]> eyrie.org Git - kerberos/wallet.git/log
kerberos/wallet.git
9 years agoCreate the directory for the Wallet::Object::Duo classes release/1.2
Russ Allbery [Tue, 9 Dec 2014 04:53:15 +0000 (20:53 -0800)]
Create the directory for the Wallet::Object::Duo classes

9 years agoAdd new files to the distribution
Russ Allbery [Tue, 9 Dec 2014 04:48:23 +0000 (20:48 -0800)]
Add new files to the distribution

9 years agoAdd stopwords for new modules
Russ Allbery [Tue, 9 Dec 2014 04:23:49 +0000 (20:23 -0800)]
Add stopwords for new modules

Change-Id: Ie1d2bcee19ace444f6f7083814133593b160d97d

9 years agoUpdate skip count for systems without sqlite3
Russ Allbery [Tue, 9 Dec 2014 04:23:40 +0000 (20:23 -0800)]
Update skip count for systems without sqlite3

Change-Id: I2180daf5055a90ae52b8a851f514993004da5303

9 years agoBe more robust when setting up the Perl tests
Russ Allbery [Tue, 9 Dec 2014 04:23:01 +0000 (20:23 -0800)]
Be more robust when setting up the Perl tests

If the Test library directory already exists, we previously failed.
Use mkdir -p and probe for it in Autoconf to avoid this.

Change-Id: I1ad9f1a83af1f2ebfe1b2337aaab99913b4edeea

9 years agoRelease 1.2
Russ Allbery [Tue, 9 Dec 2014 02:43:29 +0000 (18:43 -0800)]
Release 1.2

Change-Id: I4157db0f690542db0eb1bfbcb7e15bfee890cd65

9 years agoMerge pull request #1 from jonrober/master
Russ Allbery [Tue, 9 Dec 2014 02:33:09 +0000 (18:33 -0800)]
Merge pull request #1 from jonrober/master

Duo and rename updates

9 years agoStanford.pm: Fixed incorrect name for new duo host parsing function
Jon Robertson [Thu, 6 Nov 2014 23:37:23 +0000 (15:37 -0800)]
Stanford.pm: Fixed incorrect name for new duo host parsing function

Change-Id: Ica75f6614476088a9952cd7d97749d27811aed7e

9 years agoUpdated NEWS with pending release notes
Jon Robertson [Tue, 28 Oct 2014 03:13:09 +0000 (20:13 -0700)]
Updated NEWS with pending release notes

Change-Id: I7730b4779180d7ad85dd4d1b6e71d8576a27a662

9 years agoAdded rename support for file objects
Jon Robertson [Wed, 15 Oct 2014 00:06:43 +0000 (17:06 -0700)]
Added rename support for file objects

File objects now support a rename command, which will rename the object
and move the file to the right spot in the file store under its new
name.

Change-Id: I10ea2b8012586d69f0894905cfba54a738f3e418

9 years agoAdded rename support for file objects
Jon Robertson [Wed, 15 Oct 2014 00:06:43 +0000 (17:06 -0700)]
Added rename support for file objects

File objects now support a rename command, which will rename the object
and move the file to the right spot in the file store under its new
name.

Change-Id: I10ea2b8012586d69f0894905cfba54a738f3e418

9 years agoAdded rename support for file objects
Jon Robertson [Wed, 15 Oct 2014 00:06:43 +0000 (17:06 -0700)]
Added rename support for file objects

File objects now support a rename command, which will rename the object
and move the file to the right spot in the file store under its new
name.

Change-Id: I10ea2b8012586d69f0894905cfba54a738f3e418

9 years agoAdded Duo to the Stanford policy module
Jon Robertson [Tue, 14 Oct 2014 20:23:40 +0000 (13:23 -0700)]
Added Duo to the Stanford policy module

Duo object types currently all assume that the name of the object is the
hostname of the server it's for.

Change-Id: Ieb5ba144cd39d6aeb3a20466c75a2836a170744f

9 years agoWallet::Object::Duo: Respect new type field in Duo table
Jon Robertson [Tue, 14 Oct 2014 05:35:23 +0000 (22:35 -0700)]
Wallet::Object::Duo: Respect new type field in Duo table

Make all the searches and creations for the Duo table add or search for
the type field as well.  This avoids one Duo type clobbering another
for the same object name.

Change-Id: I62192c3616f43c7acd8ce3f94db8a0e43e77e317

9 years agoAdded Duo RDP object to the wallet
Jon Robertson [Tue, 14 Oct 2014 05:28:30 +0000 (22:28 -0700)]
Added Duo RDP object to the wallet

Change-Id: Ic728297fa830ffdd40c1580e32a81f8c5123f66a

9 years agoFixed new SQLite schema to include IF EXISTS on drop tables
Jon Robertson [Tue, 14 Oct 2014 05:18:42 +0000 (22:18 -0700)]
Fixed new SQLite schema to include IF EXISTS on drop tables

Change-Id: I54edbb543be8bfcf0de355da3cef82c6ac1bf27f

9 years agoAdded new column in the Duo table
Jon Robertson [Fri, 10 Oct 2014 05:24:18 +0000 (22:24 -0700)]
Added new column in the Duo table

New column is required to differentiate the Duo table entries now that
we have more than one Duo object type.  Added the new field and rebuilt
schema definitions and upgrade files.

