At the time of writing, https://www.freshports.org/net-mgmt/mrtg/ showed this for Build dependencies:

perl5>=5.20<5.21 : lang/perl5.20

The last commit for this port was on 20 Jul 2016 09:33:32

If you view Mk/bsd.default-versions.mk at that time, you will find 5.20 is specified.

However, should that file get updated, the information on the port become stale.

Possible solution.

• track which ports use USES= and refresh them after a change to Mk/bsd.default-versions.mk

The sanity test for checking LIB_DEPENDS, RUN_DEPENDS, etc assumes the dependency is a 'category/port' specification. This is not always the case.

Take this commit as an example: https://www.freshports.org/commit.php?message_id=201510130745.t9D7j99D028221@repo.freebsd.org

The sanity test failures appear below (see also https://www.freshports.org/sanity_test_failures.php?message_id=201510130745.t9D7j99D028221@repo.freebsd.org)

The sanity test doesn't know about versions: xkeyboard-config>=2.5:x11/xkeyboard-config

It doesn't know about libraries:: libxcb-keysyms.so

The sanity test also doesn't know about package names: xkbcomp:x11/xkbcomp but FreshPorts does record that information:

## package_name 

xkbcomp
(1 row)

That sanity test has been disabled until it can receive more attention.

This message was generated by the FreshPorts Daemon.

You recently made this commit, which FreshPorts FreshPorts had trouble
processing:

MessageID: [email protected]
Subject  : svn commit: r399179 - in head: . Mk emulators/virtualbox-ose-additions emulators/virtualbox-ose/files x11-drivers x11-drivers/xf86-input-acecad x11-drivers/xf86-input-egalax x11-drivers/xf86-input-...

The following is a list of the ports which had errors:

http://www.freshports.org/x11-servers/xorg-nestserver/
http://www.freshports.org/x11-servers/xephyr/
http://www.freshports.org/x11-servers/xorg-dmx/
http://www.freshports.org/x11-servers/xorg-server/

The exact errors appear below.

x11-servers/xorg-nestserver:

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xorg-nestserver does
not exist: 'libxcb-keysyms.so:x11/xcb-util-keysyms' on branch 'head'.

x11-servers/xephyr:

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xephyr does not
exist: 'xkeyboard-config>=2.5:x11/xkeyboard-config' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xephyr does not
exist: 'xkbcomp:x11/xkbcomp' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xephyr does not
exist: 'libdrm.so:graphics/libdrm' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xephyr does not
exist: 'libepoxy.so:graphics/libepoxy' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xephyr does not
exist: 'xkeyboard-config>=2.5:x11/xkeyboard-config' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xephyr does not
exist: 'xkbcomp:x11/xkbcomp' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xephyr does not
exist: 'libdrm.so:graphics/libdrm' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xephyr does not
exist: 'libepoxy.so:graphics/libepoxy' on branch 'head'.

x11-servers/xorg-dmx:

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xorg-dmx does not
exist: 'xkeyboard-config>=2.5:x11/xkeyboard-config' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xorg-dmx does not
exist: 'xkbcomp:x11/xkbcomp' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xorg-dmx does not
exist: 'libdrm.so:graphics/libdrm' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xorg-dmx does not
exist: 'libepoxy.so:graphics/libepoxy' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xorg-dmx does not
exist: 'xkeyboard-config>=2.5:x11/xkeyboard-config' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xorg-dmx does not
exist: 'xkbcomp:x11/xkbcomp' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xorg-dmx does not
exist: 'libdrm.so:graphics/libdrm' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xorg-dmx does not
exist: 'libepoxy.so:graphics/libepoxy' on branch 'head'.

x11-servers/xorg-server:

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xorg-server does not
exist: 'xkeyboard-config>=2.5:x11/xkeyboard-config' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xorg-server does not
exist: 'xkbcomp:x11/xkbcomp' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xorg-server does not
exist: 'libdrm.so:graphics/libdrm' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xorg-server does not
exist: 'libepoxy.so:graphics/libepoxy' on branch 'head'.

see https://www.freshports.org/phorum/read.php?f=1&i=1492&t=1492

The "This port is required by:" section only lists ports which require this port with OPTIONS_DEFAULT.

