response 23 of 255:

May 29 05:20 UTC 2000

re #21: I don't know, Marcus. Why are you so adversarial? Did the reference
        the an account break-in raise your hackles?

re #22: You will maintain the status quo, not for any good reason, but
        out of apathy? Have you considered a run for the M-Net Board? ;)

response 25 of 255:

May 29 07:08 UTC 2000

Actually, I was asking why Larson was using words like "defense" and
"nonsensical" in #20.  He seems to think this is a trial and that there
are two sides and one side has to win and the other side has to lose,
and there seems to be some sort of confusion here that only the "right"
side is logical, and that there are no other logical viewpoints.

Last time I counted, I saw rather more than two sides to this argument.
Most of the sides were using logic in various different ways, and were
arguing from several different sets of initial premises and with various
different goals in mind.  It is quite possible for several groups,
starting from different initial premises, to each use logic equally
flawlessly to prove different and incompatible conclusions.  This
happens all the time in society, the "abortion rights" debate is perhaps
one classical example of this.  Also the inability of a particular group
to use logic "flawlessly" does not necessarily mean that group is wrong,
it may just mean they need to hire a better lawyer.  :-) One final thing
to keep in mind here is not all situations have one right answer with
all the others being wrong.  In some situations, there is no right
answer - even the best answer may be slightly wrong, and it may only be
possible to decide one answer is less wrong than another answer.  Even
this may be an optimistic viewpoint - in the field of music, is it
really fair to argue whether Beethoven is right and Bach wrong?  Or is
it the case that both are right, merely different, and that choosing one
over another is ultimately nothing more than a question of taste?

Sorry, no, I'm not upset at the reference to an account break-in; I
think so far we've been quite lucky in having so few nasty incidents.  I
know our luck is not do to any lack of persistance on the part of the
bad guys, so it is in fact usually me that is reminding other people of
the possibility of a break-in.  Indeed, I was nearly ready yesterday to
use this as an argument to encourage 20 people to consider moving nearly
a megabyte each of saved personal mail off of grex and onto some more
secure personal medium.  So, no, I'm not likely to be upset, in fact,
I'm quite willing to carry the argument a bit further.  People above
were arguing that the lack of such incidents was some sort of proof no
provision need be made for such a possibility.  This isn't necessarily
true.  It's also possible to argue that the fact we have such a
provision is in itself a deterrent from someone trying to pull a
bone-headed stunt like this.  The log provides both an audit trail of
who did the censoring, and provides the possibility of undoing such
censoring if it proves to be fradulent.  If it's redirected to
/dev/null, this trail is lost - it becomes possible for people to start
blaming bugs in the system, or an evil "they" that either stole their
account or otherwise broke things.

There is actually already a sort of analogous situation - PicoSpan's
design assumes that fair-witnesses are more trust-worthy, so even though
"censor" records responses that are deleted, "kill" makes no attempt to
save items that are killed.  Kill is a final thing - *poof*, the item is
gone.  I believe on grex we've already had at least one case where a
fair-witness wasn't happy with the way things were going and killed
every item they could.  Since we did backups even less often at the
time, and the conference was fairly active, this really meant those
things were gone, without possibility of recovery.  Is this a good or a
bad thing?  I think different people would argue differently, but I
think it is safe to say this would not have happened if item deletion
weren't as final as it in fact was.

response 27 of 255:

May 29 10:27 UTC 2000

If the scribble log is "so boring", why do people read it?

response 29 of 255:

May 29 14:26 UTC 2000

Well, I think the problem is that something like "chmod 600 /bbs/censored"
*should* be policy debate.

response 31 of 255:

May 29 17:02 UTC 2000

I'd prefer if this motion said something more like:

  Shall the picospan "scribble" and backtalk "erase" commands
  permanently make the text of responses inaccessible to non-staff
  users.

This allows (1) backup tapes, and (2) a censored log file that is
readable only by staff.

Maintaining a non-publically readable log is useful in many cases.  For
example, if user A erases his response, and user B responds by saying
"how dare you call me a cuttlefish!" then it is possible for a staff
person with the permission of user A to verify that actually user A was
just horribly embarrassed by a spelling error and said nothing of the
sort.

Or if you wander away from your computer, and someone else comes by and
starts erasing all your responses, then you can ask staff to restore
them.

Once on HVCN and twice on Starting Grid, users have suddenly found all
their responses in a conference erased.  This may be some kind of bug in
Backtalk, though the log files made it look like it was done manually
(with minutes of delay between each erasure as if a human were going
through doing it manually).  In all cases, we were able to restore all
the accidentally erased content from the log files.

In any case, log files are good to have.  I would strongly prefer that
the motion be worded in such a way to allow non-public logs of erased
responses to be maintained.

response 33 of 255:

May 29 17:54 UTC 2000

I am very strongly in the camp supporting responsibility by not allowing
ex post facto permanent removal of posted responses, however, I am also
sensitive to the concerns spelled out by Joe Saul. 

The best solution I can come up with is a variation on hhsrat's delayed
posting suggestion: 

Would it be possible to make adjustments to picospan and backtalk that
would allow users to censor their own responses (save to a staff-only
accessible logfile and remove from bbs) for some period less than one week
after intially posted, beyond which time the hide/expurgate command would
still be allowed but erase/scribble would not? 

I think this could be a fairly straightforward modification, by adding a
subroutine to the commands which checks the date field of the response
against the current date and then either proceeds or exits.



This would allow posts to be reconsidered and removed within a reasonably
short period of time, before potential damage amasses, and also would
prohibit removal beyond some arbitrary point at which it could be assumed
that either the discussion has too deeply incorporated the response in
question or that it is simply no longer an issue and is merely part of the
record.

(I have written such date-checking logic, which is why I think it would be
a reasonable technical solution, though not in any language applicable in the
Grex environment.)

Thoughts?

response 35 of 255:

May 29 20:00 UTC 2000

Re resp:20 - I suppose I should be flattered to have my "4 instead
of 5" comment elevated to being the sum total of Grex's "defense",
but I was only pointing out a typo in one of md's responses. Sheesh.

response 37 of 255:

May 29 22:12 UTC 2000

Since it is possible that retracted text might conceivably be required as
evidence which could be subject to subpoena, complete removal is unwise. 
However, permitting a user to hide his remarks from all but staff gives Grex
a defense of its own that it took all reasonable steps to mitigate the
situation by not preventing the user from lessening the damage by removing
it from view.

Jan's suggestion parallels my own thoughts.

response 39 of 255:

May 29 22:27 UTC 2000

re #25: Is there some reason you are unable to address the points I raised?
        Perhaps you need a better lawyer?

re #35: I think perhaps you misunderstood md's post, then. I read him as
        proposing two different solutions he would consider to rate "5"
        on his scale.

re #37: "I subpoena your records on this issue." "Sorry. Our records were
        destroyed, in accord with a long-standing policy, in the ordinary
        course of business." It's not a problem.

response 41 of 255:

May 30 00:54 UTC 2000

If you say something, you should have to stand by the consequences.  My
opinion.  Scribble and any hiding of information is dumb, in my opinion.

response 43 of 255:

May 30 02:04 UTC 2000

How much illegal content is present, and/or eliminated now?

We are clutching at straws.  

We either provide an open bbs, or we close down to the wills of those
persecuting us (remember our legal battle?)

response 45 of 255:

May 30 02:42 UTC 2000

I like janc's suggestion in response #31, and his and others' associated
reasoning.  So I'm going to modify my motion accordingly:

Shall the picospan "scribble" and backtalk "erase" commands
permanently make the text of responses inaccessible to non-staff users?