PGXN

Isok -- A query centered monitoring tool for PostgreSQL

    pgxn install pg_isok

    CREATE SCHEMA isok;
    CREATE EXTENSION pg_isok SCHEMA isok;

Note

Any rebuild of Isok requires the installation of the m4 macro pre-processor.^[2] Your operating system almost surely makes available an m4 package.

Only the "pre-built" PGXN distribution can be installed without the use of m4.

If desired, some potentially dangerous features of Isok can be disabled at build time.

These are the make variables that control the build options:

To use these variables, set them to any value when running make. For example, to disable all optional features run:

The build configuration is documented in the doc/pg_isok--${VERSION}.config file, and installed with the rest of the documentation.

If you are running in the cloud, or some other managed instance where you do not have permissions on the host running PostgreSQL, you will not be able to install Isok as an extension. In these cases you can still install Isok, but you must first build its SQL and then manually execute it.

Of course, this installation method can always be used, as there is always a way to execute SQL.

To build a “cloud version” of Isok, suitable for installation by SQL execution, you would type something like:

The resulting sql file is in the sql/ directory.

To customize the build, any of the above variables may also be set. The TARGET_SCHEMA variable must be set; the objects produced by the generated SQL must be located within a designated schema. It is highly recommended that the TARGET_SCHEMA be lower-case and otherwise be a PostgreSQL name which does not require quoting.

To install, first create the schema and then execute the sql. The command line interaction, if you use the psql command line client interface, would look something like:

$ psql
psql (15.13 (Debian 15.13-0+deb12u1))
Type "help" for help.

me=> CREATE SCHEMA isok;   -- The TARGET_SCHEMA used to build the sql
CREATE SCHEMA
me=> \i sql/pg_isok_cloud--1.0.0.sql
<lots of output redacted>
me=> \q
$

You must re-build different SQL, with a different TARGET_SCHEMA, in order to install into a different schema.

With appropriate OS-level permissions, run:

With this step complete, you are ready to install the Isok extension into any schema of any database in the cluster.

Once an extension has been installed in the OS, regression tests can be run to test whether Isok is operating correctly. Running the regression tests when Isok is installed by SQL execution is unsupported.

The same build variables must be set when running the regression tests as when the system was built. (The PGXN distribution sets no variables, the default.) Should you set a different collection of variables than when building, some tests will fail and others may fail to run at all.

The following example runs the default set of regression tests:

    CREATE SCHEMA isok;
    CREATE EXTENSION pg_isok SCHEMA isok;

Uninstalling from the OS does the opposite of installing. It removes the extension from the PostgreSQL server's filesystem.

To uninstall with pgxnclient, run:

To uninstall using make, run:

Because Isok is pure SQL, uninstalling it from the OS does not remove any functionality from existing instances installed with CREATE EXTENSION. Uninstalling does, however, remove the ability to use the CREATE EXTENSION to install Isok in a schema.

Running:

removes the extension from all schemas in all PostgreSQL clusters.

To remove an installation of Isok from an individual schema, drop the schema with DROP SCHEMA schemaname CASCADE;.

Tip

Use PostgreSQL's dollar quoting when inserting queries into ISOK_QUERIES using INSERT statements. This avoids problems that would otherwise arise involving the use of quote characters inside quoted strings.

A TEXT value. A unique name for the query. The IQName value cannot be changed. This column may not be empty; it must contain characters, and it must contain at least one non-whitespace character. This column may not be NULL. This column may not contain whitespace characters. This column must be unique when compared in a case-insensitive fashion.

A BOOLEAN value. TRUE when the query finds conditions that are errors, FALSE when the query finds conditions that are warnings. See ISOK_RESULTS (and the Introduction to Isok) for more on warnings and errors.

This column may not be NULL.

A TEXT value. Code classifying the query. The legal values for this column are defined by the IQ_TYPES support table.

This column may not be NULL.

A timestamp. Date and time the query was first run by Isok. The value of this column is NULL if the query has never been run.

A timestamp. Date and time the query was most recently run by Isok. The value of this column is NULL if the query has never been run.

A BOOLEAN value. This column controls the value placed in the ISOK_RESULTS.Keep_Until column when run_isok_queries() inserts new rows in ISOK_RESULTS.