Change-Id: Icf538eaded93f4f2820984c087d4850a586a7db1

9 years agoCreated tests for new Duo object types
Jon Robertson [Thu, 9 Oct 2014 16:55:21 +0000 (09:55 -0700)]
Created tests for new Duo object types

Change-Id: I818be125f3195316b44e650ba6e05b8e0b831ea6

9 years agoNEWS: Added item for Duo object expansion
Jon Robertson [Thu, 9 Oct 2014 16:19:22 +0000 (09:19 -0700)]
NEWS: Added item for Duo object expansion

Change-Id: I91b8b5fd4043effe8b23a62624c47519976ace64

9 years agoSplit Duo type out into multiple sub-types
Jon Robertson [Thu, 9 Oct 2014 03:33:56 +0000 (20:33 -0700)]
Split Duo type out into multiple sub-types

The existing functionality is now in the duo-pam object type.  The old
duo type now returns output in a generic config file, and new types for
the Duo auth proxy in LDAP and Radius proxies are added.

Change-Id: I1525d79b44dafcf3ef85368297baefafcb5dc179

9 years agoUpdate JIRA references in TODO, remove finished items
Russ Allbery [Sun, 3 Aug 2014 03:42:16 +0000 (20:42 -0700)]
Update JIRA references in TODO, remove finished items

JIRA's structure has been reorganized, so change the JIRA issue
references to be accurate again.  Add JIRA issues for the things
that hadn't been entered into JIRA yet.

Change-Id: I385d19b09108d94430bbe142a2b3f6f80b5e4c39
Reviewed-on: https://gerrit.stanford.edu/1588
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoRegister the duo object type on database creation release/1.1
Russ Allbery [Wed, 16 Jul 2014 23:44:37 +0000 (16:44 -0700)]
Register the duo object type on database creation

Change-Id: I1c53e0503b29d7add289d26e67b11f9789ba8ad8
Reviewed-on: https://gerrit.stanford.edu/1576
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoGo back to recording the ADMIN ACL in history
Russ Allbery [Wed, 16 Jul 2014 23:20:10 +0000 (16:20 -0700)]
Go back to recording the ADMIN ACL in history

This turned out to not be necessary for testing since I was already
using sqlite3 to load an unversioned schema.  Remove the offending
line and restore the old code with some cleanup.

Change-Id: I282b6f3b4754e4899222be6366b77a47f0cb7189
Reviewed-on: https://gerrit.stanford.edu/1575
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoParse the Duo key_file in relaxed JSON mode
Russ Allbery [Wed, 16 Jul 2014 22:52:51 +0000 (15:52 -0700)]
Parse the Duo key_file in relaxed JSON mode

When reading the Duo object configuration to retrieve the Duo
admin server, parse the JSON in relaxed mode to match the behavior
of Net::Duo itself.  Otherwise, we get hung up on trailing commas
that Net::Duo doesn't care about.

Change-Id: I0a7347b22e379fe5dfe5fdabaec3e23420cf9a63
Reviewed-on: https://gerrit.stanford.edu/1574
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoInclude the Duo type in the name of Duo integrations
Russ Allbery [Wed, 16 Jul 2014 22:51:13 +0000 (15:51 -0700)]
Include the Duo type in the name of Duo integrations

Eventually, there will be multiple object types for different Duo
integrations, and they will need to have unique names.  Add the
Duo type in parentheses after the name to help ensure this.

Change-Id: I679130f9136077fc6bf5d8c6c9ad98ec83b400d0
Reviewed-on: https://gerrit.stanford.edu/1573
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoMove perl/create-ddl into PERL_FILES to copy it during build
Russ Allbery [Wed, 16 Jul 2014 20:31:45 +0000 (13:31 -0700)]
Move perl/create-ddl into PERL_FILES to copy it during build

Otherwise, there are warnings from Build.PL due to the file missing
from the manifest.

Change-Id: I32db0199bfda25ab8235ab965bfbbca8bee180b8
Reviewed-on: https://gerrit.stanford.edu/1572
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoAdd more Perl dependencies, bump Perl version
Russ Allbery [Wed, 16 Jul 2014 20:13:27 +0000 (13:13 -0700)]
Add more Perl dependencies, bump Perl version

Also set module_name (uselessly) to avoid warnings from
Module::Build.

Change-Id: I53426a096f4133f27aa3315b4be24385a3476793
Reviewed-on: https://gerrit.stanford.edu/1571
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoUpdate MANIFEST for new Perl tests and test reorganization
Russ Allbery [Wed, 16 Jul 2014 20:13:14 +0000 (13:13 -0700)]
Update MANIFEST for new Perl tests and test reorganization

Change-Id: I984f48d667acab4cfcb7e0c115773e34e6335d65
Reviewed-on: https://gerrit.stanford.edu/1570
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoInclude wallet-rekey-periodic in the distribution
Russ Allbery [Wed, 16 Jul 2014 20:09:47 +0000 (13:09 -0700)]
Include wallet-rekey-periodic in the distribution

Change-Id: I3dd9ae38d638cddf2307f5e07cb4a2e01422e172
Reviewed-on: https://gerrit.stanford.edu/1569
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoWork around Automake distcheck process
Russ Allbery [Wed, 16 Jul 2014 19:39:06 +0000 (12:39 -0700)]
Work around Automake distcheck process

Automake insists on not using DESTDIR for distcheck and instead
relying on prefix, but we don't want Perl module installation to
follow prefix since that may result in a module install directory that
isn't in Perl's search path.  So, if and only if we're running under
distcheck, we pass the prefix in as --install_base.

When copying the Test::RRA Perl modules into the perl/t/lib tree,
use separate mkdir and $(INSTALL_DATA) instead of cp -R.  The latter
copies the read-only permissions, and then distclean cannot remove
the files.

