Language rewrite - any success stories?

On Jan 29, 12:48 pm, g....TakeThisOut@mail.ru (Timofei Shatrov) wrote:
> On Tue, 29 Jan 2008 09:48:17 +0100, Jakub Debski <debski.ja....TakeThisOut@wp.pl> tried to
> confuse everyone with this message:
>
> >awhite pretended :
SNIP
>
> and recompile just this line which takes a millisecond. Or execute
>
> (setf number-of-monsters-per-level 100)
>
> The game could be already running during all of that, so you just go to
> a newly generated level and test the balance.

Or you could have an external properties file with the number-of-
monster-per-level
--
Slash
 >> Stay informed about: Language rewrite - any success stories? 

On Jan 29, 12:06 am, awhite <spud... DeleteThis @iinet.net.au> wrote:
> A ramble.
>
> In finding a new reason not to work hard at my game, I've lately become
> enamoured of Common Lisp.

Dont let love guide this sort of hard choices. Everything that starts
has an end... in time your love for Lisp will die, and you may regret
your choice..

> But the edit/compile/link/test cycle is getting me down. I've been
> looking for a faster-to-develop language, and Paul Graham's essays have
> convinced me that I'm missing out on something big.

What's wrong about that cycle? if you plan ahead your development and
build your components carefully, you wouldnt have to do this too
often... just to see if your work works

> So... Has anyone else tried a total rewrite in a different language? And
> succeeded? There are probably a fair few attempts out there for various
> projects - crawl5 (short-lived as it was) and langband come to mind. Has
> anyone else succeeded?

I haven't had a reason to do it, but I could tell you about several
people who "failed".

>
> Adam

--
Slash
 >> Stay informed about: Language rewrite - any success stories? 

On Jan 29, 4:15 pm, g....TakeThisOut@mail.ru (Timofei Shatrov) wrote:
> On Tue, 29 Jan 2008 10:30:06 -0800 (PST), Slash <java.ko....TakeThisOut@gmail.com> tried to
> confuse everyone with this message:
>
>
>
> >On Jan 29, 12:48 pm, g....TakeThisOut@mail.ru (Timofei Shatrov) wrote:
> >> On Tue, 29 Jan 2008 09:48:17 +0100, Jakub Debski <debski.ja....TakeThisOut@wp.pl> tried to
> >> confuse everyone with this message:
>
> >> >awhite pretended :
> >SNIP
>
> >> and recompile just this line which takes a millisecond. Or execute
>
> >> (setf number-of-monsters-per-level 100)
>
> >> The game could be already running during all of that, so you just go to
> >> a newly generated level and test the balance.
>
> >Or you could have an external properties file with the number-of-
> >monster-per-level
>
> Well, what about tweaking monsters' AI, or something like that?

Or you could have an external script (LUA?) defining the monster's
AI

(Sorry, couldn't resist)

> Also, you'll
> need to somehow force the game to reload this properties file, though this is a
> minor inconvenience. And also such file is a great enabler for cheating, and I'm
> notoriously determined not to make cheaters' life easier

True, true ... >D
>
> --
> |Don't believe this - you're not worthless ,gr---------.ru
> |It's us against millions and we can't take them all... | ue il |
> |But we can take them on! | @ma |
> | (A Wilhelm Scream - The Rip) |______________|

--
Slash
http://roguetemple.com
http://slashie.net
 >> Stay informed about: Language rewrite - any success stories? 

Interesting that this should come up today. Yesterday I decided I was
finally going to get around to learning Lisp, so I got a book from the
library this morning and I've been working my way through it. It's an
omen!
Probably this will mean I end up doing a total rewrite of the game I
was working on (which happens to be a roguelike) in Lisp. Fortunately
for me this was at a very early stage (lots of ideas, not much code)
so this is not a very big task.

Everyone who is talking about compilation speed: The Lisp fan club
claim that Lisp is faster for development in terms of how long it
takes you to write code (see http://www.paulgraham.com/avg.html for a
prominent example of this). Compile time isn't the big issue here -
it's the expressiveness of the language. Or so they say. This is not
why I'm learning it, I could write a game in C++ much faster than
learning a whole new language, but I will be interested to discover
the truth of the matter.

