Jeremy Kitchen wrote:

On Tuesday 27 April 2004 05:29 pm, Rick Macdougall wrote:

That would be a fun script to write.  See what mail server *sent* to
your mail server based on the headers and get a sampling that way.

but that would be incredibly biased.

For instance, I'd say probably 70% of the mail servers that send mail to me run qmail, if not a higher percentage. In a windows shop, you might get 70% exchange, etc. That's not a 'representative sample'


As I stated, we are an ISP, not an end user. I think our stats will tend to reflect what the majority of ISP (hence majority of high volume mail servers) are using.

Still have to think of a good way to do it though.



Reply via email to