Change-Id: Ic1879defad993c76384f7c207cd04cb67889a7ac
Reviewed-on: https://gerrit.stanford.edu/1568
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoChange the Perl install rule to reflect Module::Build
Russ Allbery [Wed, 16 Jul 2014 19:08:58 +0000 (12:08 -0700)]
Change the Perl install rule to reflect Module::Build

Change-Id: I36565462a248cef0ff1560b5a1d89a20353d566f
Reviewed-on: https://gerrit.stanford.edu/1567
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoAdd perl/t/data/perl.conf to PERL_FILES
Russ Allbery [Wed, 16 Jul 2014 18:57:22 +0000 (11:57 -0700)]
Add perl/t/data/perl.conf to PERL_FILES

Change-Id: I0d56ea7b64cdcc43bf59f803077d076414b1a1ce
Reviewed-on: https://gerrit.stanford.edu/1566
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoFix test skip numbering for keytab tests
Russ Allbery [Wed, 16 Jul 2014 18:56:25 +0000 (11:56 -0700)]
Fix test skip numbering for keytab tests

Change-Id: I5143d487d6b3623bc2be1724ed766b8709feb506
Reviewed-on: https://gerrit.stanford.edu/1565
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoAdd missing directories to PERL_DIRECTORIES
Russ Allbery [Wed, 16 Jul 2014 18:24:59 +0000 (11:24 -0700)]
Add missing directories to PERL_DIRECTORIES

Change-Id: I5390ec1ea8ba90394454a75acb54f1f4a25f9c83
Reviewed-on: https://gerrit.stanford.edu/1564
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoRelease 1.1
Russ Allbery [Wed, 16 Jul 2014 18:21:32 +0000 (11:21 -0700)]
Release 1.1

Change-Id: I823bb20d129e4c1efdb607821adc3b134c2f6276
Reviewed-on: https://gerrit.stanford.edu/1563
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoMove wallet-rekey-periodic license into POD
Russ Allbery [Wed, 16 Jul 2014 18:21:07 +0000 (11:21 -0700)]
Move wallet-rekey-periodic license into POD

To ensure that the generated man page also contains a license, move
the license text down into the POD.

Change-Id: Iaeedfbffccd1510d50bf1f84f396e1b1f8fda8fc
Reviewed-on: https://gerrit.stanford.edu/1562
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoAdd license statements to new SQL files
Russ Allbery [Wed, 16 Jul 2014 18:18:44 +0000 (11:18 -0700)]
Add license statements to new SQL files

Change-Id: I518a175998aa77920b08c43e3a6b890bbab59280
Reviewed-on: https://gerrit.stanford.edu/1561
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoUse xcalloc instead of xmalloc with calculated sizes
Russ Allbery [Wed, 16 Jul 2014 06:14:51 +0000 (23:14 -0700)]
Use xcalloc instead of xmalloc with calculated sizes

Ensure we don't have integer overflow when allocating the array
of struct iovec for store commands.

Change-Id: I0777ca8ef050b9773dba0c03b5e8533d3b2d5486
Reviewed-on: https://gerrit.stanford.edu/1560
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoReturn the name of the ACL instead of the numeric ID
Russ Allbery [Wed, 16 Jul 2014 05:57:31 +0000 (22:57 -0700)]
Return the name of the ACL instead of the numeric ID

The owner and getacl commands now return the current name of the ACL
instead of its numeric ID, matching the documentation of owner.

Change-Id: Ic47aad48bd1454ed4bffff7030b0492d74eee4fa
Reviewed-on: https://gerrit.stanford.edu/1559
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoFix ordering of table drops on wallet-admin destroy
Russ Allbery [Wed, 16 Jul 2014 05:07:48 +0000 (22:07 -0700)]
Fix ordering of table drops on wallet-admin destroy

Fix the ordering of table drops during a wallet-admin destroy action
to remove tables with foreign key references before the tables they
are referencing.  Should fix destroy in MySQL and other database
engines that enforce referential integrity.

Change-Id: I9b37c516f67acdf1d9e25222f067df6749e8c769
Reviewed-on: https://gerrit.stanford.edu/1558
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoUpdate TODO for more Commerzbank contributions, other changes
Russ Allbery [Wed, 16 Jul 2014 04:51:09 +0000 (21:51 -0700)]
Update TODO for more Commerzbank contributions, other changes

Change-Id: I478ed7812a4d25641ee85846e4092e17536e5a1d
Reviewed-on: https://gerrit.stanford.edu/1557
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoFix wallet-backend parsing of expires
Russ Allbery [Wed, 16 Jul 2014 03:48:35 +0000 (20:48 -0700)]
Fix wallet-backend parsing of expires

Fix wallet-backend parsing of the expires command to expect only one
argument as the expiration.  This was correctly documented in the
wallet client man page, but not in wallet-backend, and it accepted two
arguments (a date and time).  However, Wallet::Server did not and
would just ignore the time.  Now wallet-backend correctly requires the
date and time be passed as a single argument.

Change-Id: I8e51a576ea8781502f4eb983462ceca867b002be
Reviewed-on: https://gerrit.stanford.edu/1556
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoUse DateTime objects uniformly, improve expires parsing
Russ Allbery [Wed, 16 Jul 2014 03:46:57 +0000 (20:46 -0700)]
Use DateTime objects uniformly, improve expires parsing

Always use DateTime objects for every date field in the database,
and translate them into the local time zone for display when
pulling them out of the database.  This should provide better
portability to different database backends.

Change the parsing of expires arguments to use Date::Parse, thus
supporting a much broader variety of possible date and time
formats and allowing easy conversion to a DateTime object.
Document the new dependency.