When this column is TRUE, each row returned by the query is stored in ISOK_RESULTS with a Keep_Until value of infinity. This prevents run_isok_queries() from deleting the query result row when run, when the query no longer returns the result row.

When this column is FALSE, the ISOK_RESULTS.Keep_Until value of any new rows that run_isok_queries() inserts is NULL.

This column may not be NULL.

A PostgreSQL name value. The PostgreSQL role to use to run the query.

Because different roles have differing access to database content, it can be useful to run queries with different roles in effect.

This column is not validated against existing roles.

Note that the name data type casts (transparently) to TEXT.

When this column is NULL, the effective role is not changed.

A TEXT value. The PostgreSQL schema search_path to have in effect when the query is run.

The syntax of the search path is that used by SET search_path ... and returned by SHOW search_path;.

Because queries may not always contain schema names to qualify database objects, a single query can return different results depending on the search_path in effect. So it can be useful to run different queries when different schema search paths are in effect.

Care must be taken when setting the search path because the search path can be set to anything, regardless of which schemas exist or are available to the user. It is quite easy to set a search path that searches no schemas. PostgreSQL will not produce any warnings or errors should you do so.

When this column is NULL, the schema search path is not changed.

A TEXT value. A query which checks for database integrity violations. The query need not end in a semi-colon. The query must return 3 columns. Although these columns are referred to by name below, the names the query gives to the columns does not matter.

A TEXT value. A comment on the query. This may be as verbose as necessary. This column may not be NULL. This column may be empty; it need not contain characters, but it may not contain only whitespace characters..

An BIGINT value This column uniquely identifies the row containing the result of a database integrity query. The IRID value cannot be changed and is automatically generated with a PostgreSQL sequence.

A TEXT value. The ISOK_QUERIES.IQName value identifying the query which produced the result.

A timestamp value. Date and time the query result was first produced by Isok. This column may not be NULL.

A timestamp value. Date and time the query result was most recently produced by Isok. This column may not be NULL.

A PostgreSQL name value. The role (user) which was the current role when the query was last executed. Note that the name data type casts (transparently) to TEXT.

This column is not validated against existing roles.

This column may not be NULL.

An array of PostgreSQL name values. All schemas that were, implicitly or not, in the search_path, and also available to the Last_Role, when the result was returned. For more information, see the documentation of the current_schemas() function.

Note that the name data type casts (transparently) to TEXT. This column may not be NULL.

A timestamp value. Date and time the query result was resolved; that is, marked not a concern. The Isok system does not display resolved results, although of course the ISOK_RESULTS table can always be manually queried.

The value of this column may be NULL. This occurs both when the query result is a data integrity error and when it is a data integrity warning that has not yet been resolved.

A timestamp value. Isok suppresses display of the result when the current time is before this time. Use of this column allows resolution of data integrity problems to be deferred, and hence not clutter up the output of Isok with noise that might hide other problems.

When this column is NULL Isok displays the query result.

A TEXT value. Code classifying the query result. The legal values for this column are defined by the IR_TYPES support table.

This column may be NULL when the query result is unclassified.

A timestamp value. This column controls whether or not run_isok_queries() deletes the row when the ISOK_QUERIES.Query is re-run and the query does not return the row's QR_ID.

A query result that the query no longer returns is kept until the given time is reached, when the value of this column is not NULL. When the value of this column is NULL, a query result row that is no longer returned is is always deleted. For further detail see the Deletion of Old Results section of the run_isok_queries() documentation.

A TEXT value. This is a unique, unique per query that is, identifier for the query result. It is the first column produced by the related ISOK_QUERIES.Query.

This column may not be NULL.

A TEXT value. This is the message, the second column, produced by the most recent execution of the ISOK_QUERIES.Query.

A JSONB value. The value of the third, optional, column returned by most recent execution of the query. This may contain any JSON deemed useful. This column serves as a catch-all container for any additional data that needs to be tracked regarding a reported condition.

The value of this column may be NULL. This is the default when the ISOK_QUERIES.Query does not return a third column.

See PostgreSQL's documentation on the JSON data types for information on how to access, index, and efficiently query the JSONB data type.

A TEXT value. Any notes regarding this particular query result. This column may not be NULL. This column may be empty; it need not contain characters, but it may not contain only whitespace characters.

The IQ_TYPES table is keyed by the IQType column. This column may not contain whitespace characters. This column must be unique when compared in a case-insensitive fashion.

