Mailing List Archive

Analog newsletter, February 2003
| TO UNSUBSCRIBE from this list:
| Usenet version: news://
| List archives:
Dear analog users,

Welcome to the list, if you've joined since the last mail. I've just
discovered that it's nearly nine months since I last wrote to you all,
so it's long past time that I updated you on what's going on with

In this mail:
1. Version 5.3
2. Next version
3. Geotargeting -- your opinions wanted!
4. News in brief
5. Donations news

1. Version 5.3
After a bit of a break from new versions, I released version 5.30 at the
end of November, and 5.31 at the beginning of January.

These versions contain some commonly-requested features. For example,
analog now reads gzipped logfiles natively, without the need for an
external program. Also host inclusions and exclusions can now use IP
address ranges (such as 131.111.20-21.*) and subnet masks (such as

Of course, you can upgrade at

2. Next version
Work is progressing fast on the next version of analog. I've completely
rewritten the output functions, and Per Jensen, Jeremy Wadsack and
myself are working on writing new output formats. This has been on a lot
of people's wishlists for a long time, but it was impossible without a
substantial rewrite, which I've now done. It's not quite ready for
public consumption yet, but I hope to have a beta up some time next

3. Geotargeting -- your opinions wanted!
A few people have been asking for geotargeting to be added to analog,
and I'd value your opinions on this.

For those who don't know, "geotargeting" refers to looking up numerical
internet addresses in a database stored on your machine, to try and work
out which country, and possibly city, they're from. The advantages are
that you don't have to do slow DNS lookups to find out where your
visitors came from, and that you can find out the countries of people
who came from .com addresses. But obviously it's only as good as the
database you're using.

The current state of play seems to be roughly this. There are some free
databases, which are mostly derived from "whois" lookups. They claim
accuracy of up to 90% on country-level lookups (which is actually not
great -- it means 1 in 10 lookups are located in the wrong country).
There are also some commercial databases which use proprietary
techniques. They claim accuracy of 99% on country-level lookups and
mid-90% on city-level lookups; but they cost hundreds or thousands of
dollars a year.

Enough preamble. What I want to know is:
1) Is there demand for integrating a free database into analog, even if
the accuracy isn't great?
2) Is there demand for integrating a commercial database into analog,
even if the cost is in the range indicated above?
3) Does anyone have any views on / experience of particular packages?

4. News in brief
This year has been pretty busy for analog so far. As well as version
5.31, here are some other news items since the turn of the year.

You can now read the analog-help and analog-announce mailing lists as
newsgroups, courtesy of Gmane. See the links at

Documentation for old versions of analog moved to a new location,
when SourceForge withdrew the facility to use CVS from CGI scripts.
Thanks to Ian Jackson for giving me space on this server.

Clockwatchers Hosting of Florida became the 100th company on the
"analog-friendly ISP" list
If any other web hosting companies who supply analog reports to their
clients want to be on the list, just mail me.

5. Donations news
First, thank you to those of you who have donated money over the last
year. I'm determined that analog should stay free (in both senses), but I
really appreciate the generosity of those who want to thank me financially.

You can now send donations through PayPal in multiple currencies. As
well as US dollars, you can now use pounds sterling, euros, Canadian
dollars and yen.

And if you haven't donated, please consider it. Even small donations
make me feel appreciated and give me motivation to work on analog!

Stephen Turner, Cambridge, UK
"The question of whether a computer can think is no more interesting than
the question of whether a submarine can swim." (Edsger W. Dijkstra)