The bug 74 was simply closed - that was wrong.
I was killed today while sitting in the driver stand of an japanise
express train at full speed.
The server had a bit lag and the track was above a lava area with a
short ascending slope in front of that area.
PS: please refrain from silently closing bugs - it gives the
impression that bug reports are ignored and therefor useless; that
will lead sooner or later to forks of advtrains
I closed bug 74 because I had found and fixed the cause of this bug.
There, players were damaged by the train damage system itself. In this
issue, however, it appears that the lava was the cause of your death.
(which means that that is, although similar to #74, a different issue)
Sadly, there's nothing I can do about server lags. If the server fails
to update the entity position and the client supposes it is inside lava,
then I can't prevent the damage easily.
A solution would be to make the player immortal while in a train.
Support for this in the engine is however not merged yet, see
(PS: the reason why the close of #74 was so silent is that I can't close
a bug and write a comment in one single e-mail, like it used to work
some time ago. OK, I was just lazy and only wrote one e-mail.)
As advtrains now enforces a 0.2s globalstep for itself I think this is
a selfmade lag/delay problem. With a server setting of 0.1s the player
would be moved much faster away from the wrong place - and with lag the
situation is even getting worse.
I think that the decision to enforce a 0.2s globalstep for advtrains
shows here it's very bad side.