BSDstats is an attempt to produce reasonably accurate statistics on
deployments of BSD operating systems. This is useful for marketing,
project advocacy, hardware vendors and purchasers, device driver
maintainers, and port maintainers.
The major problem that we are trying to address is vendors who do not
feel that BSD operating systems present enough of a market to bother
supporting.
Meant to be run monthly or during bootup, this script submits
statistics anonymously to a central server (http://bsdstats.org).
Potentially sensitive details like IP addresses and hostnames are not
stored by or relayed through the BSDstats server.
===> The following configuration options are available for bsdstats-7.0_3:
TOR=off: Submit securely through TOR anonymity network
===> Use 'make config' to modify these settings
sysutils/bsdstats: mark as BROKEN.
The mastersite simply returns a PHP fatal error.
Noticed via maintainer-timeout in reply to PR.
PR: ` 251504
Hat: bugmeister
Remove WWW entries moved into port Makefiles
Commit b7f05445c00f has added WWW entries to port Makefiles based on
WWW: lines in pkg-descr files.
This commit removes the WWW: lines of moved-over URLs from these
pkg-descr files.
Approved by: portmgr (tcberner)
Add WWW entries to port Makefiles
It has been common practice to have one or more URLs at the end of the
ports' pkg-descr files, one per line and prefixed with "WWW:". These
URLs should point at a project website or other relevant resources.
Access to these URLs required processing of the pkg-descr files, and
they have often become stale over time. If more than one such URL was
present in a pkg-descr file, only the first one was tarnsfered into
the port INDEX, but for many ports only the last line did contain the
port specific URL to further information.
There have been several proposals to make a project URL available as
a macro in the ports' Makefiles, over time.
(Only the first 15 lines of the commit message are shown above )
sysutils/bsdstats: Set timeout to wait response from server
Currently 300.statistics hangs up if server accepts connection but
doesn't respond to the request. So set timeout to wait response from
server.
PR: 265389
Approved by: maintainer timeout
MFH: 2022Q3
sysutils/bsdstats: make reporting opt-in
- disables reporting by default
- does not use the pkg-install script to "modify the currently running
system" as per the FreeBSD Porters Handbook Chapter 9.2.
- does not output noise if monthly_statistics_* variables aren't enabled
- add a post-upgrade message informing the user it's now opt-in and may
need enabled
sysutils/bsdstats: Fix occurence of /usr/local
Change an occurence of /usr/local that I inadvertently introduced with
r533478 into %%PREFIX%%.
Approved by: gerald (mentor)
sysutils/bsdstats: Update to 7.0
- Use pkg query %n:%v:%o instead of pkg info -o "*" to get list of
installed packages. This simplifies processing and accuracy of split for
version numbers on the report server by not having to guess where that
split is.
- Check for existence of %%PREFIX%%/etc/bsdstats.conf and if it exists,
restrict ports submitted to those listed in there.
- Add a note to pkg-message.in about %%PREFIX%%/etc/bsdstats.conf
- Get rid of the LIMITED config option: LIMITED is enabled if and only if
%%PREFIX%%/etc/bsdstats.conf exists.
While here, also modernize the port using option helpers.
Submitted by: scrappy
Approved by: gerald (mentor)
Differential Revision: https://reviews.freebsd.org/D24644
Update sysutils/bsdstats to 6.2
- Fix files/300.statistics.in so that it can work with the new output
format of pciconf -l in 13.0-CURRENT.
- Fix a bug in device reporting.
- Add support for ports reporting from MidnightBSD systems, whose ports
tree is based on FreeBSD ports tree.
- Changes to using /usr/bin/nc on DragonFly BSD, whose ports tree is based
on FreeBSD ports tree.
- Assume maintainership of sysutils/bsdstats. [1]
PR: 229347
Submitted by: scrappy
Approved by: yuri (maintainer) [1], gerald (mentor)
- Use anticongestion instead of simply sleeping.
- Bump PORTREVISION
svn r316342 added a common anticongestion routine that all periodic scripts
should
use instead of simply calling sleep. To test if it's available, a port should
check for the anticongestion_sleeptime variable.
PR: 218283
Submitted by: asomers
Approved by: Yuri Victorovich <yuri@rawbw.com> (maintainer)
- Add DIG command to SUB_LIST, not overwrite the list
- Use correct OSVERSION for the dig(1) / drill(1) change
- Don't add anything to rc.conf and periodic.conf when WITH_PACKAGE_BUILDING is
set to pet poudriere and tinderbox
- Re-word pkg-message
- Bump PORTREVISION
PR: ports/175879
Submitted by: Alex Keda <admin@lissyara.su>
Approved by: maintainer
Reported by: Alex Keda, 2A, and dvl
Patched by: A.J. Kehoe and me
To prevent fallout of lowering __FreeBSD_version in releng/10.0 branch,
adjust OSVERSION evaluation in ports that specifically use '100050N'.
Approved by: affected maintainers (implicit)
Sponsored by: The FreeBSD Foundation
In the rc.d scripts, change assignments to rcvar to use the
literal name_enable wherever possible, and ${name}_enable
when it's not, to prepare for the demise of set_rcvar().
