A point of clarification: how you proceed in the negotiation process
in getting approval (whether by resubmitting your actual FAQ's or by
sending us email where you discuss the changes you wish to make)
shouldn't make make any systematic difference in how soon we get
around to dealing with it.
(We *DO* sometimes require that people post at least once rather than
send their FAQ's in email, so we can confirm that their local site's
news software isn't ridiculously broken. And our form letters ask by
default maintainers to resubmit the whole shebang because we know that
will get us all the info that we need to find out. The *.answers
submissions guidelines do explicitly describe whether you need to
resubmit your postings or just send us email describing changes when
you get to the point where your postings are already approved and you
make some changes to them and want to get reapproval.)
----------------
Now, as for the idea embodied by:
> it seems to me that it would be a lot more efficient for all involved
> if, once an FAQ wiggles its way to the top of the submission queue,
> further correspondence on the FAQ could be expedited so that the
> overall approval time for an FAQ could remain fairly low.
There's some trade-offs here. It would take some additional time to
try to pick out the entries in the queue which are known to be
resubmissions rather than new resubmissions; also, ultimately, the
reduced waiting time for resubmissions would have to be paid for by
increased waiting time for new submissions. This might be bad since
people with new submissions have no idea at all if their submissions
got to us or disappeared into the ether. (Our plans to install a
quick-and-dirty auto-reply to acknowledge receipt of submissions this
summer got sidetracked since the organization which donates the use of
rtfm.mit.edu hardware was to have upgraded at the end of the summer
[from the current DECstation 5000/25 to a SPARCstation 20], so the
auto-reply feature is waiting until the hardware cutover.)
Now, in practice, I will admit that sometimes when I see replies from
FAQ maintainers addressing some problem which I just recently sent
email to them about, I'll often yank them out of queue out of order.
----------------
> Meanwhile, the readers
> of the (non-*.answers) groups are wondering where this FAQ is,
> and I can't tell if I should post it to the groups or not, since
> I don't know whether it's about to be approved or not.
We often give out the advice below, but I just checked and realized
that it was not actually in the *.answers submissions document. I've
just added this text in the section describing how to submit FAQ
postings, after the paragraph apologizing for long delays.
| Note that because of the potential length of delays involved in
| getting your postings approved, you may wish to continue posting
| your FAQ in its home newsgroup(s) in the meantime.
That should help reduce the problem you describe.
----------------
> (Sheesh, I work in a government lab, and *they're* not even this
> inefficient and bureacratic!)
Hey! At least government employees get paid to impose their
bureacratic ways on everybody. :)
I think your complaints were legitimate ones. As for re-engineering
the process (to make us more efficient), there's some effort on our
part right now to further automate a lot of the background work
(example: we already make liberal use of form letters, but even
customizing the form letter we send describing the first cut of
changes necessary to make something conform takes a lot of time when
there are hundreds of submissions in the queue....)
Also, additional delays specifically attributable to the start of the
new academic term aside, we've taken a net loss of two moderators this
year and a couple of people we're hoping to bring up to speed are
still not up to speed. I recognize (Fred Brook's _The Man-Month Myth_
is sitting on my shelves) that some problems require more than just
throwing additional man-power at it, but the current state of
*.answers moderation is such that more man-power *IS* effective.
----------------
Random statistics: here's a count tallied by month of the number of
*.answers queue entries that the moderators dealt with that month.
941 February-1994
1020 March-1994
671 April-1994
822 May-1994
902 June-1994
703 July-1994
963 August-1994
== Ping Huang
[
Usenet Hypertext FAQ Archive |
Search Mail Archive |
Authors |
Usenet
]
[
1993 |
1994 |
1995 |
1996 |
1997
]
© Copyright The Landfield Group, 1997
All rights reserved