It would be handy if it could also list ports which require this port in any configuration of enabled or disabled options (perhaps with a note or asterix to point out that it does not require this port by default).

Thanks for this amazing site. I already use it for years.
Since the support for branches the rss feed contains everything which I think is a bit confusing if you do not follow a branch or even more when you do follow a branch.
Is it possible to get the rss per branch? Or only the changes on head/trunk?

Clicking on Distfiles Availability results in a big message saying "This service has been discontinued."

Portscout is the canonical replacement and should be used instead.

There are several ports that are restricted to only certain architectures. I know that Openjdk (for example) is restricted to i386, AMD64.

When searching with Depends Build/Lib/Run and sounds like for gimp or anything with Plain Text selected first line is:

"NOTE: Instead of using 'sounds like' as instructed, the system used 'containing'. See the notes below for why this is done.<br>graphics/cimg

with unparsed <br> so first result is in the same line

review cf99655

kibana50-x-pack installs into /usr/local/www/kibana50/plugins/x-pack/x-pack

Kibana expects the plugin to be in /usr/local/www/kibana50/plugins/x-pack instead (package.json is required in this directory).

Can be verified by running

# /usr/local/bin/node /usr/local/www/kibana50/src/cli_plugin list
which returns

[email protected]
when installed correctly, but

/usr/local/www/kibana50/src/cli_plugin/list/list.js:26
        throw new Error('Unable to read package.json file for plugin ' + filename);
        ^

Error: Unable to read package.json file for plugin x-pack
    at /usr/local/www/kibana50/src/cli_plugin/list/list.js:26:15
    at Array.forEach (native)
    at list (/usr/local/www/kibana50/src/cli_plugin/list/list.js:13:44)
    at Command.processCommand (/usr/local/www/kibana50/src/cli_plugin/list/index.js:38:25)
    at Command.<anonymous> (/usr/local/www/kibana50/src/cli/command.js:90:18)
    at Command.listener (/usr/local/www/kibana50/node_modules/commander/index.js:301:8)
    at emitTwo (events.js:87:13)
    at Command.emit (events.js:172:7)
    at Command.parseArgs (/usr/local/www/kibana50/node_modules/commander/index.js:610:12)
    at Command.parse (/usr/local/www/kibana50/node_modules/commander/index.js:458:21)
    at Object.<anonymous> (/usr/local/www/kibana50/src/cli_plugin/cli.js:54:9)
    at Module._compile (module.js:409:26)
    at Module._extensions..js (module.js:416:10)
    at Object.require.extensions.(anonymous function) [as .js] (/usr/local/www/kibana50/node_modules/babel-core/lib/api/register/node.js:214:7)
    at Module.load (module.js:343:32)
    at Function.Module._load (module.js:300:12)

otherwise

Is the daily_refreshes required?

I'm planning on forking this repository and modernizing the codebase a bit. This will mostly be as a learning experience on my part, but if you like the direction I'm going, I would love to merge it back at some point.

In order to allow others to work on this project, it would be nice to have a copy of the database schema. A lot of it can be inferred from existing queries scattered throughout, but it would be easier to have a single place to view everything. A few rows of sample data would be wonderful too. pg_dump with the -s flag can export schema only, which would provide a great starting point.

I'm not too familiar with Postgres, but the POSTGRESQL.sql file looks it sets up permissions on existing tables, right?

Each search result, and each details page now displays:

ONLY_FOR_ARCHS: nil
NOT_FOR_ARCHS: nil

For something with an ONLY_FOR_ARCHS defined, you get:

ONLY_FOR_ARCHS: amd64
NOT_FOR_ARCHS: nil

The line should not display if the value is nil

see https://duck.co/ideas/idea/936/grab-bsd-port-summary-from-freshports

Some (many? all?) search queries are returning duplicate results.

