Newsgroups: comp.lang.smalltalk
Path: cantaloupe.srv.cs.cmu.edu!bb3.andrew.cmu.edu!newsfeed.pitt.edu!gatech!newsfeed.internetmci.com!news.kei.com!nntp.coast.net!torn!nott!cunews!tina.mrco.carleton.ca!knight
From: knight@acm.org (Alan Knight)
Subject: Re: 128 ... initial count
X-Nntp-Posting-Host: tina.mrco.carleton.ca
Message-ID: <knight.821285038@tina.mrco.carleton.ca>
Sender: news@cunews.carleton.ca (News Administrator)
Reply-To: knight@acm.org (Alan Knight)
Organization: The Object People
References: <30F3B0ED.33AB@Cherniak.ON.CA>
Date: Wed, 10 Jan 1996 14:43:58 GMT
Lines: 50

In <30F3B0ED.33AB@Cherniak.ON.CA> Dennis Smith <quenton@Cherniak.ON.CA> writes:

>I asked for mail from anyone reading / posting-to this group
>about a week ago.

>So far, 76 responses, plus at least 5 that I know of (including me)
>who have not responded.

>That kind of supports the 128 number so far.

Oh come on. The 128 number is completely ludicrous. In order to
demonstrate this, I actually did some work <Gasp>.

I write a best of comp.lang.smalltalk column for The Smalltalk Report.
This means I keep an archive of a large percentage of the postings to
this group. I took my archive starting from July 3, 1995
(approximately 6 months ago), ran it through
   grep "From:"
   sort
   uniq

That gives me a list of lines containing From:. Some of those are
garbage (e.g. Smalltalk code fragments containing from: as a keyword)
and some of them are duplicate names that don't happen to be exact
duplicates (e.g. I might show up as knight@acm.org,
knight@mrco.carleton.ca or alan@objectpeople.on.ca). I went through
the file and manually eliminated the obvious duplicates and garbage.

I get 651 distinct people who have POSTED to this newsgroup in the
last 6 months. Some of those are almost certainly duplicates. Let's
say one third, to be extremely liberal. Some of th remainder might be
cross-postings from people who don't actually read the group. Let's
say that's a third of what remains. That still leaves 290 people
posting. I expect that there are many people who read and don't post.
This also doesn't include people who post things that I didn't save
(e.g. simple questions). I expect that a search of a comprehensive
archive would turn up more posters.

I believe that in direct mailings they consider a fraction of a
percentage response good. I think it very unlikely that the same
technique applied to postings would give >50% response.

I can post the list of posters if anybody cares.

-- 
 Alan Knight                | The Object People
 knight@acm.org             | Smalltalk and OO Training and Consulting
 alan@objectpeople.on.ca    | 509-885 Meadowlands Dr.
 +1 613 225 8812            | Ottawa, Canada, K2C 3N2

