Resets and repops

[claw at null.net]

On 22/03/97 at 12:19 AM, Adam Wiggins <nightfall at inficad.com> said: >
[ChrisL:]

>> Hehn.  I like coming up with ideas that break models.  The Crytalline
>> Tree is one we should probably revisit.

>Crystalline Tree?  Wazzat?

An old scenario, much like the Elven forest/Sceptre/Castle Krak
scenario we used to bounce ideas off.

The actual scenario was called "Bubba and the Crystalline Tree", and
dealt fairly simply with our friend Bubba who managed to climb a
crystalline tree.  This was no mean feat as Bubba is a very heavy chap
whose weight *should* break the tree, dropping bubba to the ground. 
The scenario is:

-- Bubba is in the Crystalline Tree.

-- Bubba has some sort of state changing effect which lessens his
weight such that the tree does not break.  Typical examples are: he is
holding on to a large number of helium balloons, he is enchanted to be
lighter, etc.


-- Bubba now changes his weight value.  This could be by letting go of
the balloons, picking up another object, receiving a "heavy" spell,
having his "light" spell wear off, having one of the objects he is
carrying receive or lose a heavy/light spell, having one of the
objects he is carrying chage its weight for any other reason etc.

-- Having the tree respond appropriately to Bubba's weight change.

>> >Another solution is to establish a convention that says that distance
>> >viewers should check for and call a specific action on locations they
>> >are viewing. That action could cause the viewers to see the missing
>> >NPC's, without actually having to create them. (Dirty trick, I know!)
>> 
>> Sounds like a neat one to me.  Forgery is fine, as long as you can
>> guarantee that the forgery is indistinguishable from the original.  eg
>> What happens when you summon a mobile that is a forgery?  If you don't
>> do summons, and I seem to recall you don't, try something like a
>> magical spell which can be delivered to a distant room (fireball? 
>> charm?  transmogrify?).

>Sounds like you need a layer of abstraction which interprets all "do
>something interesting with this object" calls (ie damage, heal,
>change form, anything even mildly active) which quickly loads up the
>real thing. And, at this point, are you really saving much time or
>space?

Nahh.  That's way too expensive.  Most of us already have ways of
trapping method calls for which we don't have a defined method (and
usually calling a default method instead).  All you need to do instead
is to instantiate the actual object upon receipt of such a call, and
continue working with the forgery the rest of the time.  

Cheap, effective, and workable.  Real objects only get built when
something asks for them and a forgery can't do the job.

>> I'm buried in acceptance testing of a market data delay server.  So
>> far we're up to 5,000 messages per seconds (whee!) which is way over
>> spec, but they just changed the environment on me and I can't keep the
>> thing up for more than 2 seconds before it goes into a frantic core
>> loop.  (Yes!  You too can generate 100 core files per second!  Yes! 
>> You too can eat disk space faster than you ever imagined!  Yes!  It
>> gets faster when you kill the synch daemon...)

>Sounds like fun.  I 'get' to work with Win95 + DirectX all day long.

Ahh, there's one of my saving graces.  I refuse to work with MS
products.  It keeps life a lot more enjoyable.

>After rebooting both my code machine and my debug machine a dozen
>times a day, it's somewhat refreshing to come home and type 'uptime'
>on my Linux machine and see 104 days.

Shoot.  I reboot OS/2 no more than once a fortnight or so, and then
usually because I've coded something stupid.

--
J C Lawrence                               Internet: claw at null.net
----------(*)                              Internet: coder at ibm.net
...Honourary Member of Clan McFud -- Teamer's Avenging Monolith...