In cases where I had to hand-edit unusual instances also
modify formatting slightly to be more uniform (and in
some cases, correct). This includes adding some $FreeBSD$
tags, and most importantly moving rcvar= to right after
name= so it's clear that one is derived from the other.
- Forced commit to note that some files were repocopied:
- files/300.statistics -> files/300.statistics.in
- files/bsdstats -> files/bsdstats.in
- pkg-message -> files/pkg-message.in
Repocopy by: marcus
- Ensure executable permissions on periodic script in case files subdirectory
get permissions stripped
PR: ports/147255
Submitted by: Chris Rees <utisoft@gmail.com>
Remove a reference to hostname that was in the script, but was never
actually used anywhere ...
PR: ports/106298
Submitted by: Andre Albsmeier <Andre.Albsmeier@siemens.com>
The point of bsdstats is to be as painless as possible, but also to be
run monthly ... change the behaviour such that when it is run with the
-nodelay option, it *only* reports the operating system, even if both
devices and ports reporting are also enabled ...
... this when, when run as part of the original port install, *or*
on system reboot if enabled in /etc/rc.conf, it is 'quick n painless',
and we get the critical number we are looking for ...
As per discussion on -advocacy, resurrect having the installer prompted
if they want to enable it in /etc/rc.conf, instead of making it automatic.
Put a message explaining that for desktop/laptop environments, enabling in
/etc/rc.conf is recommended ...
By popular request, I've added a bsdstats.sh script to etc/rc.d so that
it can be enabled to run on reboot (ie. for laptops, office computers, etc)
since not everyone runs a server 24/7 ...
I've added to pkg-install a yesno question as to whether it should be enabled
on reboot, seperate from the monthly periodic run ...
Only depend on netcat if we are running an older version of FreeBSD
that doesn't already include it ...
PR: ports/106446
Submitted by: KIMURA Yasuhiro <yasu@utahime.org>
v5.0 of the BSDstats script ...
This version brings in optional port reporting.
This version is backwards compatible with the older versions, so an
upgrade isn't required, except if you wish to enable port reporting
Make pkg-install a bit smarter ... if the questions have already been
answered in /etc/periodic.conf, don't ask them again, just print
out the message ...
Better for unattended port upgrades ...
Fix PATH so that it includes /usr/local/bin, so that dig is found on
systems where its installed via ports
Log communications, by default, to /var/log/bsdstats, so that one
knows if things are successful ... now have it so that if any phase
FAILs, it exits and doesn't just go to the next, which most likely
won't work either ...
since we're removing a bunch of the sleep(s), let's increase the couple
that remain so that we have pretty much the same overall 'randomness' for
reporting systems ...
fix the code a bit to allow for setting the checkin_server to something
different ... instructions soon to follow on web site as to how to use
this to use apache to redirect requests through it as a 'light proxy'
Clean up the code a wee bit, since the non-FreeBSD case(s), where OSs don't
have pciconf, just keeps growing ...
Fix up code for use of /etc/rc.conf.local for those systems that don't use
periodic.conf
Submitted by: Thorsten Glaser <tg@mirbsd.de>
Apparently Sparc doesn't like using openssl to generate a random value ... it
created a 41 year 'sleep' time for one ...
Fix it to use sum instead of openssl
Add 'unset HTTP_USER_AGENT' ... the backend is setup to only accept specific
USER_AGENTs, so if someone has this set in their shell when running it manually,
it will break
Add random sleep 'breaks' before doing the fetchs, to help reduce impact on
the server when all reporting hosts submit their data ... get them someone
staggered ...
Submitted by: Andre Albsmeier <Andre.Albsmeier@siemens.com>
Several cleanups, and extensions to allow NetBSD to submit ...
- uri_escape IDTOKEN properly
- use ftp vs fetch
- use /etc/rc.conf instead of /etc/periodic.conf
Submitted by: David Brownlee <abs@absd.org>
First, set PATH similar to /etc/rc.shutdown, so that we don't have to
hardcode openssl, to allow for those using ports for it to easily make
use of acript
Second, check timestamp of /var/db/bsdstats to make sure that the 15
minutes have passed before submitting, so that ppl don't get that 'false
success' when they ctl-C
Submitted by: Oliver Fromme <olli@lurza.secnetix.de>
Ensure that data returned from server on initial challenge is, in
fact, only the data that we expect.
Submitted by: Matthew Seaman <m.seaman@infracaninophile.co.uk>
This is mostly a rewrite of the bsdstats script:
- eliminate use of ip/hostname as unique keys
- reduce # of fetch's down to 4
- move to using bsdstats.org
- put in a time delay on initial challenge/response
to make it a bit more difficult for fakers
Add some verbosity to the script, letting ppl know that a report has
been sent, and/or that it is disabled and how to enable it ...
Submitted by: Paul Schmehl <pauls@utdallas.edu>
Record Operating System itself, in case someone in the NetBSD,
OpenBSD or DragonflyBSD decides they want to add to the stats, so
that we know where the numbers belong ...