The IR_TYPES table is keyed by the IRType column. This column may not contain whitespace characters. This column must be unique when compared in a case-insensitive fashion.

Limiting access to Isok is a clear first-step. Installing Isok into a dedicated schema goes a long way toward helping with this. When a schema is created, only the owner has access.^[6]So, unless GRANTs are issued, access is limited by default.

Remember also, the ISOK_RESULTS table contains query output that may contain sensitive information to which access should be restricted. And, even if this is not true today, it may become true when additional queries are added to ISOK_QUERIES.

Even the queries in ISOK_QUERIES could, possibly, contain sensitive information.

The executed queries, the ISOK_QUERIES.Querys, can be any SQL statement. Obviously, what executes matters. Less obviously, the ownership of and permissions granted on every object referenced by every query also matters.

Really, when multiple schemas are in the search_path, it is the ownership of and permissions granted on every object that might be referenced by every query that matters.

The ownership and permissions of referenced objects matter because these factors ultimately control what any given query actually does. If a user has, for example, permission to alter a view with some given name, or replace a table having that name with a view that has the same name, then the user can change what happens when that name is used in a query. The user can write a view that does anything. Or at least anything that the role which runs run_isok_queries() is allowed to do.

Imagine, the new view could call a function, say, in place of a table that was referenced, and that function could do anything at all. Even while still returning the replaced table's rows, so as to produce a results identical to that produced before the system was altered.

That is the issue. The user executing the saved query is dependent upon the goodwill of all the users who have enough access to alter any of the objects involved when the query is executed.

The ISOK_QUERIES.Search_Path column allows setting of the search_path on a per-query basis. The security implications of changing the search_path may be the hardest to reason through. The crux of the problem is that different users may have different permissions on the search path's schemas, and on the objects the schemas contain. This opens up the possibility that a malicious user may create an object, say, a view or a function, in a schema which appears earlier in the search path than the schema holding the object the query expects to find. If this is the case, the query will use the malicious object instead of the expected object.

The PostgreSQL documentation contains an analysis of this situation, in the context of writing SECURITY DEFINER functions. However, the analysis in the PostgreSQL documentation is not entirely applicable to Isok. In the case of Isok, even when Isok changes the effective role, the position in the search path of the temporary table schema, pg_temp, is less relevant. Because temporary tables are not shared between connections, the creation of a malicious object in the temporary schema must be done in the current connection. And so the issue is no different from that which occurs when any other malicious object is created in the current connection. In either case, there is a security lapse that occurs dynamically, at some point in the current connection.

Having said that, moving pg_temp to the end of the search path does make it harder to “mask” an existing object with a malicious object. Because all roles have permission to create objects in pg_temp, a malicious actor would not be able to mask an existing object with an object in pg_temp if pg_temp is at the end of the search path. For this reason it may make sense to always put pg_temp at the end of the search path whenever Isok is used.

The PostgreSQL documentation's observation remains valid: Malicious users with the ability to change objects in the search path may inject malicious objects.

The role in effect does have security implications. But changing a role for the duration of a query's execution, with ISOK_QUERIES.Role, has fewer security implications than it might seem.

Changing the current role does open up the possibility that database objects to which the new role has access may be changed. But this door is already open. A new role cannot be assumed without some chain of SET option grants from the session_user [definition here(-ish)] to the current role. So a malicious actor always has access to the same set of roles, regardless of whether Isok is involved or not.

What might be surprising is that, even though a role may SET ROLE to another, perhaps with less privileges, it is always possible to use RESET ROLE (or SET ROLE NONE) and reset the current role to the session_role. There is no sandboxing. If the session sets a role before running run_isok_queries(), there is the possibility that a malicious actor might undo the assumption of the role. This could then affect the role used to execute any queries that run_isok_queries() has not yet executed.

Don't expect that a SET ROLE to a role of lesser privileges makes running run_isok_queries() any safer.

There is no one-size-fits-all solution. Even disabling Isok's ability to dynamically alter the current search path and the current role does not address the fundamental issues. Even more so because, to be useful, run_isok_queries() may need an expansive set of permissions to do its job.

One possible strategy is to always supply values in the ISOK_QUERIES.Role ISOK_QUERIES.Search_Path columns. At least that way the context of each query's execution is always known.

