[306] in Athena User Interface

home help back first fref pref prev next nref lref last post

Re: Current thinking on gZephyr

daemon@ATHENA.MIT.EDU (Richard Tibbetts)
Thu Jul 20 12:40:01 2000

Message-Id: <200007201639.MAA12874@hikari-no-ken.mit.edu>
To: Dan Winship <danw@helixcode.com>
cc: Richard Tibbetts <tibbetts@MIT.EDU>, aui@MIT.EDU, tibbetts@MIT.EDU
In-reply-to: Your message of "Thu, 20 Jul 2000 12:20:22 EDT."
             <200007201620.MAA01245@twelve-monkeys.helixcode.com> 
Date: Thu, 20 Jul 2000 12:39:47 -0400
From: Richard Tibbetts <tibbetts@MIT.EDU>

On 7/20 you wrote:
> > - Allowing people to log classes will likely create a political
> >   problem for Ops. For example, if someone asks that class
> >   seekrit-stuff be logged, how will Ops deal if the majority of
> >   subscribers of seekrit-stuff don't want it to be logged.
> 
> Ops's official position last I knew is that they won't log or acl
> classes for people other than IS because there's no way to know who
> "owns" a given class. This is not a solvable problem in zephyr.

Since we hold all the cards, it is an eminently solvable problem. For
instance (each of these is a possible solution):

1)Make every moira list into an acl'd and (if requested) logged class
  with the classname "moira-<listname>" and then hack up "appropriate"
  client support.

2)Collect statistic for a while on what classes are used, and
  blacklist those from every being logged or acl'd.

3)Make every moira list into an acl'd and (if requested) logged class
  with the classname "<listname>" and deal with the annoyance.

4)Just start logging and acl'ing classes and deal with the shitstorm.

I am particial to answers 1 or 3. I bet Ops would prefer 1 to 3.
Personally I think that we should do 3 and just grow a spine.

tibbetts
Who still thinks this is not part of AUI's problem

-*- http://www.mit.edu/~tibbetts -*- finger tibbetts@monk.mit.edu -*-

home help back first fref pref prev next nref lref last post