Re: do_set check

From: Daniel A. Koepke (dkoepke@circlemud.org)
Date: 07/12/01


On Wed, 11 Jul 2001, George Greer wrote:

> The command worked...for 5 minutes.

Five minutes or less; probably less.  Or is saved permanently.  There are
three possible (visible) outcomes for 'set file' on an online player, as
it stands:

  1) The command had a visible effect, for up to 5 minutes.
  2) The command had no visible effect (overwritten before observed).
  3) The command had a visible, lasting effect.

The first two are the obvious cases.  The third would happen if the set
character data was loaded by starting to login just before the unset,
online character quits.  The person logging in can then type the wrong
password, causing the set character to overwrite the now offline character
data.

In other words, "the command worked...for 5 minutes," isn't the only
observable outcome.  Actually, the outcome is mostly unpredictable and can
be influenced by seemingly unrelated activities by unrelated parties, that
can occur within a fairly arbitrary interval (however close you are to an
autosave).

> Moving from bad situation to bad situation isn't an improvement, ...

And keeping a bad situation, but presenting a warning about or an extra
step to evoke the braindead behavior is an improvement?  If assuming that
the immortal wants something set when they type 'set file' (and not some
random effect) is wrong, then there's only one legitimate reply: tell them
they can't do it and return.  Adding 'set force file' isn't necessary.  I
assume if the person can create a need for 'set file' to work on online
players (without comitting to memory), they can remove the three lines the
warning would constitute; otherwise, it's undesirable to be able to evoke
random behavior from a valid command.


-dak

--
   +---------------------------------------------------------------+
   | FAQ: http://qsilver.queensu.ca/~fletchra/Circle/list-faq.html |
   | Archives: http://post.queensu.ca/listserv/wwwarch/circle.html |
   +---------------------------------------------------------------+



This archive was generated by hypermail 2b30 : 12/06/01 PST