Nethack Round-robin #2

G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

The buck has been passed to Mark who will take the game from 1001-2000.

I will ask Mark to post here with an update!

Who's next?

Cheers, Dave
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Dave Hammond wrote:
> The buck has been passed to Mark who will take the game from 1001-2000.
>
> I will ask Mark to post here with an update!
>
> Who's next?

I'd love to have a go at it.

As someone suggested in the other thread, it may be a good idea to try
the next one on NAO, which would at least enforce no cheating and game
compatability. Plus, it'd be fun to watch. :)
 

Manuel

Distinguished
May 6, 2004
97
0
18,630
Archived from groups: rec.games.roguelike.nethack (More info?)

[Nethack round-robin game]

Please sign me up, I'd like to share this experience :)

Mind you, I'm not yet a good player, but probably you
won't be too picky until you really plan on ascending.

What do you say?

--
Manuel.
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Andy Johnson <usenet@take-this.out.swervy.tk> wrote in
news:d3f768$mts$1@joe.rice.edu:

> Canageek wrote:
>> Sorry, but what is NAO?
>>
>
> NAO = nethack.alt.org. If you telnet there you can play nethack on a
> remote server. Or watch other people play; that can be educational.
>

But slow as heck!
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Canageek wrote:
> Sorry, but what is NAO?
>

NAO = nethack.alt.org. If you telnet there you can play nethack on a
remote server. Or watch other people play; that can be educational.
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

"Andy Johnson" <usenet@take-this.out.swervy.tk> wrote in message
news:d3ecj2$rn9$1@joe.rice.edu...
> Dave Hammond wrote:
>> The buck has been passed to Mark who will take the game from 1001-2000.
>>
>> I will ask Mark to post here with an update!
>>
>> Who's next?
>
> I'd love to have a go at it.
>
> As someone suggested in the other thread, it may be a good idea to try the
> next one on NAO, which would at least enforce no cheating and game
> compatability. Plus, it'd be fun to watch. :)

As this one has started...lets leave it to run as a trial, and take it from
there....we can only learn from our mistakes!

Hopefully Mark will do his 1000 and pass it to you!

Cheers, Dave
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Dave Hammond wrote:
> "Andy Johnson" <usenet@take-this.out.swervy.tk> wrote in message
> news:d3ecj2$rn9$1@joe.rice.edu...
> > Dave Hammond wrote:
> >> The buck has been passed to Mark who will take the game from
1001-2000.
> >>
> >> I will ask Mark to post here with an update!
> >>
> >> Who's next?
> >
> > I'd love to have a go at it.
> >
> > As someone suggested in the other thread, it may be a good idea to
try the
> > next one on NAO, which would at least enforce no cheating and game
> > compatability. Plus, it'd be fun to watch. :)
>
> As this one has started...lets leave it to run as a trial, and take
it from
> there....we can only learn from our mistakes!
>
> Hopefully Mark will do his 1000 and pass it to you!
>
> Cheers, Dave

I have done my 1,000 turns (which flew by), and then passed it onto
Andy, along with a notepad (.txt) doc of noteworthy info. I have sort
of posted a rushed update, but set it up as a different subject (a bit
silly of me), but I'll give you a less rushed one now.

Started on Dlv 3, quaff-tested acid and confusion (people say its
dangerous, but I see little harm in quaff testing), then found 2 of the
same type of scrolls one after the other (as well as some rings/wands),
so I read the abovementioned scrolls, thinking it was ID/light, but
turned out to be blessed enchant weapon, so our long sword is now upto
+3... very handy! (with another unknown enchant weapon being held).
Also engrave IDed magic missiles, so that might come in handy.

Just made it into sokoban (after having a minor consultation with the
oracle) and started doing the puzzle when my time ran out(hope Andy
knows his Sokoban!), but no nasty monsters encountered yet (explored
upto Dlv 6). Also got him from Exp lv 2 to 5, so she's now ready to dip
the long sword for excalibur if she wants (though if we can get mojo
through an alter excalibur is somewhat useless, and makes it less
likely to get more artifacts).

One last thing Andy, not that she ('Robin') just became hungry, dont
forget to feed her!!

That should be enough of an update for most of you surely,

Mark
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

David Damerell wrote:

> Quoting Dave Hammond <davidjhammond@hotmail.com>:
>> The buck has been passed to Mark who will take the game from 1001-2000.
>> I will ask Mark to post here with an update!
>> Who's next?
>
> I infer you are using Windows from your implicit assumption that everyone
> uses what you do, but perhaps you should mention it explicitly?

