The 1980 posts ARE at Google! And Google's dates are unreliable!
Joe Bernstein
doesn't allow me to set followups. This post is going to the set of
groups on which I remember the previous thread about 1980 posts gone
missing having appeared, but I think followups are likely to belong in
alt.fan.dejanews alone.
Ever since the first-searchable date on the Advanced Group Search page
rolled over from May 11, 1981 to May 12, 1981, I've been wondering
just how trustworthy Google's dates are - it's not as though May 11,
1981 posts had suddenly stopped appearing in searches, after all. As
a corollary, I've been wondering whether the 1980 posts might really
be in the archive. My first attempt to check this stopped when I
found that net.v7bugs, one of the really old groups, had no posts
dated earlier than May 11, 1981.
Well, I should have looked harder. If you look at the
allegedly-oldest net.v7bugs post in "Original Format", it turns out to
bear a date of August 13, 1980.
(I'm still not sure how original the format actually is - I would have
bet money that as of August 13, 1980, the group was still named
NET.v7bugs, but that's not what Google's version of the header says.)
Nor is this all. In fact, the allegedly-oldest *four* posts in
net.v7bugs at Google are all, according to their Googlified headers,
older than their alleged dates.
<http://groups.google.com/groups?threadm=anews.Aduke.488&output=gplain>
Header says August 13, 1980
Presented as May 11, 1981
<http://groups.google.com/groups?selm=anews.Aucbvax.300&output=gplain>
Header says August 12, 1980
Presented as May 11, 1981 (and as a followup to the August 13, 1980
one!)
<http://groups.google.com/groups?selm=anews.Aduke.583&output=gplain>
Header says January 6, 1981
Presented as May 15, 1981
<http://groups.google.com/groups?selm=anews.Aduke.574&output=gplain>
Header says December 12, 1980
Presented as May 19, 1981
The next two, presented as May 21 and May 22, 1981, have headers to
match. But according to my records, Bruce Jones's A-News Archive held
a post to NET.bugs.v7 (an anachronistic name; NET.v7bugs was renamed
to net.v7bugs before it was renamed to net.bugs.v7) dating to May 18,
1980. As you can see, this post isn't among the above four. This
post is also a linchpin in my argument for mid-May, 1980 as the date
at which topical newsgroups were first created, so I'm a wee bit
worried by this.
However, have no fear. The *next* post up the list gives us:
<http://groups.google.com/groups?selm=anews.Aucsfcgl.112&output=gplain>
Header says August 29, 1980
Presented as May 22, 1981
So I am heartened to continue:
<http://groups.google.com/groups?selm=anews.Aucbcory.194&output=gplain>
Header says August 24, 1980 (and the post contains internal evidence
for this)
Presented as May 24, 1981
<http://groups.google.com/groups?selm=anews.Aucbvax.107&output=gplain>
Presented as May 25, 1981
Header says (drum roll please)
May 18, 1980
!
The previous poster said there were nine posts in the A-News Archive
that dated from 1980. We have here six, and I'm not even done with
net.v7bugs yet!
<http://groups.google.com/groups?selm=anews.Aucbvax.114&output=gplain>
Presented as May 25, 1981
Header says May 18, 1980
<http://groups.google.com/groups?selm=anews.Aucbvax.117&output=gplain>
Presented as May 25, 1981
Header says May 27, 1980
<http://groups.google.com/groups?selm=anews.Aucbvax.110&output=gplain>
Presented as May 25, 1981
Header says May 18, 1980
<http://groups.google.com/groups?selm=anews.Aucbvax.106&output=gplain>
Presented as May 26, 1981
Header says May 18, 1980
Just for kicks, it now occurs to me to try
<http://groups.google.com/groups?selm=anews.Aucbvax.100&output=gplain>
No luck. But I'm wondering now what a real post from May 1981 from
ucbvax would have as its Message-ID.
I thought I was home free when I found
<http://groups.google.com/groups?selm=anews.Aucbvax.425&output=gplain>
Presented as June 1, 1981
but nope:
Header says October 13, 1980!
So I now try searching by date; I mean not searching by date-limits,
but by typing "1980" into the main search field. Let's see. Alas,
that nets me mainly SF-LOVERS DIGESTS with a moderator whose computer
is backdating by a year, according to copious internal evidence. Oh,
well, we're back to fishing, but this is *far* more 1980 posts already
than either I or the previous poster expected, given that this is only
one of *five* groups in which I have record of pre-May 11, 1981 posts
in the A-News Archive.
I don't know if I have the fortitude to attempt this with the other
groups (fa.unix-wizards, net.general, net.micro, and net.news, all of
which lived a lot longer under those names than did net.v7bugs, you
see). But here's some more I see in net.v7bugs:
<http://groups.google.com/groups?selm=anews.Aduke.473&output=gplain>
July 10, 1980 vs. May 27, 1981
<http://groups.google.com/groups?selm=anews.Agrumpy.314&output=gplain>
August 24, 1980 vs. June 10, 1981
(I have thus far tested 20 posts, of which fourteen are older than May
11, 1981. Note that the cross-implication is that either traffic or
archive quality for May-June 1981 is much poorer than expected.)
I had planned to test the rest of the surviving posts, but not
tonight, and maybe not at all: the remaining two on the second page
of threads checked out as dating to the dates presented (one of these
is no big surprise, since it's a post by Henry Spencer at Toronto, who
*could* not have posted before May 11, 1981...), their authors are the
sources of the next several posts, and there is a notable shortage, on
the entire first page of threads, of posts from sites like duke, phs,
ucbvax, etc.
Regardless. I knew there were certainly 3 posts from 1980 in the
archive, and the previous poster knew there were 9. An hour's work
has now found me 13 in just one group. Who knows what more awaits?
On the downside, I *really* hope someone from Google can work out why
the dates are behaving so oddly. Until then, it may not be wise to
rely on the date of *any* posting you haven't checked original format
for, or it may: without knowing the reason...
Joe Bernstein
--
Joe Bernstein, writer and accounting clerk (as if you couldn't tell)
j...@sfbooks.com <http://turing.postilion.org/these-survive/>
under which URL the *other* archive of 1980 posts, which suddenly
looks a good deal smaller, will soon reside.