On Jan 29, 5:06 am, awhite <spud....DeleteThis@iinet.net.au> wrote:
> A ramble.
>
> In finding a new reason not to work hard at my game, I've lately become
> enamoured of Common Lisp.
>
> Don't get me wrong... C++ is a great language. It's a fun language. It's
> very very powerful, and I've invested a lot of money in books on
> techniques, structure, coding styles, "don't do that" practices, and
> others. But it's quite slow to develop in. An edit-compile-link-test run
> takes a lot of time. Tools can speed that up, but on linux (my primary
> platform), there's really only g++ and icc.
>
> Common Lisp is fun, quick, and just as capable for roguelike development
> (a couple of quite good roguelikes have been written in it).
>
> Conventional wisdom is to never rewrite, but to refactor, refactor,
> refactor. I agree with that, so my current codebase is uglier than
> necessary - it still contains vestiges of being written for multiple
> executable targets (server, client, and client/server), multiple players,
> and even multiple simultaneous displays (SDL and curses, though I'm
> tempted to keep those).
>
> But the edit/compile/link/test cycle is getting me down. I've been
> looking for a faster-to-develop language, and Paul Graham's essays have
> convinced me that I'm missing out on something big.
>
> So... Has anyone else tried a total rewrite in a different language? And
> succeeded? There are probably a fair few attempts out there for various
> projects - crawl5 (short-lived as it was) and langband come to mind. Has
> anyone else succeeded?
>
> Adam
 >> Stay informed about: Language rewrite - any success stories? 

In article <479eb44e$0$9770$5a62ac22@per-qv1-newsreader-
01.iinet.net.au>, spudboy RemoveThis @iinet.net.au says...
>
> A ramble.
>
> In finding a new reason not to work hard at my game, I've lately become
> enamoured of Common Lisp.
>
> Don't get me wrong... C++ is a great language. It's a fun language. It's
> very very powerful, and I've invested a lot of money in books on
> techniques, structure, coding styles, "don't do that" practices, and
> others. But it's quite slow to develop in. An edit-compile-link-test run
> takes a lot of time. Tools can speed that up, but on linux (my primary
> platform), there's really only g++ and icc.

MSVC with pre-compiled headers makes the edit-compile-run cycle almost
instant. I just checked - on my roguelike Lair, a partial compile in
debug takes about 5 seconds, while a full rebuild/link in either debug
or release mode is about 15-20 seconds.

Four-year-old laptop, good at the time.

- Gerry Quinn
--
Lair of the Demon Ape (a coffee-break roguelike)
<http://indigo.ie/~gerryq/lair/lair.htm>
 >> Stay informed about: Language rewrite - any success stories? 

In article <479ee294$0$10842$5a62ac22@per-qv1-newsreader-
01.iinet.net.au>, spudboy.DeleteThis@iinet.net.au says...

> Using the REPL (Read Evaluate Print Loop) just seems faster, as
> essentially there's no real difference between writing a program, and
> using the debugger. You can write a function and debug it straight away.
> No need to stop, compile, fire up the debugger, edit, recompile, fire up
> the debugger, ad nauseum. That's the bit about c++ that's currently
> depressing me.

Code and debug instantly is how it is with C++ for me - the delay is at
most a few seconds.

- Gerry Quinn
--
Lair of the Demon Ape (a coffee-break roguelike)
<http://indigo.ie/~gerryq/lair/lair.htm>
 >> Stay informed about: Language rewrite - any success stories? 

On 29 Jan, 18:34, Slash <java.ko....RemoveThis@gmail.com> wrote:
> Dont let love guide this sort of hard choices. Everything that starts
> has an end... in time your love for Lisp will die, and you may regret
> your choice..
> <snip>
> --
> Slash

Yeah, it's like marrying in Vegas, out of a "the hell with it!"-
moment

OTOH I liked the suggestion of writing a 7DRL, just to give you a
better taste of the language and how it relates to RL/game
programming. It's the safest way, and will probably turn into a good
learning experience. But don't try to make it into your DreamRL in the
first try.


....I wasn't gonna say this, but...

....you're missing out on the bandwagon. LISP isn't hot anymore,
Haskell is -- it's purely functional


Jotaf
 >> Stay informed about: Language rewrite - any success stories? 

grue.TakeThisOut@mail.ru (Timofei Shatrov) writes:

> On Tue, 29 Jan 2008 10:30:06 -0800 (PST), Slash <java.koder.TakeThisOut@gmail.com>
> tried to confuse everyone with this message:
>
>>On Jan 29, 12:48 pm, g....TakeThisOut@mail.ru (Timofei Shatrov) wrote:
>>> On Tue, 29 Jan 2008 09:48:17 +0100, Jakub Debski
>>> <debski.ja....TakeThisOut@wp.pl> tried to confuse everyone with this message:
>>>
>>> >awhite pretended :
>>SNIP
>>>
>>> and recompile just this line which takes a millisecond. Or execute
>>> (setf number-of-monsters-per-level 100) The game could be already
>>> running during all of that, so you just go to a newly generated
>>> level and test the balance.
>>
>>Or you could have an external properties file with the number-of-
>>monster-per-level
>
> Well, what about tweaking monsters' AI, or something like that? Also,
> you'll need to somehow force the game to reload this properties file,
> though this is a minor inconvenience.