Change-Id: I2ee8eaa6aa6ae9925ac419e49234ec9880d4fe95
Reviewed-on: https://gerrit.stanford.edu/1555
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoRecord ACL names in the ACL history
Russ Allbery [Wed, 16 Jul 2014 03:33:18 +0000 (20:33 -0700)]
Record ACL names in the ACL history

Change-Id: I0d7a088bb34dda2fc554b9f104c2a33e5faf879e
Reviewed-on: https://gerrit.stanford.edu/1554
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoRecord the ACL name in the acl_history table
Russ Allbery [Wed, 16 Jul 2014 03:29:19 +0000 (20:29 -0700)]
Record the ACL name in the acl_history table

Store the current name of the ACL with each history row, and index
the name.  This will eventually allow retrieval of history by name
for ACLs that have been deleted, although the rest of the code is
not yet in place.

The initial creation and membership of the ADMIN ACL during database
initialization or reinitialization is no longer recorded in the
acl_history table, since otherwise it produces errors due to the
missing ah_name field when building the database with schema 0.07.
There should be some better solution to this, but this will be okay
for the time being.

Change-Id: I015a00c972e0c2730c3d449952fcfe9b79c6e54f
Reviewed-on: https://gerrit.stanford.edu/1553
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoAdjust create-ddl for the move of modules to lib
Russ Allbery [Wed, 16 Jul 2014 03:23:50 +0000 (20:23 -0700)]
Adjust create-ddl for the move of modules to lib

Now needs a use lib 'lib' in order to run out of the source
directory.

Change-Id: Ia8645eae6c6699db919968d42f057b06e42150a2
Reviewed-on: https://gerrit.stanford.edu/1552
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoUse DateTime objects in the database layer, not strings
Russ Allbery [Tue, 15 Jul 2014 23:50:13 +0000 (16:50 -0700)]
Use DateTime objects in the database layer, not strings

Pass in DateTime objects for the date fields in the database instead
of formatted time strings.  This provides better compatibility with
different database engines.  Document in README the need to install
the DateTime::Format::* module corresponding to the DBD::* module used
for the server database.

Change-Id: Id25796da718d734ac96ca27ccea9045b0c80c03f
Reviewed-on: https://gerrit.stanford.edu/1551
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoDrop use POSIX from several tests that don't use it
Russ Allbery [Tue, 15 Jul 2014 23:49:20 +0000 (16:49 -0700)]
Drop use POSIX from several tests that don't use it

Change-Id: I61379e8963569d26c9b9c31d1727f3cca4567f8e
Reviewed-on: https://gerrit.stanford.edu/1550
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoAlso drop the new duo table when destroying the database
Russ Allbery [Tue, 15 Jul 2014 23:44:44 +0000 (16:44 -0700)]
Also drop the new duo table when destroying the database

In Wallet::Admin, add duo to the list of tables to drop when
either destroying or reinitializing the database.

Change-Id: I78790927f7d53b8d596e6ccb7c2340a341e404ae
Reviewed-on: https://gerrit.stanford.edu/1549
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoDocument enctype reference table population in NEWS
Russ Allbery [Tue, 15 Jul 2014 23:09:08 +0000 (16:09 -0700)]
Document enctype reference table population in NEWS

Change-Id: I3b3a035817e7e8f1c0e9709505490ce0ec299f3d
Reviewed-on: https://gerrit.stanford.edu/1548
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoAdd contrib wallet-rekey-periodic script
Russ Allbery [Tue, 15 Jul 2014 22:29:09 +0000 (15:29 -0700)]
Add contrib wallet-rekey-periodic script

Add a new contrib script, wallet-rekey-periodic, which is used at
Stanford to periodically rekey hosts from cron.

Change-Id: Ic1f515da44e55623f7d6864f9a3cebf24c08e13b
Reviewed-on: https://gerrit.stanford.edu/1547
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoTest for Perl strict and minimum version
Russ Allbery [Sat, 12 Jul 2014 05:36:11 +0000 (22:36 -0700)]
Test for Perl strict and minimum version

Fix strictness issues across the whole code base, and ensure that
all Perl scripts enable warnings.  (Hopefully enabling warnings
won't cause problems for the server.)

Change-Id: I4dee49f7a6bcbeeee21d74bf61a1fd26514f832c
Reviewed-on: https://gerrit.stanford.edu/1532
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoUpdate Perl POD tests to rra-c-util 5.5
Russ Allbery [Sat, 12 Jul 2014 04:49:26 +0000 (21:49 -0700)]
Update Perl POD tests to rra-c-util 5.5

Now that the Perl directory has been reorganized, it's easier to
synchronize with the standard rra-c-util tests.

Change-Id: I97a03d06ff964edcc85ab8788af281b7bc321235
Reviewed-on: https://gerrit.stanford.edu/1531
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoSwitch to Module::Build for the Perl module
Russ Allbery [Sat, 12 Jul 2014 04:39:23 +0000 (21:39 -0700)]
Switch to Module::Build for the Perl module

The wallet server now requires Perl 5.8 or later (instead of 5.006 in
previous versions) and is now built with Module::Build instead of
ExtUtils::MakeMaker.  This should be transparent to anyone not working
with the source code, since Perl 5.8 was released in 2002, but
Module::Build is now required to build the wallet server.  It is
included in some versions of Perl, or can be installed separately from
CPAN, distribution packages, or other sources.

Also reorganize the test suite to use subdirectories.

