In article <[EMAIL PROTECTED]>, Uri Guttman <[EMAIL PROTECTED]> wrote:

> but perl was the number one example of OS used there which should be
> trumpeted to the world IMO. any manglement that says no to OS should be
> made to read that article (or be sent quotes from it).

[http://www.egovos.org/pdf/dodfoss.pdf]

i feel uneasy about using poorly developed statistics.  if python had come
out ahead we would be villifying the report instead, i think.

i don't think we can say anything about Perl from this evidence.  the broad
conclusions about open source are valid, i think, but beyond that is really
poor research methods.

the number of Perl projects is mostly meaningless.  notice that
"Perl CGI scripts" is in the list of approved software.  i think this means
that a single CGI script may have counted towards the total Perl total.
i think a web site using a Perl CGI script was given the same weight as 
the apache web server they ran on.  this inflates the importance of Perl
and deflates the importance of apache, which also has numbers hidden in 
PHP and so on.

in my opinion, if we use the result of research we know to be lacking to
support our position, we're virtually telling lies.  the broad conclusions
of the report get it right, i think, but beyond that we can't say much.
remember, this report assess the affect on DoD if FOSS went away, not
whether DoD should convert everything to FOSS, or any particular tool
in particular.

Reply via email to