Indeed. Just moments ago I did this. I wanted to walk up to an enemy and
hit him (for testing purposes; I'm not vicious) but he wouldn't stay
still. Since I had access to my program, it was trivial to change his AI
to make him quit moving around.

Sure, in C++ I could have quit the game, changed the code, waited for it
to compile, fixed my typo, successfully compiled, started the game, and
gone to find an enemy. And then done the whole thing in reverse when I
wanted his old behavior back. But I think I'd kind of rather be able to
do it instantly without disturbing the game world.

I don't know why this group is so skeptical about the value of
interactive development. It's a big time saver. While a couple of
seconds (minutes?!) of time wasted on compilation now and then may not
add up to a lot of hours wasted, it's awfully tedious, and means you
can't test your new code as easily as you otherwise could.
 >> Stay informed about: Language rewrite - any success stories? 

On Jan 29, 5:12 pm, Paul Donnelly <paul-donne....RemoveThis@sbcglobal.net> wrote:
> g....RemoveThis@mail.ru (Timofei Shatrov) writes:
> > On Tue, 29 Jan 2008 10:30:06 -0800 (PST), Slash <java.ko....RemoveThis@gmail.com>
> > tried to confuse everyone with this message:
>
> >>On Jan 29, 12:48 pm, g....RemoveThis@mail.ru (Timofei Shatrov) wrote:
> >>> On Tue, 29 Jan 2008 09:48:17 +0100, Jakub Debski
> >>> <debski.ja....RemoveThis@wp.pl> tried to confuse everyone with this message:
>
> >>> >awhite pretended :
> >>SNIP
>
> >>> and recompile just this line which takes a millisecond. Or execute
> >>> (setf number-of-monsters-per-level 100) The game could be already
> >>> running during all of that, so you just go to a newly generated
> >>> level and test the balance.
>
> >>Or you could have an external properties file with the number-of-
> >>monster-per-level
>
> > Well, what about tweaking monsters' AI, or something like that? Also,
> > you'll need to somehow force the game to reload this properties file,
> > though this is a minor inconvenience.
>
> Indeed. Just moments ago I did this. I wanted to walk up to an enemy and
> hit him (for testing purposes; I'm not vicious) but he wouldn't stay
> still. Since I had access to my program, it was trivial to change his AI
> to make him quit moving around.

I miss this sort of functionality when working on POWDER. There is a
lot of #if 0 code whose purpose is just to set up situations for
debugging. One of my plans is to add a wish engine - not for a wand
of wishing, but just for testing.

> Sure, in C++ I could have quit the game, changed the code, waited for it
> to compile, fixed my typo, successfully compiled, started the game, and
> gone to find an enemy. And then done the whole thing in reverse when I
> wanted his old behavior back. But I think I'd kind of rather be able to
> do it instantly without disturbing the game world.

MSVC has had "edit and continue" for years now. I can break my game,
change the code, hit continue and it will recompile the relevant files
and restart me where I left off. Since only a few lines of one file
changed, this turn around time is likely the same as that in an
interpreted language.

Interactive development isn't something unique to interpreted
languages.

> I don't know why this group is so skeptical about the value of
> interactive development. It's a big time saver. While a couple of
> seconds (minutes?!) of time wasted on compilation now and then may not
> add up to a lot of hours wasted, it's awfully tedious, and means you
> can't test your new code as easily as you otherwise could.

First, I don't recall people saying they doubted there is value to
interactive development. They merely doubt there is enough value to
justify abandoning their favorite languages. I will thus step in and
fulfill your claim by doubting there is any value at all :>

There are two theories of code testing that I subscribe to. Both are,
of course, contradictory.

Theory #1 states that you should write correct code in the first
place. Under this theory, this sort of "just in time" code writing is
a dangerous habit as it fails to punish the author for having made the
mistake. Throwing something together and then fixing what you notice
broken isn't a wise strategy - you will only notice *some* of your
errors. Fixing the errors you find doesn't help the errors you didn't
find. Let's be really generous and say testing and debugging finds
90% of the errors. If my sloppy coding results in 9 errors (quickly
fixed, mind you, in my edit & continue world), we can sadly shake our
head since there is likely an unknown error still there. On the other
hand, the three-minute-to-compile person took more care in time in
writing the algorithm. As a result, their testing found only one
error - of course, taking twice the time to fix. However, they can
rest more easily suspecting that there only one tenth of an unknown
error left in the code.

I remember a game programmer in rec.games.programmer who claimed she
never used a debugger since invoking a debugger is an admission you
don't know your code. I still think that is extreme, but it is
something to keep in mind.

Theory #2 wonders why we can't just have some self discipline and
write good code AND use accelerated tools to debug the mistakes that
do occur. At this point a large debate on human nature starts and
everyone leaves the debate with their own view reinforced.
--
Jeff Lait
(POWDER: http://www.zincland.com/powder)
 >> Stay informed about: Language rewrite - any success stories? 