Are saved game files not platform independent? If so, why not?

--
Benjamin Lewis

Evelyn the dog, having undergone further modification, pondered the
significance of short-person behavior in pedal-depressed panchromatic
resonance and other highly ambient domains... "Arf", she said.
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Raisse the Thaumaturge wrote:

> Benjamin Lewis wrote:
>
>> Are saved game files not platform independent?
>
> Not at all. Not even on different versions of the same platform, which
> Dylan and I found out when trying to exchange Linux save files.

Hmm. Well, that seems completely unnecessary, with a modicum of extra work
to the save/restore code. I wonder if there are any problems besides
big/little endian issues, which are easily solvable. Perhaps things like
shopkeeper names are the issue?

--
Benjamin Lewis

Evelyn the dog, having undergone further modification, pondered the
significance of short-person behavior in pedal-depressed panchromatic
resonance and other highly ambient domains... "Arf", she said.
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Quoting Dave Hammond <davidjhammond@hotmail.com>:
>The buck has been passed to Mark who will take the game from 1001-2000.
>I will ask Mark to post here with an update!
>Who's next?

I infer you are using Windows from your implicit assumption that everyone
uses what you do, but perhaps you should mention it explicitly?
--
David Damerell <damerell@chiark.greenend.org.uk> Distortion Field!
Today is Wednesday, April.
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Benjamin Lewis <bclewis@cs.sfu.ca> wrote:
>Are saved game files not platform independent? If so, why not?

They are not; they are written out in the host's native numeric format.

You would do best to ask the DevTeam the reasons why.
--
Martin Read - my opinions are my own. share them if you wish.
My roguelike games page (including my BSD-licenced roguelike) can be found at:
http://www.chiark.greenend.org.uk/~mpread/roguelikes.html
Good riddance to the Pope.
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Benjamin Lewis wrote:

> Are saved game files not platform independent?

Not at all. Not even on different versions of the same platform, which
Dylan and I found out when trying to exchange Linux save files.

Raisse, killed by a system shock
--
irina@valdyas.org LegoHack: http://www.valdyas.org/irina/nethack/
Status of Raisse (piously neutral): Level 8 HP 63(67) AC -3, fast.
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Quoting Benjamin Lewis <bclewis@cs.sfu.ca>:
[Savefiles not intercompatible.]
>Hmm. Well, that seems completely unnecessary, with a modicum of extra work
>to the save/restore code.

Many people have said this. No-one seems to have done it. Go figure.
--
David Damerell <damerell@chiark.greenend.org.uk> flcl?
Today is Thursday, April.
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

On Wed, 13 Apr 2005, Benjamin Lewis wrote:

> Raisse the Thaumaturge wrote:
>> Benjamin Lewis wrote:
>>
>>> Are saved game files not platform independent?
>> Not at all. Not even on different versions of the same platform, which
>> Dylan and I found out when trying to exchange Linux save files.
> Hmm. Well, that seems completely unnecessary, with a modicum of extra work
> to the save/restore code. I wonder if there are any problems besides
> big/little endian issues, which are easily solvable. Perhaps things like
> shopkeeper names are the issue?

All the game structures are saved as a dump of binary data. Even the
slightest difference in their internal representation will make the
savefile non-portable.

I'm afraid that ensuring such portability would take quite a lot of work.

--
,-=========================================-. Rule #35: That which
| 1KB | does not kill you has
`-------------------------------------------' made a tactical error.
- The Seven Habits of Highly Effective Pirates
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Adam Borowski wrote:

> On Wed, 13 Apr 2005, Benjamin Lewis wrote:
>
>> Raisse the Thaumaturge wrote:
>>> Benjamin Lewis wrote:
>>>
>>>> Are saved game files not platform independent?
>>> Not at all. Not even on different versions of the same platform, which
>>> Dylan and I found out when trying to exchange Linux save files.
>> Hmm. Well, that seems completely unnecessary, with a modicum of extra
>> work to the save/restore code. I wonder if there are any problems
>> besides big/little endian issues, which are easily solvable. Perhaps
>> things like shopkeeper names are the issue?
>
> All the game structures are saved as a dump of binary data. Even the
> slightest difference in their internal representation will make the
> savefile non-portable.
>
> I'm afraid that ensuring such portability would take quite a lot of work.