Change-Id: Id06120ba2bad1ebbfee3d8a48ca2f25869463165
Reviewed-on: https://gerrit.stanford.edu/1530
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoUpdate to rra-c-util 5.5 and C TAP Harness 3.1
Russ Allbery [Sat, 12 Jul 2014 03:18:41 +0000 (20:18 -0700)]
Update to rra-c-util 5.5 and C TAP Harness 3.1

Update to rra-c-util 5.5:

* Use Lancaster Consensus environment variables to control tests.
* Use calloc or reallocarray for protection against integer overflows.
* Suppress warnings from Kerberos headers in non-system paths.
* Assume calloc initializes pointers to NULL.
* Assume free(NULL) is properly ignored.
* Improve error handling in xasprintf and xvasprintf.
* Check the return status of snprintf and vsnprintf properly.
* Preserve errno if snprintf fails in vasprintf replacement.

Update to C TAP Harness 3.1:

* Reopen standard input to /dev/null when running a test list.
* Don't leak extraneous file descriptors to tests.
* Suppress lazy plans and test summaries if the test failed with bail.
* runtests now treats the command line as a list of tests by default.
* The full test executable path can now be passed to runtests -o.
* Improved harness output for tests with lazy plans.
* Improved harness output to a terminal for some abort cases.
* Flush harness output after each test even when not on a terminal.

Change-Id: I05161eb3d3be49a98f7762e876cb114da0c84e9a
Reviewed-on: https://gerrit.stanford.edu/1529
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoShow the error message if verifier-netdb creation fails
Russ Allbery [Sat, 12 Jul 2014 03:04:23 +0000 (20:04 -0700)]
Show the error message if verifier-netdb creation fails

Change-Id: I1acd7894316fc96943b9eda5e3a9abb9d229646c
Reviewed-on: https://gerrit.stanford.edu/1528
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoList all SQL files separately in Makefile.am
Russ Allbery [Sat, 12 Jul 2014 02:55:07 +0000 (19:55 -0700)]
List all SQL files separately in Makefile.am

Otherwise, they won't be copied properly to the perl directory
when building out of tree.

Change-Id: Id178d1b58d14a3595f372e25744dbdc61a81ff34
Reviewed-on: https://gerrit.stanford.edu/1527
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoClean up foreign keys and indices for history tables
Russ Allbery [Sat, 12 Jul 2014 02:26:46 +0000 (19:26 -0700)]
Clean up foreign keys and indices for history tables

Previous versions had erroneous foreign key constraints between the
object history table and the objects table.  Remove those constraints,
and an incorrect linkage in the schema for the ACL history, and add
indices for the object type, name, and ACL instead.

Change-Id: Ie0ff2448caa82c7a533a1b9ff5c13029bb6ae4ef
Reviewed-on: https://gerrit.stanford.edu/1526
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoFix test cleanup code to run during global destruction
Russ Allbery [Sat, 12 Jul 2014 01:56:33 +0000 (18:56 -0700)]
Fix test cleanup code to run during global destruction

If we don't run the code to delete the wallet database very late,
destruction of the SQLite objects may recreate the database file.
Move the unlink to an END block to avoid this problem.

Change-Id: Ib5ec2cbd08ba24c99f3dfa1daa01488931c5958f
Reviewed-on: https://gerrit.stanford.edu/1525
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoAdd some missing stopwords in POD documentation
Russ Allbery [Sat, 12 Jul 2014 01:56:09 +0000 (18:56 -0700)]
Add some missing stopwords in POD documentation

Change-Id: Ieb29af8a3bcf61cdebb1d8b7b2a5b8868a3586dc
Reviewed-on: https://gerrit.stanford.edu/1524
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoUndo email address change for verifier test
Russ Allbery [Sat, 12 Jul 2014 01:24:09 +0000 (18:24 -0700)]
Undo email address change for verifier test

The global replacement of my email address broke this test as
well since I was using my address as part of the test.  Revert
to the previous email address, since it's just example data.

Change-Id: Iaebcf4b4556308d5a69fbe2c8b59b05304dfbf66
Reviewed-on: https://gerrit.stanford.edu/1523
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoFix verifier-netdb test for Heimdal, email address
Russ Allbery [Sat, 12 Jul 2014 01:22:39 +0000 (18:22 -0700)]
Fix verifier-netdb test for Heimdal, email address

Undo the email address renaming for verifier-netdb as well, since
this has to use rra@stanford.edu for right now.  Adjust so that the
test is not skipped with Heimdal Kerberos user space.

Change-Id: I0a0e16e1044fe6057a310cc419477a34052ffcda
Reviewed-on: https://gerrit.stanford.edu/1522
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoFix verifier-ldap-attr test from email address change
Russ Allbery [Sat, 12 Jul 2014 01:20:46 +0000 (18:20 -0700)]
Fix verifier-ldap-attr test from email address change

This has to use rra@stanford.edu to work at all (and badly needs
some sort of mock LDAP environment rather than what it's doing now).

Change-Id: I0961c7f97633eb7e29e391804a6803195564d74b
Reviewed-on: https://gerrit.stanford.edu/1521
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoAdjust Heimdal keytab object tests to not compare keytabs
Russ Allbery [Sat, 12 Jul 2014 01:12:21 +0000 (18:12 -0700)]
Adjust Heimdal keytab object tests to not compare keytabs

For some reason, two keytabs are comparing inequal even after
masking the timestamp but both keytabs work for authentication.
Stop doing a data comparison and instead attempt authentications
with both keytabs as a more reliable test.

Change-Id: I4bd0712d492b78f09e95ffbed3461d97613d9d0a
Reviewed-on: https://gerrit.stanford.edu/1520
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoAdjust Wallet::Admin test for new schema version
Russ Allbery [Sat, 12 Jul 2014 00:50:56 +0000 (17:50 -0700)]
Adjust Wallet::Admin test for new schema version

We now have a 0.09 schema, so test upgrading from 0.07 to 0.09
in two steps.