On Jan 29, 8:17 pm, Paul Donnelly <paul-donne....DeleteThis@sbcglobal.net> wrote:
> Jeff Lait <torespondisfut....DeleteThis@hotmail.com> writes:
> > MSVC has had "edit and continue" for years now. I can break my game,
> > change the code, hit continue and it will recompile the relevant files
> > and restart me where I left off. Since only a few lines of one file
> > changed, this turn around time is likely the same as that in an
> > interpreted language.
>
> > Interactive development isn't something unique to interpreted
> > languages.
>
> That's good to hear, although you seem a little hazy on what constitutes
> an interpreted language.

I am very hazy for a very good reason. "interpreted" is really a
function of the implementation of the language, not the language
itself. BASIC can be compiled and C++ could be interpreted. I think,
however, there is an understanding of their being a sort of divide
between "compile and test" and "change source code on the fly", with
the latter generally made possible through an interpreter.

> >> I don't know why this group is so skeptical about the value of
> >> interactive development. It's a big time saver. While a couple of
> >> seconds (minutes?!) of time wasted on compilation now and then may not
> >> add up to a lot of hours wasted, it's awfully tedious, and means you
> >> can't test your new code as easily as you otherwise could.
>
> > First, I don't recall people saying they doubted there is value to
> > interactive development. They merely doubt there is enough value to
> > justify abandoning their favorite languages. I will thus step in and
> > fulfill your claim by doubting there is any value at all :>
>
> And I merely observed more than the expected level of naysaying about
> interactive development.

Naysaying because the implication was a shift in language, not because
the thing itself was not good. I would be curious what you think
about my claim that interactive development is bad because, like GOTO,
it begets bad habits.
--
Jeff Lait
(POWDER: http://www.zincland.com/powder)
 >> Stay informed about: Language rewrite - any success stories? 

On Tue, 29 Jan 2008 09:48:17 +0100, Jakub Debski <debski.jakub.TakeThisOut@wp.pl> tried to
confuse everyone with this message:

>awhite pretended :
>> But the edit/compile/link/test cycle is getting me down.
>
>In Lisp you don't have to edit/compile/link/test?
>

It's more like edit/compile what you edited/test what you edited. You
don't need to compile the whole file, just the top-level forms you edited
and you don't need to run the entire application to test it, just the specific
function. Moreover you can edit/compile while the application is running and
it will change accordingly.

For example you want to change the number of monsters per level. In C you
edit the line

int number-of-monsters-per-level = 100;

and recompile the whole file, which takes several seconds. Then you start new
game, generate new character, and go to the level to test the balance of
gameplay.

In CL you edit the line

(defparameter number-of-monsters-per-level 100)

and recompile just this line which takes a millisecond. Or execute

(setf number-of-monsters-per-level 100)

The game could be already running during all of that, so you just go to
a newly generated level and test the balance.

--
|Don't believe this - you're not worthless ,gr---------.ru
|It's us against millions and we can't take them all... | ue il |
|But we can take them on! | @ma |
| (A Wilhelm Scream - The Rip) |______________|
 >> Stay informed about: Language rewrite - any success stories? 

Jakub Debski wrote:

> awhite pretended :
>> But the edit/compile/link/test cycle is getting me down.
>
> In Lisp you don't have to edit/compile/link/test?

Not really. The REPL loop makes the compile/link stage instantaneous.

>> So... Has anyone else tried a total rewrite in a different language? And
>> succeeded? There are probably a fair few attempts out there for various
>> projects - crawl5 (short-lived as it was) and langband come to mind. Has
>> anyone else succeeded?
>
> http://www.joelonsoftware.com/articles/fog0000000069.html

Without even following the link, I know which story that'll be.
Good example. But Netscape had *working* code. Ok, it wasn't perfect, but it
worked pretty darn well for a lot of people.

I don't have releasable code. I'm programming purely for my own benefit
and/or amusement. Although I'd like to have a releasable game one day, it's
not a high priority. Much like the original designers of /Rogue/ - I'm
mostly looking for a game I'd enjoy myself.

Mostly, my OP was just ranting out loud. I'm not sure what the right answer
is - C++ is fun, though I'm a little low on motivation at the moment.

Lisp is new and (because of that) more fun, so I'm more motivated.

A
 >> Stay informed about: Language rewrite - any success stories? 

After serious thinking Adam White wrote :
> I don't have releasable code. I'm programming purely for my own benefit
> and/or amusement. Although I'd like to have a releasable game one day, it's
> not a high priority.

You don't really want to finish your game, just to play with
programming and learn something new? Why are you asking about success
rate then?
It will be the success for sure, because you will play with new code
and learn new language.
I don't want to be rude, but this is how I see your posts.

regards,
Jakub
 >> Stay informed about: Language rewrite - any success stories?