Subject: Re: automatic package statistics
To: None <mcmahill@mtl.mit.edu>
From: Tim Rightnour <root@garbled.net>
List: tech-pkg
Date: 10/14/1999 09:33:01
On 14-Oct-99 mcmahill@mtl.mit.edu wrote:
>> > Besides, just because someone installs a package, doesn't mean he
>> > likes it, or that he he uses it. These are all three orthogonal.

The statistics won't be perfect.. but what rough statistics do tell someone
like me is:

a) 90 people installed pkg a, and 4 people installed pkg b.   This means that
the features advertized in pkg a are more important to users than b.  Perhaps I
should update a, rather than b if I have limited time.

b) Working on the above.. If I see other pkgs, similar to a, vs b.. I should pkg
those first.

c) If nobody installs c, perhaps I should make sure it still builds.

d) If I tell author of linux software that "hey, 90 NetBSD users downloaded and
installed your software"  Perhaps that linux software writer will become more
inclined to integrate patches, and support our OS better in thier software.

I think looking for a currently in use count is going too far, and not going to
be effective.  Just getting a rough idea, would however be very nice.

---
Tim Rightnour <root@garbled.net>
NetBSD: Free multi-architecture OS http://www.netbsd.org/
NetBSD Mailing lists on the web: http://mail-index.netbsd.org/