Change-Id: I0e7af4371ba78aa69a9b7be59239f10c86e1fade
Reviewed-on: https://gerrit.stanford.edu/1519
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoAdjust ACL test for new SQLite autoincrement behavior
Russ Allbery [Sat, 12 Jul 2014 00:43:28 +0000 (17:43 -0700)]
Adjust ACL test for new SQLite autoincrement behavior

Similar to server, the ID of the last ACL created may vary
depending on whether SQLite reuses the last autoincrement key
when the highest-numbered record is deleted.  Accept either
possibility.

Change-Id: I2dd3b3cbbdf32931820ff799ca06f751c37a4cbd
Reviewed-on: https://gerrit.stanford.edu/1518
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoAdjust server test for new SQLite autoincrement behavior
Russ Allbery [Sat, 12 Jul 2014 00:39:01 +0000 (17:39 -0700)]
Adjust server test for new SQLite autoincrement behavior

SQLite now, when the highest-numbered record with an autoincrement
key is deleted, will reuse that number instead of incrementing
further.  Adjust the test suite so that this ambiguity is never
encountered, since it's not part of what we're testing.

Change-Id: I08c0c1c3fae82556d0f016b95db2992bdded1775
Reviewed-on: https://gerrit.stanford.edu/1517
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoAdd new object type for Duo integrations
Russ Allbery [Sat, 12 Jul 2014 00:19:59 +0000 (17:19 -0700)]
Add new object type for Duo integrations

A new object type, duo (Wallet::Object::Duo), is now supported.  This
creates an integration with the Duo Security cloud multifactor
authentication service and allows retrieval of the integration key,
secret key, and admin hostname.  Currently, only UNIX integration
types are supported.  The Net::Duo Perl module is required to use this
object type.  New configuration settings are required as well; see
Wallet::Config for more information.  To enable this object type for
an existing wallet database, use wallet-admin to register the new
object.

Change-Id: I2c0dac75e81f526b34d6b509c4bdaecb43dd4a9d
Reviewed-on: https://gerrit.stanford.edu/1516
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
9 years agoIgnore new Perl MYMETA.json file
Russ Allbery [Fri, 11 Jul 2014 23:43:45 +0000 (16:43 -0700)]
Ignore new Perl MYMETA.json file

Change-Id: I64dcd9fc393edba60f3a8d17ff2b59f8d51e131f
Reviewed-on: https://gerrit.stanford.edu/1515
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
10 years agoAdded /cron principals to Stanford policy and summary
Jon Robertson [Tue, 27 May 2014 18:28:35 +0000 (11:28 -0700)]
Added /cron principals to Stanford policy and summary

cron principals were being rejected due to not being a part of the
Stanford Policy module.

Change-Id: Ic67a8e2bce8474431163b74d97c2bf1fb184a4b7
Reviewed-on: https://gerrit.stanford.edu/1488
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
10 years agoChange my email address to eagle@eyrie.org
Russ Allbery [Sun, 13 Apr 2014 23:36:36 +0000 (16:36 -0700)]
Change my email address to eagle@eyrie.org

Change-Id: I4c2b5d7c807d6c27dd18a3b92eef66d21287d21e
Reviewed-on: https://gerrit.stanford.edu/1481
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
10 years agoRandomize the password on Heimdal principal creation
Russ Allbery [Thu, 9 Jan 2014 02:16:21 +0000 (18:16 -0800)]
Randomize the password on Heimdal principal creation

When creating new principals in a Heimdal KDC, generate a long, random
password as the temporary password of the disabled principal before
randomizing keys.  This is necessary if password quality is being
enforced on create calls.  Since the principal is always inactive
until the keys have been randomized, the password should not need to
be secure (and indeed is not cryptographically random).

Change-Id: If519a82475bb0d387a19d16ef1e024b0da64779a
Reviewed-on: https://gerrit.stanford.edu/1374
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
10 years agoFix test skip count in the verifier-ldap-attr test
Russ Allbery [Tue, 7 Jan 2014 05:32:14 +0000 (21:32 -0800)]
Fix test skip count in the verifier-ldap-attr test

When the correct Kerberos tickets aren't available, this test
skipped the wrong number of tests.

Change-Id: Icf27178fe88027f38764285bb671560e051f9105
Reviewed-on: https://gerrit.stanford.edu/1373
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
10 years agoAdd additional stopwords for POD spelling tests
Russ Allbery [Tue, 7 Jan 2014 05:11:08 +0000 (21:11 -0800)]
Add additional stopwords for POD spelling tests

Change-Id: I7eefcb5eab2e35e8d45baa6e868f1f00867c6b62
Reviewed-on: https://gerrit.stanford.edu/1372
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
10 years agoUpdate Test::RRA modules from the current rra-c-util
Russ Allbery [Tue, 7 Jan 2014 05:09:55 +0000 (21:09 -0800)]
Update Test::RRA modules from the current rra-c-util

Further documentation of the changes will come later when the rest
of rra-c-util files have been updated and the package makes more
use of these modules, but this fixes a spelling error test failure.

Change-Id: Ia885c4ab103235a1f6a2bf2b86d5a32c93751d89
Reviewed-on: https://gerrit.stanford.edu/1371
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
10 years agoAdd AM_PROG_AR to configure.ac for newer Automake
Russ Allbery [Tue, 7 Jan 2014 05:09:35 +0000 (21:09 -0800)]
Add AM_PROG_AR to configure.ac for newer Automake

Change-Id: I422a014c3a2611324c2cdebd364d81f2e91aadd7
Reviewed-on: https://gerrit.stanford.edu/1370
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
10 years agoFix wallet-rekey on keytabs containing multiple principals
Russ Allbery [Tue, 7 Jan 2014 05:09:00 +0000 (21:09 -0800)]
Fix wallet-rekey on keytabs containing multiple principals

Fix wallet-rekey on keytabs containing multiple principals.  Previous
versions assumed one could concatenate keytab files together to make a
valid keytab file, which doesn't work with some Kerberos libraries.
This caused new keys downloaded for principals after the first to be
discarded.  As a side effect of this fix, wallet-rekey always appends
new keys directly to the existing keytab file, and never creates a
backup copy of that file.

