Stage 1: FAQ maintenance ceases.
If the maintainer was posting by hand, it ceases appearing in the newsgroup;
if with the FAQ checker or some other automated approach, it keeps getting
posted but in most cases becomes out of date.
Response: news.answers moderators notice this stage when the routine archive
maintenance finds something not posted in over 90 days. If there's an
autoposting we may never notice until someone reading the newsgroup
complains. Once we notice, we ask the maintainer to clarify his/her
intentions. At this point the FAQ is either resurrected or advances either to
stage 2 or stage 3; the issue is when to go to stage 3.
Stage 2: FAQ delisted from LoPIP.
By this I mean the regularly-autogenerated List of Periodic Information
Postings that gets posted regularly to certain newsgroups. I don't actually
know who uses this list, but would be happy to be told it's widely read. To
the *.answers moderators, we do this by placing a mark in the database entry.
Stage 3: FAQ deleted from archives.
By this I mean the directories whose names are derived from the archive name,
ftp://rtfm.mit.edu/pub/usenet/your/archive/name
This requires explicit action by the *.answers moderators to unlink the files
from all the directories in which they occur (all other parts of the 'archive'
that correspond to specific newsgroups and can be spammed are flushed after 90
days). This results also in it disappearing from the mirrors, like
www.ftp.org.
In the past Stage 2 and Stage 3 were loosely linked -- when a FAQ was truly,
sincerely dead, it went through both stages in the same editing session.
However, sometimes we decide a 'dead' FAQ might still be useful to a potential
new maintainer, so we keep it in the archives although marking it dead for the
LoPIP. For example, if the last posting was a bare 3 months ago, and the
maintainer's mail bounces, it may be worth hanging on for a while to see if
s/he picks it up again. Also, sometimes a moderator would just `forget stage
3. Thus there are likely, to this day, entries deleted from the LoPIP that
are still in the archives.
Stage 4: FAQ database entry moved to dead file.
This is really only relevant to *.answers moderators, because it's just a
convenience issue. As far as I can tell, it has been past practice to leave
dead FAQ entries in the FAQ database forever, marked as dead and not to be
listed. I recently started moving some of the database entries to a separate
'dead' list, just for efficiency (routine archive maintenance is VERY slow,
which is why it wasn't done for 18 months until I restarted it in July).
As far as I can tell, we don't ever completely delete entries from our
databases (I use the term loosely) which serve as our collective memory.
The issue:
There are 2 problems:
- Sometimes a posting doesn't quite enter stage 2 because we wrote to query
the maintainer but didn't notice we never got a response. Currently
noticing requires an active moderator going through our database looking for
certain comments and checking whether anything has been posted in the
meantime.
- Today I handled a posting that definitely required a stage 2/stage 3
separation. The maintainer doesn't want to post anymore because it's too
much trouble, but would do so "if there is demand". To me this means it's
really, truly in stage 2 and will remain there -- if he'd repost on demand,
why not make it even more available on demand from www.faqs.org and other
mirrors? But it's no longer a periodic posting.
Thus I finally think there needs to be a clearer Policy on when to go to stage
3, and since it's possibly of interest to current FAQ maintainers (all of whom
will eventually become former FAQ maintainers in the long term, since in the
long term we're all dead anyway), it seems appropriate to make suggestions and
ask for input instead of just announcing a policy or working something out
with the inactive *.answers moderators.
It seems to me the choices are:
For the first problem (limbo state - transition from 1 to 2)
A. Continue informally as now. It's no problem if postings never quite enter
stage 2.
B. Go to stage 2 immediatly on discovering an outdated FAQ. Write to the
maintainer and resurrect the FAQ if s/he replies quickly - after all, it
only REALLY disappears from the LoPIP later in the month when the
autogeneration happens.
I'd relly like to go with B unless it's objectionable.
For the second problem (transition from 2 to 3):
A. Continue informally as now. It's no problem if stage 3 happens years after
Stage 2, or immediately after stage 2.
B. Never go to stage 3. It doesn't cost anything to keep a few more files
around.
C. Develop specific policies for the following 3 separate situations
- When a maintainer says a FAQ is dead (currently: Do stage 2 and 3
together)
- When a maintainer won't post but still keeps the info around (currently:
never enter stage 3 unless a moderator happens to feel like re-querying
several months later.
- When a maintainer never responds but mail doesn't bounce (currently, we
likely don't notice)
- When queries to a maintainer bounce (currently: policy a little unclear.
When I resumed archive maintenance, I deleted anything that hadn't been
posted for a year for which the mail bounced, but if I keep up regular
maintenance I'll start getting such situations after a mere 3 months,
which seems too short for complete deletion)
I'd prefer C because I'm a detail-obsessed control freak :-) but can cope with
A or B.
Advice? questions? comments? I'm mostly looking for what matters to FAQ
maintainers; at the moment I'm not too concerned about workload implications
for *.answers moderators, although that will certainly become an issue later
in the discussion.
[
Usenet Hypertext FAQ Archive |
Search Mail Archive |
Authors |
Usenet
]
[
1993 |
1994 |
1995 |
1996 |
1997
]
© Copyright The Internet FAQ Consortium, 1997
All rights reserved