example:

http://www.freshports.org/search.php?stype=name&method=match&query=bind9&num=10&orderby=category&orderbyupdown=asc&search=Search&format=plaintext&minimal=1

produces the following output:

dns/bind9-devel
dns/bind910
dns/bind910
dns/bind911
dns/bind911
dns/bind99
dns/bind99

see http://news.freshports.org/2015/12/29/not-all-ports-have-packages/

re: https://www.freshports.org/sanity_test_failures.php?message_id=201605161739.u4GHdj5r070412@repo.freebsd.org

We have two ports named linux-c7-flashplugin11, which is OK, because one is on a branch.

freshports.org=# select id, category_id, date_added, element_pathname(element_id) from ports where id in (select id from ports_active where name = 'linux-c7-flashplugin11');
id | category_id | date_added | element_pathname
-------+-------------+-------------------------------+---------------------------------------------------
39414 | 2 | 2016-10-19 14:50:44.746043+00 | /ports/head/www/linux-c7-flashplugin11
39543 | 2 | 2016-10-24 16:57:46.761717+00 | /ports/branches/2016Q4/www/linux-c7-flashplugin11
(2 rows)

freshports.org=#

39543 was created by a commit to the master port in r424578. linux-c7-flashplugin11 did not exist on the branch, so it was created.

The problem with 39543 is the ports table has only the bare details. The entry was never updated from the Makefile. It might be because the update process for branches messed up.

Hi Dan,

we have no freshports for a whole day. In future it would be great to have a note in freebsd-ports@ or in your twitter account for cases like that with likely ETAs on the issue. Thank you!

There are commits, they just do not show. Why?

Can a port's distsize attribute be displayed for each port?

From twitter:
I can sense a new table coming on: distinfo (port_id, distname, sha256, size...)

Should be simple to extract SIZE and parse the line for each file.

Compare these two URLS:

http://www.freshports.org/commit.php?message_id=201601062012.u06KC53V085831@repo.freebsd.org

http://www.freshports.org/commit.php?message_id=201601062012.u06KC53V085831@repo.freebsd.org&files=yes

Not sure if this is an actual issue...

At https://www.freshports.org/date.php?date=2014/3/31 the top of the page shows:

< Latest commits | 1 April | 29 March > | Number of commits: 118

I expect:

< Latest commits | 1 April | 30 March > | Number of commits: 118

Additionally, https://www.freshports.org/date.php?date=2014/3/25 and before have no commits

Today, if you search for something in a pkg-plist, it only looks at the contents of this file. It should probably combine the file contents with PLIST_FILES and PLIST_DIRS.

% make -VPLIST_FILES
libexec/apache24/mod_auth_openid.so /usr/local/share/licenses/ap24-mod_auth_openid-0.8_1/catalog.mk  /usr/local/share/licenses/ap24-mod_auth_openid-0.8_1/LICENSE /usr/local/share/licenses/ap24-mod_auth_openid-0.8_1/MIT`

A better approach might be to call
make generate-plist
from each port dir and then add the resulting file to the DB. This will merge pkg-plist + PLIST_FILES + PLIST_DIRS.

The resulting file is in ${WRKDIR}/.PLIST.mktmp

This doesn't cover all cases (like any ports that auto-plist) but gets olivier closer to what he is asking for.

ADDED BY dvl: Current situation: FreshPorts displays only information contained in the pkg-config file. This proposed change will increase the number of ports with at least some 'pkg info -l' type information

Hey

I've installed the php-composer but it dosn't work. It throws an error that "mbstring" is missing. After installing the right mbstring it was fixed.

Maybe it can be helpfull to install it together with php-composer. Otherwise it can be usefull to give an post install notice like "Make sure that phpXX-mbstring is installed, if not install phpXX-mbstring. XX is the version of php you have installed."

Kind Regards

Moved here from dlangille/freshports#5

At present, each port and category uses this method for providing content:

ErrorDocument 404 /missing.php

Let's change that by doing a rewrite to start with so every URL goes to missing.php (for example).

At http://www.freshports.org/ports-deleted.php, click on any SVNWeb link and you get a 404. The URL is wrong (e.g. https://ports/head/accessibility/at-poke/?pathrev=362407).

The hostname is missing from the SQL query.

see https://www.freshports.org/phorum/read.php?f=1&i=1492&t=1492

I wonder if database connections can be reserved?

