Changes in USENET archives (Was Re: Chain of strictly decreasing closed sets is countable)

From: Dave Rusin (rusin_at_vesuvius.math.niu.edu)
Date: 03/04/05 

Date: 4 Mar 2005 18:56:53 GMT

In article <1109957253.426095.261130@l41g2000cwc.googlegroups.com>,
Dave L. Renfro <renfr1dl@cmich.edu> wrote:

>If you're only interested in a mathematical reply,
>skip past the following rant.

I would like to chime in here because the rant is very important to me too,
and I think it affects anyone else who, like me, sees value in the
permanent accessibility of (a small portion of) old sci.math posts.

To recap a few of Dave's points, things are bad because of three things
to have hit USENET recently:

  1. Google Groups has a "New and Improved" interface which suffers
     several new deficiencies.
  2. Google is starting to lose articles from its archives.
  3. Math Forum has changed its interface too and old URLs no longer work.

The upshot is that in a very short time period we have gone from having
two sources of archival information to having none in some cases.

Here is my own example. I needed to find a 1999 article of mine, with
     Message-ID: <7cofdk$ib0$1@gannett.math.niu.edu>
I was unable to find it at google by typing some keywords into its
the search engine. I did succeed at Math Forum, finding it at
     http://www.mathforum.org/epigone/sci.stat.math/phimplilchol
so I'm not hallucinating when I claim it really existed.

Seeing Dave Renfro's comments, I went to find the article in the
"New" math forum but was unable to do so from the old URL.
In fact, although I had saved all the data you see here, I hadn't
made a note of what the article was about, so I couldn't today do
a keyword search either!

In fact, what I did today was to dig up a May 2002 file I had made
when I was bored one day: I saved the results of a search on my name
at google. THAT file I could grep for the message ID and so I had
the few words that show up in a google search. Those I searched in
Math Forum, so I now know the file still exists there, but under
a completely different URL:
     http://mathforum.org/kb/message.jspa?messageID=1525844

So this is what we have now:

    Google, which _used to have_ a copy of the article, on public display,
    but which now has no trace of the article in its groups archives
    (though a search of the _web_ through Google still finds the article
    at Math Forum -- pointing to the old URL, unfortunately, but thanks to
    the Google Cache I was able to read it)

    Math Forum, which has the article and a searchable index, but which
    offers no way to find the article given only its old URL.

Both of the archives now serve up old articles in the proportional font
that makes a mess of our beloved ASCII graphics. Yuk.

On top of all this, although Math Forum has thoroughly revamped their
interface, they didn't make the one change I have been asking them to
make for some time: they still don't turn off the "reply-to" feature
for threads that have been dormant for a decade or more!

Is there a point to my rant? I'm not sure. If you want "action items",
try these:

1. If you have old posts that you value, find them and make copies.
Recordable CDs are cheap. If you've got your own web site, you might
want to put things there; if not, save the CDs until you do have your
own site.

2. Keep an eye on these archives to stay abreast of changes, in case
you are in the habit of using them or encouraging others to do so.

3. I believe it to be true that the keepers of these archives read
their email and take requests and complaints to heart. If you notice
that something of yours has gone missing, or if you want to make
a suggestion, do so. One letter might get ignored, but then again,
maybe not. Multiple letters would be harder to ignore.

dave

Relevant Pages

  • Re: Changes in USENET archives (Was Re: Chain of strictly decreasing closed sets is countable)
    ... Dave Rusin wrote: ... Google is starting to lose articles from its archives. ... Math Forum has changed its interface too and old URLs no longer work. ...
    (sci.math)
  • Re: RAID 1
    ... >> repeated context that also clutters the archives, ... I suppose I should have given up on usenet ... he didn't have google to help sift through the cruft and probably ... It shows parts of your 100 most recent posts with links to the full text ...
    (comp.os.linux.networking)
  • Re: Google Groups Changes?
    ... >Is/was there some information provided that will return Google ... and obfuscated the ever-burgeoning public newsgroup archives ... post-election desperation--given that the liberals no longer ...
    (sci.astro.amateur)
  • Re: Google Groups Changes?
    ... >Is/was there some information provided that will return Google ... and obfuscated the ever-burgeoning public newsgroup archives ... post-election desperation--given that the liberals no longer ...
    (sci.physics)
  • Re: Google Groups Changes?
    ... >Is/was there some information provided that will return Google ... and obfuscated the ever-burgeoning public newsgroup archives ... post-election desperation--given that the liberals no longer ...
    (sci.astro)