Can you give an example of what might be different across platforms besides
endian-ness? Does the NetHack source use variables that are different
sizes depending on platform (e.g. using int rather than defining something
like uint32, say)?

Anyway, I'll have a look at the code if I get some spare time.

--
Benjamin Lewis

Evelyn the dog, having undergone further modification, pondered the
significance of short-person behavior in pedal-depressed panchromatic
resonance and other highly ambient domains... "Arf", she said.
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Benjamin Lewis wrote:
> Adam Borowski wrote:
>
> > On Wed, 13 Apr 2005, Benjamin Lewis wrote:
> > All the game structures are saved as a dump of binary data. Even
the
> > slightest difference in their internal representation will make the

> > savefile non-portable.
> >
> > I'm afraid that ensuring such portability would take quite a lot of
work.

>From what I know of coding (which isn't much mind), dumping it into
binary would make it load faster. It also makes it harder to go in and
modify your charater than if they stored it in ASCII and made it
portable. I'm sure someone could make a decompiler, then recompile it
into another systems binary....
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

David Damerell <damerell@chiark.greenend.org.uk> writes:
> Quoting Benjamin Lewis <bclewis@cs.sfu.ca>:
> [Savefiles not intercompatible.]
> >Hmm. Well, that seems completely unnecessary, with a modicum of extra work
> >to the save/restore code.

> Many people have said this. No-one seems to have done it. Go figure.

Actually, last summer Topi Linkala began some work to convert bones files
to a platform-independent format and back with external utility programs,
and after that the next step could have been converting save files..
But last I heard of his project, it was still far from working properly.
Looks like he hasn't had time to finish what he started.

--
Jukka Lahtinen
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Benjamin Lewis <bclewis@cs.sfu.ca> wrote:

> Adam Borowski wrote:
>
> > All the game structures are saved as a dump of binary data. Even the
> > slightest difference in their internal representation will make the
> > savefile non-portable.
> >
> > I'm afraid that ensuring such portability would take quite a lot of work.
>
> Can you give an example of what might be different across platforms besides
> endian-ness?

Padding bytes. And no, you're not guaranteed an option to turn those off
on every platform you want to port NetHack to.

Richard
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

rlb@hoekstra-uitgeverij.nl (Richard Bos) writes:
> > Can you give an example of what might be different across platforms besides
> > endian-ness?
>
> Padding bytes. And no, you're not guaranteed an option to turn those off
> on every platform you want to port NetHack to.

Which is why it will be necessary to invent XML, XDR or plain old flat
ASCII... only someone already has.

(And no, these things existing doesn't make portable save files easy,
just easier. XML for saves, levels and bones would bring coolness in
ways beyond just portability, though.)
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Douglas Henke wrote:

> rlb@hoekstra-uitgeverij.nl (Richard Bos) writes:
>>> Can you give an example of what might be different across platforms
>>> besides endian-ness?
>>
>> Padding bytes. And no, you're not guaranteed an option to turn those off
>> on every platform you want to port NetHack to.
>
> Which is why it will be necessary to invent XML, XDR or plain old flat
> ASCII... only someone already has.

Or you could just write out all the fields you wanted as a stream of
bytes -- that's the solution I was thinking of.

--
Benjamin Lewis

Evelyn the dog, having undergone further modification, pondered the
significance of short-person behavior in pedal-depressed panchromatic
resonance and other highly ambient domains... "Arf", she said.
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Douglas Henke <henke@kharendaen.dyndns.org> wrote:

> rlb@hoekstra-uitgeverij.nl (Richard Bos) writes:
> > > Can you give an example of what might be different across platforms besides
> > > endian-ness?
> >
> > Padding bytes. And no, you're not guaranteed an option to turn those off
> > on every platform you want to port NetHack to.
>
> Which is why it will be necessary to invent XML, XDR or plain old flat
> ASCII... only someone already has.

Nah... they're human-readable, and therefore too easily modifiable with
just a text editor. Even if you don't care about malicious hackery, it
makes it too easy to rummage around in the file, just out of curiosity,
and accidentally damage it. Besides. it would make the save files
several times larger, and they're large enough as it is for some of the
smaller platforms NetHack runs on.
A fixed binary file format, in which the values of the game data have
been marshalled into an explicit number of bytes in an explicit order,
is just as simple as a textual format - in some ways easier - and much
safer and smaller.
But neither of these options is as simple as a direct dump of the
program's internal memory, and apparently there isn't so much demand for
portable save files that someone has been willing to change a couple of
simple fwrite() calls into some relatively extensive byte manipulation.
After all, most of us play our games on our own, on a single computer,
and don't share them around.