DBI connect([redacted])failed: FATAL: remaining connection slots are reserved for non-replication superuser connections at [filename redacted] line [line number redacted].

see https://www.freshports.org/ports-mgmt/portlint/ where there is a value for master sites.

However, there is no such value now:

[dan@jester:/usr/local/FreshPorts/ports-jail/usr/local/repos/PORTS-head/ports-mgmt/portlint] $ make -V _MASTER_SITES_ALL -f /usr/local/FreshPorts/ports-jail/usr/local/repos/PORTS-head/ports-mgmt/portlint/Makefile 

[dan@jester:/usr/local/FreshPorts/ports-jail/usr/local/repos/PORTS-head/ports-mgmt/portlint] $

Could this be related to the change from from 'make master-site-all' to 'make -V _MASTER_SITES_ALL'?

The sanity test for checking LIB_DEPENDS, RUN_DEPENDS, etc assumes the dependency is a 'category/port' specification. This is not always the case.

Take this commit as an example: https://www.freshports.org/commit.php?message_id=201510130745.t9D7j99D028221@repo.freebsd.org

The sanity test failures appear below (see also https://www.freshports.org/sanity_test_failures.php?message_id=201510130745.t9D7j99D028221@repo.freebsd.org)

The sanity test doesn't know about versions: xkeyboard-config>=2.5:x11/xkeyboard-config

It doesn't know about libraries:: libxcb-keysyms.so

The sanity test also doesn't know about package names: xkbcomp:x11/xkbcomp but FreshPorts does record that information:

freshports.org=# select package_name from ports where package_name = 'xkbcomp';
package_name 
--------------
xkbcomp
(1 row)

That sanity test has been disabled until it can receive more attention.

This message was generated by the FreshPorts Daemon.

You recently made this commit, which FreshPorts FreshPorts had trouble
processing:

MessageID: [email protected]
Subject  : svn commit: r399179 - in head: . Mk emulators/virtualbox-ose-additions emulators/virtualbox-ose/files x11-drivers x11-drivers/xf86-input-acecad x11-drivers/xf86-input-egalax x11-drivers/xf86-input-...

The following is a list of the ports which had errors:

http://www.freshports.org/x11-servers/xorg-nestserver/
http://www.freshports.org/x11-servers/xephyr/
http://www.freshports.org/x11-servers/xorg-dmx/
http://www.freshports.org/x11-servers/xorg-server/


The exact errors appear below.

x11-servers/xorg-nestserver:

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xorg-nestserver does
not exist: 'libxcb-keysyms.so:x11/xcb-util-keysyms' on branch 'head'.

x11-servers/xephyr:

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xephyr does not
exist: 'xkeyboard-config>=2.5:x11/xkeyboard-config' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xephyr does not
exist: 'xkbcomp:x11/xkbcomp' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xephyr does not
exist: 'libdrm.so:graphics/libdrm' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xephyr does not
exist: 'libepoxy.so:graphics/libepoxy' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xephyr does not
exist: 'xkeyboard-config>=2.5:x11/xkeyboard-config' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xephyr does not
exist: 'xkbcomp:x11/xkbcomp' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xephyr does not
exist: 'libdrm.so:graphics/libdrm' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xephyr does not
exist: 'libepoxy.so:graphics/libepoxy' on branch 'head'.

x11-servers/xorg-dmx:

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xorg-dmx does not
exist: 'xkeyboard-config>=2.5:x11/xkeyboard-config' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xorg-dmx does not
exist: 'xkbcomp:x11/xkbcomp' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xorg-dmx does not
exist: 'libdrm.so:graphics/libdrm' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xorg-dmx does not
exist: 'libepoxy.so:graphics/libepoxy' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xorg-dmx does not
exist: 'xkeyboard-config>=2.5:x11/xkeyboard-config' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xorg-dmx does not
exist: 'xkbcomp:x11/xkbcomp' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xorg-dmx does not
exist: 'libdrm.so:graphics/libdrm' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xorg-dmx does not
exist: 'libepoxy.so:graphics/libepoxy' on branch 'head'.

x11-servers/xorg-server:

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xorg-server does not
exist: 'xkeyboard-config>=2.5:x11/xkeyboard-config' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the RUN_DEPENDS of x11-servers/xorg-server does not
exist: 'xkbcomp:x11/xkbcomp' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xorg-server does not
exist: 'libdrm.so:graphics/libdrm' on branch 'head'.

