faster pace than in the first codebase and it
was eventually surpassed in features. So, in a nutshell, I rewrote it
cautiously, tried to evaluate at all times if it was time to bail on
the new thing, took what I needed from the old, and I'm pretty happy
with the result.
Jotaf
personjerry
I think i'll try that, make a new rl while trying to mainly adapt the
good old code back in (no pun intended). Looking over the zip, this
wasn't my best revision, actually, heh.
Thanks for the comments guys.
keith
> My code is really bad, I know.
Even though it's bad, try to make it multiplayer. If you see some users
using your game, it should motivate to develop it further