Change-Id: I5f863239ce4ebba66b35ff09454f2897367bd359
Reviewed-on: https://gerrit.stanford.edu/1369
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
10 years agoFlesh out TODO with lots more work
Russ Allbery [Sat, 2 Nov 2013 00:11:13 +0000 (17:11 -0700)]
Flesh out TODO with lots more work

Change-Id: I8f63cfd9692039f37ecfd46ab6072aa2f71c344d
Reviewed-on: https://gerrit.stanford.edu/1328
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
10 years agoACL.pm: Fix a place where the acl history was getting raw timestamp
Jon Robertson [Thu, 17 Oct 2013 05:37:20 +0000 (22:37 -0700)]
ACL.pm: Fix a place where the acl history was getting raw timestamp

The acl_history table needed to get the DateTime object rather than the
raw epoch timestamp in one place.  This was causing errors adding new
lines to the history.

Change-Id: I9c971819484cd0b26cb2561549246c284afc55a1
Reviewed-on: https://gerrit.stanford.edu/1325
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
10 years agoChanged postgres schema file to remove reference
Jon Robertson [Thu, 17 Oct 2013 05:35:34 +0000 (22:35 -0700)]
Changed postgres schema file to remove reference

The reference from object_history to the objects table needed to be
removed.  We still want the relationship in the DBIx::Class files, but
we don't want the relationship enforced as we want to keep history
entries for deleted objects.

Change-Id: Id927404b996fe171a8f5fc0747ccb0abddcbe1f2
Reviewed-on: https://gerrit.stanford.edu/1324
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
10 years agoClarify use of the tivoli-key file type in Stanford policy
Russ Allbery [Sun, 15 Sep 2013 00:42:50 +0000 (17:42 -0700)]
Clarify use of the tivoli-key file type in Stanford policy

tivoli-key used to always contain the entirety of the TSM.PWD file,
but it's more useful to store only the encryption key in password
form.

Change-Id: Id770691fb756b7675ec0fe2eee1308a8974c9c3f
Reviewed-on: https://gerrit.stanford.edu/1309
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoAdd use Wallet::Config to Wallet::ACL::LDAP::Attribute
Russ Allbery [Tue, 28 May 2013 22:56:22 +0000 (15:56 -0700)]
Add use Wallet::Config to Wallet::ACL::LDAP::Attribute

Normally this will already be loaded, but since we use it directly,
make sure it is loaded.

Change-Id: Ibc4ca874b659f316268957cbf77ead9d49bc3ca0
Reviewed-on: https://gerrit.stanford.edu/1205
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoFix documentation of ldap_map_principal hook
Russ Allbery [Tue, 28 May 2013 22:55:39 +0000 (15:55 -0700)]
Fix documentation of ldap_map_principal hook

Fix the Wallet::Config documentation for the ldap-attr verifier to
reference an ldap_map_principal hook, not ldap_map_attribute, matching
the implementation.

Change-Id: I258edcf69d4dcb3d2ec8dc66db4b768d91645fc4
Reviewed-on: https://gerrit.stanford.edu/1204
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoFix several bugs in Wallet::Object::Keytab enctype handling
Russ Allbery [Tue, 16 Apr 2013 03:47:41 +0000 (20:47 -0700)]
Fix several bugs in Wallet::Object::Keytab enctype handling

There was a missing resultset() call in one place and the wrong
resultset used in a different place, causing the enctype management
code to not work.

Change-Id: I796169c5968ec164f90f3cd75541dd346dd50fdf
Reviewed-on: https://gerrit.stanford.edu/1070
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoImprove the keytab enctype test suite
Russ Allbery [Tue, 16 Apr 2013 03:46:42 +0000 (20:46 -0700)]
Improve the keytab enctype test suite

Fix recognition of the syntax error from Heimdal's klist -ke, which
doesn't exit with status 1.  Assume that if we didn't see any known
enctypes, we're dealing with Heimdal.  Remove the code to populate the
enctype table, since we do that in Wallet::Admin now.  Show the error
if adding an enctype fails.

Change-Id: I40da967ef6868e5cd51a1238e1c6324386468134
Reviewed-on: https://gerrit.stanford.edu/1069
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoPopulate the enctype table by default on new install
Russ Allbery [Tue, 16 Apr 2013 03:45:46 +0000 (20:45 -0700)]
Populate the enctype table by default on new install

We actually know the enctypes that are in most common use, so rather
than making the user poke them into the database manually, save them
a step and put them in.  We still need some mechanism to remove the
DES enctype and add new ones, though.

Change-Id: I2eda7e29897ec16a04a10f0c7289878c853b7531
Reviewed-on: https://gerrit.stanford.edu/1068
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoAllow afs/* principals in the Stanford naming policy
Russ Allbery [Tue, 16 Apr 2013 01:45:12 +0000 (18:45 -0700)]
Allow afs/* principals in the Stanford naming policy

Change-Id: I59db1e8638a602d9c118ac01da17280c9ed7d005
Reviewed-on: https://gerrit.stanford.edu/1067
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoFurther clarify the ssl-key Stanford naming policy
Russ Allbery [Sat, 13 Apr 2013 02:37:20 +0000 (19:37 -0700)]
Further clarify the ssl-key Stanford naming policy

Adam requested some clarification on whether the name of the object
should be fully-qualified or not (since we didn't in the legacy
naming scheme).