NOTE: this particular sanity test is very experimental
A port specified in the LIB_DEPENDS of x11-servers/xorg-server does not
exist: 'libepoxy.so:graphics/libepoxy' on branch 'head'.

Eg, commit r431138

I think this could be solved by a CSS guru

Only the first 100 commits are shown, even if you click on other pages.

Example:

http://www.freshports.org/sysutils/apcupsd/?page=2
http://dev.freshports.org/sysutils/apcupsd/?page=2

Hello!
I've recently installed cacti-1.1.24 and it depends on rrdtool-1.7.0
While trying to install ntop 5.0.1_13 I've got an error,

_ntop 5.0.1_13 depends on package: rrdtool 12>=1.2 not found
rrdtool12-1.2.30_7 conflicts with installed package(s):
rrdtool-1.7.0_

So now, I have to choose what to run - cacti+rrdtool-1.7 or ntop+rrdtool-1.2

I wonder if there's a way to build and use ntop with fresh rrdtool.

Any help will be appreciated.

FreshPorts should integrate with portscout and show information about new versions. portscout doesn’t have an API per say, but the RSS feeds are easy enough to monitor and contains information about new releases.

https://portscout.freebsd.org/rss/[email protected]

Suggestion for using the data:

$package_name version $current_version is available in ports. The $package_name maintainer was notified about a new version $new_version_from_portscout on $date_from_feed."

Boost 1.63 boost/config/stdlib/libcpp.hpp has code

#if __cplusplus < 201103
...
#  define BOOST_NO_CXX11_HDR_MUTEX
...

So the macro is always defined despite of libc++ has std::mutex!
Boost.Context - C++11 library never uses std::mutex as a consequence.

Just a crazy idea. You've likely seen these badges used by many GitHub repos to report continous integration status and other stuff. Why not make FreshPorts generate a badge for a port by request? The badge should simply show DISTVERSION, like this:

This has numerous purposes among which is raising awareness of FreeBSD ports/Freshports among users, a convenient way to know the project is supported by FreeBSD and a way to know that the port is at latest version (or whether it falls behind).

The simplest way to implement this is just a make a per-port link (e.g. http://www.freshports.org/x11/xproto/badge or http://www.freshports.org/x11/xproto/?badge) which redirects to shields.io like https://img.shields.io/badge/FreeBSD_port-${DISTVERSION}-blue.svg. Or you could just generate the same .svg by yourself to not depend on third party site.

pkg_plist is deprecated by #50

makefile - why store makefile? It's in the repo. We can link to that instead.

From swills: http://www.freshports.org/x11-themes/ -- how is adwaita-qt5 listed twice?

Answer: Commits on quarterly branches.

When uploading a watch list, if it contains certain ports, the following error is returned, and the watch list is unchanged:

ERROR: more than one row returned by a subquery used as an expression CONTEXT: SQL statement "UPDATE watch_list_staging SET element_id = ( SELECT ports.element_id FROM ports, categories, element WHERE element.name = watch_list_staging.port AND categories.name = watch_list_staging.category AND ports.element_id = element.id AND ports.category_id = categories.id) WHERE user_id = UserID" PL/pgSQL function watchliststagingprocess(integer) line 17 at SQL statement sql = select WatchListStagingProcess(13320)

I find that any of these ports in the watch list result in the problem:

• dns/bind911
• security/ca_root_nss
• ftp/curl
• multimedia/ffmpeg
• devel/jsoncpp
• www/mod_php56
• lang/perl5.24
• lang/php56 (and probably all its slave ports)
• ports-mgmt/pkg
• security/sudo

On 22 Oct, we had:

 Port count: 26568 http://www.FreshPorts.org/categories.php
New ports:     8 http://www.FreshPorts.org/ports-new.php?interval=today

