Playing NerfHack on the Hardfought EU server my game crashed when
my rogue intended to descend the stairs (from dungeon level 3 to
mines) for the first time. (Maybe a bones level issue?) - Recover
of the game was possible. If the devs are interested in observing
the circumstances in the ttyrecs (if available on the server) the
time of crash was 2025-01-16 T 07:34 UTC.

Janis

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 20.01.2025 19:24, hackemslashem wrote:

On 1/16/25 8:43 AM, Janis Papanagnou wrote:

Playing NerfHack on the Hardfought EU server my game crashed when
my rogue intended to descend the stairs (from dungeon level 3 to
mines) for the first time. (Maybe a bones level issue?) - Recover
of the game was possible. If the devs are interested in observing
the circumstances in the ttyrecs (if available on the server) the
time of crash was 2025-01-16 T 07:34 UTC.

Janis

Thanks for the report - I believe I was able to fix this issue. It was
caused by cloning an item with a name, causing a conflict with two names sharing the same address.

On Hardfought (EU) with NerfHack 2.1.0
Dungeon level 7, descending the stairs

crash (again); recover was possible

(Happened at 2025-01-25 T 14:05 UTC)

Janis

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 25.01.2025 15:10, Janis Papanagnou wrote:

On 20.01.2025 19:24, hackemslashem wrote:

On 1/16/25 8:43 AM, Janis Papanagnou wrote:

Playing NerfHack on the Hardfought EU server my game crashed when
my rogue intended to descend the stairs (from dungeon level 3 to
mines) for the first time. (Maybe a bones level issue?) - Recover
of the game was possible. If the devs are interested in observing
the circumstances in the ttyrecs (if available on the server) the
time of crash was 2025-01-16 T 07:34 UTC.

Thanks for the report - I believe I was able to fix this issue. It was
caused by cloning an item with a name, causing a conflict with two names
sharing the same address.

On Hardfought (EU) with NerfHack 2.1.0
Dungeon level 7, descending the stairs

crash (again); recover was possible

(Happened at 2025-01-25 T 14:05 UTC)

I'm not quite sure, but it seems my pet got lost during that process.
(May the crash have something to do with pets during level-changes?)

Janis

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 25.01.2025 15:14, Janis Papanagnou wrote:

On 25.01.2025 15:10, Janis Papanagnou wrote:

On 20.01.2025 19:24, hackemslashem wrote:

On 1/16/25 8:43 AM, Janis Papanagnou wrote:

Playing NerfHack on the Hardfought EU server my game crashed when
my rogue intended to descend the stairs (from dungeon level 3 to
mines) for the first time. (Maybe a bones level issue?) - Recover
of the game was possible. If the devs are interested in observing
the circumstances in the ttyrecs (if available on the server) the
time of crash was 2025-01-16 T 07:34 UTC.

Thanks for the report - I believe I was able to fix this issue. It was
caused by cloning an item with a name, causing a conflict with two names >>> sharing the same address.

On Hardfought (EU) with NerfHack 2.1.0
Dungeon level 7, descending the stairs

crash (again); recover was possible

(Happened at 2025-01-25 T 14:05 UTC)

I'm not quite sure, but it seems my pet got lost during that process.
(May the crash have something to do with pets during level-changes?)

Wanted to return to the upper levels; climb up the stairs

crash (again); but this time *no* recovery possible!

Janis

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 26.01.2025 15:14, hackemslashem wrote:

On 1/25/25 3:24 PM, Janis Papanagnou wrote:

[...]
Wanted to return to the upper levels; climb up the stairs

crash (again); but this time *no* recovery possible!

Very sorry for the inconvenience -

No worries! (It's just a game.) :-)

with a bug like this it may not be
possible for the fix to work until a new game is started.

I'm not sure what you are trying to say here.

Just FYI; this thread started with a post about _another_ game
where I already died. The new crashes were from a more recent
game; in _that_ game the climbing-down-the-stairs crash was
recoverable (as it was in that other earlier game), but then
that same game crashed again, but un-recoverable (upon trying
to take the upstairs).

If you run int o something like it again, please report it here https://github.com/elunna/NerfHack/issues and I'll try to fix it ASAP.

Thanks, but I cannot use github for that - it doesn't support
the browser on my primarily used computer system.

Thanks for the offer; (for me) quick fixes are not necessary.

Janis

PS: I had left the crashed (unrecoverable) save-file on the
server in case that it might help for the error tracking.
But to play a new game that file will get deleted. I might
start another game soon. (Just FYI.)

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)

On 26.01.2025 20:36, hackemslashem wrote:

On 1/26/25 3:35 PM, Janis Papanagnou wrote:

[...]

Ok I see now. Do you think it had something to do with the previous
cloning issue or something totally different? [...]

The "cloning" was your statement; I can neither confirm nor deny
your assumption (or analysis). What I tried was to mention factors
that I observed - the pet thing, for example - which, though, may
be completely irrelevant (or even wrong observations on my part).
(Inspecting the ttyrecs may also provide some information of that
sort, but I'm not too optimistic on that.)

All I can now (after the larger set of observed crashes) say is
that it has to do with level changes, all the rest is speculation
(for me). - I'd ask myself what happens when you change a level.

From other games (in other variants) I could imagine effects of
{dungeon, monster, object, artifact} generation, or a bones level
related bugs. - But now where the [second] crash happened while
climbing _up_ to an already known (and constructed) level there's
some of these assumptions more unlikely and possible others come
to mind; say problems with resurrection of monsters on the previous
level, and so on.

But as an experienced and knowing software engineer the crash may
also be an effect of an earlier coding issue (like uninitialized
or undefined pointer states) that could be completely unrelated,
but just gets visible when changing levels.

I think, when the crash happened, I saw (just for a short moment)
some information on the screen that appeared to me to look like a
stack trace (or similar); maybe the Hardfought servers maintain a
logfile of such messages that you can inspect for better hints.

Otherwise, doing code analysis and code reviews, or play-testing
with log files activated. :-)

Janis

--- SoupGate-Win32 v1.05
* Origin: fsxNet Usenet Gateway (21:1/5)