Richard
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Richard Bos wrote:
> Douglas Henke <henke@kharendaen.dyndns.org> wrote:
>
>>rlb@hoekstra-uitgeverij.nl (Richard Bos) writes:
>>
>>>>Can you give an example of what might be different across platforms besides
>>>>endian-ness?
>>>
>>>Padding bytes. And no, you're not guaranteed an option to turn those off
>>>on every platform you want to port NetHack to.
>>
>>Which is why it will be necessary to invent XML, XDR or plain old flat
>>ASCII... only someone already has.
>
> Nah... they're human-readable, and therefore too easily modifiable with
> just a text editor. Even if you don't care about malicious hackery, it
> makes it too easy to rummage around in the file, just out of curiosity,
> and accidentally damage it. Besides. it would make the save files
> several times larger, and they're large enough as it is for some of the
> smaller platforms NetHack runs on.

You could compress the XML, solving both problems you cite (you have to
want to edit it, and XML compresses really well). The whole process of
creating and parsing the XML, and compressing and decompressing, would
probably be too costly for Nethack as it currently stands, though, both
in cpu and in code. Still, platform-indepentent save files (and bones)
would be nice.

Erik
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

rlb@hoekstra-uitgeverij.nl (Richard Bos) writes:
> Douglas Henke <henke@kharendaen.dyndns.org> wrote:
> > [ cross-platform portable saves, levels and bones ]
> > Which is why it will be necessary to invent XML, XDR or plain old flat
> > ASCII... only someone already has.
>
> Nah... they're human-readable, and therefore too easily modifiable with
> just a text editor.

This is why, in the year 2005, we have things like digital signatures
and operating systems with proper file permissions. If you want to stop
people messing with the save files, then stop them doing that -- don't
just make it obscure enough that you think you'll scare most of them off.

> It would make the save files several times larger, and they're large
> enough as it is for some of the smaller platforms NetHack runs on.

That may be a legitimate concern, though I question whether anyone plays
much these days on any platform more cramped than, say, Wince on PocketPC.

> A fixed binary file format, in which the values of the game data have
> been marshalled into an explicit number of bytes in an explicit order,
> is just as simple as a textual format - in some ways easier - and much
> safer and smaller.

1) The binary format used by the game today forces neither a specific
number of bytes nor a specific order. Word length, byte order
within multibyte integers, and padding of in-memory structures
(dumped directly to file) are at the whim of the compiler and
options chosen for a particular build.

2) Development and debugging are _much_ easier if you have human-readable
files, and can leverage libraries other people have already written
and tested (e.g. for parsing XML).

3) If the size of files is a big deal, you know where bzip2 is.
 
G

Guest

Guest
Archived from groups: rec.games.roguelike.nethack (More info?)

Douglas Henke wrote:

> rlb@hoekstra-uitgeverij.nl (Richard Bos) writes:
>> A fixed binary file format, in which the values of the game data have
>> been marshalled into an explicit number of bytes in an explicit order,
>> is just as simple as a textual format - in some ways easier - and much
>> safer and smaller.
>
> 1) The binary format used by the game today forces neither a specific
> number of bytes nor a specific order. Word length, byte order
> within multibyte integers, and padding of in-memory structures
> (dumped directly to file) are at the whim of the compiler and
> options chosen for a particular build.

None of these pose a problem for a fixed binary file format, including the
fact that the length is variable. The only merits to a text format over a
binary format are those of human-readable-ness.

> 2) Development and debugging are _much_ easier if you have human-readable
> files, and can leverage libraries other people have already written
> and tested (e.g. for parsing XML).

These are libraries which are not required to begin with for a binary
format, and the code for a raw binary format would be simple enough not to
require another library.

I believe you are also greatly overstating the greater ease of development
and debugging considering the amount of debugging of the save format that
would have to be done compared to the amount of debugging for the rest of
the project.

--
Benjamin Lewis

Evelyn the dog, having undergone further modification, pondered the
significance of short-person behavior in pedal-depressed panchromatic
resonance and other highly ambient domains... "Arf", she said.