21 Oct 2016

http://www.freshports.org/devel/libastylej-arduino/
    Source code indenter, formatter, and beautifier (JNI library)

http://www.freshports.org/devel/nspr/
    Platform-neutral API for system level and libc like functions

http://www.freshports.org/www/firefox-i18n/
    Localized interface for Firefox

22 Oct 2016

http://www.freshports.org/audio/raop_play/
    Music file player for Apple Airport Express

http://www.freshports.org/devel/ncurses/
    Library for terminal-independent, full-screen output

http://www.freshports.org/games/openra/
    Multiplayer re-envisioning of early RTS games by Westwood Studios

http://www.freshports.org/security/nmap/
    Port scanning utility for large networks

http://www.freshports.org/security/zenmap/
    GUI frontend for the Nmap scanning utility

When a port dependency is removed in the ports tree, the FreshPorts page for the dependent port does not remove the port which previously depended on it from the "This port is required by" list.

This was recently noticed when graphics/mapnik was updated to no longer depend on devel/py-icu:
https://svnweb.freebsd.org/ports/head/graphics/mapnik/Makefile?r1=410719&r2=410718&pathrev=410719

After this commit, the FreshPorts page http://www.freshports.org/graphics/mapnik was correctly updated to no longer show a dependency on devel/py-icu.

However, the FreshPorts page http://www.freshports.org/devel/py-icu still incorrectly showed that "this port is required by: graphics/mapnik".

After discussion with Dan Langille on IRC, it was determined that this was caused by a cache issue on the FreshPorts server. The cache for devel/py-icu was not invalidated.

[dan@supernews:/usr/local/www/freshports.org/dynamic/caching/cache/ports/devel/py-icu] $ ls -lt total 58 -rw-rw-r-- 1 www www 28699 Feb 8 21:30 Detail.PageSize500.PageNum1.html -rw-rw-r-- 1 www www 28626 Oct 30 18:09 Detail.PageSize100.PageNum1.html [dan@supernews:/usr/local/www/freshports.org/dynamic/caching/cache/ports/devel/py-icu] $ rm Detail.PageSize* [dan@supernews:/usr/local/www/freshports.org/dynamic/caching/cache/ports/devel/py-icu] $

Sometimes I have to this manually:

This is a list of ports that do not have entries in the ports_categories table
This can be fixed with this query:
begin; insert into ports_categories select id, category_id from ports_active PA WHERE NOT EXISTS (SELECT * from ports_categories PC where PC.port_id = PA.id and PC.category_id = PA.category_id);
id | category_id | name | category | port
-------+-------------+------------+----------+---------------------
38544 | 18 | py-dbfread | textproc | textproc/py-dbfread
(1 row)

That should invoke cache clearing.

See https://www.freshports.org/lang/php70 where you will find:

Slave ports

• archivers/php70-bz2
• archivers/php70-bz2
• archivers/php70-phar
• archivers/php70-phar
• archivers/php70-zip
• archivers/php70-zip

Possibly related to #31

After a discussion on irc and a tweet at Dan, along with some additional ideas I had on my own, I think there's room for improvement with respect to specific rss feeds, so here's a few things that I've thought of, in no particular order, which new feeds could do:

• Include only newly added ports, not updated ones
• Include only broken ports
• Include only vulnerable ports

The first suggestion is mostly because sometimes a neat little thing gets added which someone might not have thought of using, and number two and three will help people who don't regularly build ports with poudriere, or who don't keep packages up-to-date.

I realize that all of that information could be scraped, but then only I would have that information, and I suspect others can use it.

https://www.freshports.org/ports-new.php has two issues for all ports displayed:

"License: not specified in port"

and the svnweb link points to https://svn.freebsd.org/ instead of the port's svnweb page.

I discovered this for /UPDATING

The cache is being created in head/UPDATING - should be UPDATING

There is nothing which clears cache for non-ports (e.g. Mk/Uses/cargo.mk)

Sometimes you want to know what this port installs. Therefore, add a link from each page port to

https://svnweb.freebsd.org/ports/head/devel/p5-File-Temp/pkg-plist?view=co

For example...