Change-Id: I52fcab71e54aee38f0c03eff774f927c5836ad03
Reviewed-on: https://gerrit.stanford.edu/1054
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoUpdate stanford.conf example to use Wallet::Policy::Stanford
Russ Allbery [Mon, 8 Apr 2013 20:02:04 +0000 (13:02 -0700)]
Update stanford.conf example to use Wallet::Policy::Stanford

Eliminate all the verification code that moved into the policy
object.  Update coding style and remove some settings that were
no longer used at Stanford.

Change-Id: I3a098bc318abe4bc9dd82e86186da012e5c2cd27
Reviewed-on: https://gerrit.stanford.edu/1025
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoUpdate stanford.conf to latest production version
Russ Allbery [Mon, 8 Apr 2013 19:56:04 +0000 (12:56 -0700)]
Update stanford.conf to latest production version

Update the example wallet.conf from Stanford's configuration to our
latest production version.

Change-Id: Ic652b7a2fadb53a688a0c0c16b5ea7e429cff79e
Reviewed-on: https://gerrit.stanford.edu/1024
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoUpdate the wallet ACL for its-apps in Stanford policy
Russ Allbery [Mon, 8 Apr 2013 19:50:24 +0000 (12:50 -0700)]
Update the wallet ACL for its-apps in Stanford policy

Change-Id: Ie4c0af7b218fcd00fc6f39514967c9e38e12b1be
Reviewed-on: https://gerrit.stanford.edu/1023
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoMake t/admin.t test robust against missing sqlite3 binary
Russ Allbery [Fri, 29 Mar 2013 20:56:45 +0000 (13:56 -0700)]
Make t/admin.t test robust against missing sqlite3 binary

If we can't find the sqlite3 binary, just skip the upgrade test.

Change-Id: I20e9ad1978a189cf059756e15794ea4d954f3867
Reviewed-on: https://gerrit.stanford.edu/994
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoadmin.t: Fixed problem with not upgrading from unversioned db
Jon Robertson [Fri, 29 Mar 2013 06:53:52 +0000 (23:53 -0700)]
admin.t: Fixed problem with not upgrading from unversioned db

Since we were reinstalling a fresh database via the same DBIx::Class
functions, the database we installed to upgrade from a non-versioned
setup was still getting a version table.  Switched to delete the
database and reload it fresh from the sqlite3 command itself.

Change-Id: Ia09bbc279ab834b5d17453b4282e18dd3a36f857
Reviewed-on: https://gerrit.stanford.edu/993
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoUpdate the comment in README about supported ACL types
Russ Allbery [Thu, 28 Mar 2013 03:34:08 +0000 (20:34 -0700)]
Update the comment in README about supported ACL types

Change-Id: I171722d03bebc812bb113d1366457d75a94e468a
Reviewed-on: https://gerrit.stanford.edu/991
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoFix return status of Wallet::Schema release/1.0
Russ Allbery [Thu, 28 Mar 2013 02:16:15 +0000 (19:16 -0700)]
Fix return status of Wallet::Schema

The module wasn't always returning a true value when loaded.

Change-Id: I998ab25509cb9079034cae6aca467024ec6b4949
Reviewed-on: https://gerrit.stanford.edu/990
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoAccept any characters in the argument to the comment command
Russ Allbery [Thu, 28 Mar 2013 01:23:27 +0000 (18:23 -0700)]
Accept any characters in the argument to the comment command

It's nice to have spaces and other special characters in comments,
so allow any character rather than applying the normal argument
filtering.

Change-Id: Iec8584f1f6893906db7245fbe571d62ebc60f72a
Reviewed-on: https://gerrit.stanford.edu/989
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoAdd NEWS entry and doc reference to Wallet::Policy::Stanford
Russ Allbery [Thu, 28 Mar 2013 00:27:39 +0000 (17:27 -0700)]
Add NEWS entry and doc reference to Wallet::Policy::Stanford

Change-Id: If833e4a6434362e04e738274a6f7fb276a9efe51
Reviewed-on: https://gerrit.stanford.edu/988
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoSkip the wa-keyring tests if WebAuth is too old
Russ Allbery [Thu, 28 Mar 2013 00:23:45 +0000 (17:23 -0700)]
Skip the wa-keyring tests if WebAuth is too old

If the WebAuth module is too old, don't just fail the tests.  Skip
them instead.  This will let the Debian package build in unstable.

Change-Id: I84c97f23ff7fbf89f2fd797898ebb4ab5e58eee6
Reviewed-on: https://gerrit.stanford.edu/987
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoRegister the wa-keyring object handler by default
Russ Allbery [Wed, 27 Mar 2013 23:14:40 +0000 (16:14 -0700)]
Register the wa-keyring object handler by default

In Wallet::Admin, add the wa-keyring object handler to the list
of initializations when creating a new database.

Change-Id: I804b47ae712ce3d96c57699fb2ba05c45f687881
Reviewed-on: https://gerrit.stanford.edu/986
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>
11 years agoDocument the new WebAuth keyring object type
Russ Allbery [Wed, 27 Mar 2013 23:13:24 +0000 (16:13 -0700)]
Document the new WebAuth keyring object type

Add a mention to NEWS and to the REQUIREMENTS section of README.

Change-Id: I560f737e9cb899046f7fe3c8d2c8c648d31041e7
Reviewed-on: https://gerrit.stanford.edu/985
Reviewed-by: Russ Allbery <rra@stanford.edu>
Tested-by: Russ Allbery <rra@stanford.edu>