Another possible strategy is to install Isok in multiple schemas, each schema dedicated to a different purpose and assigned different permissions, intended to be used by different users.

To better respond to a suspected security problem it is always very useful to have an audit trail to examine. One way to have such a trail is to install a temporal extension. These extensions track the history of database content over time. The Isok tables could be temporally tracked, to audit what queries were changed when, as well as what query results were produced or deleted when.

A conceivable, although entirely untested on our part, idea is to use a temporal extension to track changes made to the postgres database. Otherwise known as the system catalog, pg_catalog, this database contains the definitions of all objects in all databases. Tracking the catalog provides an audit trail should a malicious object be created, although this would not help if pg_temp was involved.

Some installations may even want to temporally track all their tables, although this may not be feasible for a whole host of reasons.

The files shown below deliver an Isok report, if there is something to report, by email every Tuesday morning.

The system on which they are installed must have a mail transfer agent installed, like Postfix, to begin the email delivery process. The system also must have GNU mailutils installed, or an equivalent mail command, like BSD mailx, to send the email.

Most operating systems will have packages available to install these services, and a way to configure simple defaults. However, it is non-trivial to reliably deliver email from your system directly to the rest of the Internet. The recommended approach is to send the email from your local system to a mail relay provided by your local IT professionals. (Or, your Internet Service Provider. Or, if you are hosted in the cloud, contact your hosting company.) These professionals will usually be able to supply you with what you need to know to have mail sent from your system to a system able to send email to the Internet at-large. If not, there are companies that provide this service for a nominal fee.

The service you would ask for is usually called an email relay service.

It is usually a good idea to ask your local IT professionals to help with the selection of a mail transfer agent.

This example is expected to run, as is, on most systems that have the default PostgreSQL install.

The example assumes that pg_hba.conf contains:

This line is typically present, but this is not guaranteed.

#!/usr/bin/bash
#
# Run pg_isok, and mail (with GNU mailutils) if it produces anything.
#
# Expected enviornment variables:
# CONNECTION_STRING
#    The postgres connection string (or other arguments)
#    Putting passwords in here is a bad idea, change pg_hba.conf instead?
# ISOK_SCHEMA
#    The schema in which pg_isok is installed

# The recommendation is to _not_ change this.  Instead, make an
# alias for "pg_isok_report" in /etc/aliases.
MAIL_RECIPIENT=pg_isok_report

EMPTY_FILE=$(/usr/bin/mktemp --tmpdir pg_isok_empty.XXXXXXXXXX)
OUTPUT=$(/usr/bin/mktemp --tmpdir pg_isok_output.XXXXXXXXXX)

PSQL="/usr/bin/psql ${CONNECTION_STRING}"

cleanup () {
  /usr/bin/rm -rf ${EMPTY_FILE} ${OUTPUT}
}
trap cleanup EXIT

PAGER= ${PSQL} --command="
              SELECT irid, iqname, first_seen, last_seen, last_role
                   , last_schemas, deferred_to, category, keep_until
                   , qr_id, qr_message, qr_extra, notes
                FROM ${ISOK_SCHEMA}.isok_results
                LIMIT 0;
  " \
  > ${EMPTY_FILE} 2>&1

PAGER= ${PSQL} --command="SELECT * FROM ${ISOK_SCHEMA}.run_isok_queries();" \
  > ${OUTPUT} 2>&1

cmp --quiet ${EMPTY_FILE} ${OUTPUT} \
  || { /usr/bin/mail -s 'Isok output' ${MAIL_RECIPIENT} \
         < ${OUTPUT} ; }

[Unit]
Description=Tuesday report from pg_isok

[Timer]
# See: man 7 systemd.time
OnCalendar=tuesday *-*-* 3:00
RandomizedDelaySec=60m
Persistent=true

[Install]
WantedBy=timers.target

[Unit]
Description=Run pg_isok's run_isok_queries() function and email when there's\
a result
ConditionACPower=true

[Service]
#
# Configuration is done here (and in /etc/aliases, see pg_isok_report)
#

# The postgres connection string (or other arguments to psql)
# Putting passwords in here is a bad idea, change pg_hba.conf instead?
Environment="CONNECTION_STRING=mydatabase"

# Put a connection string variable assignment containing secrets in this file:
# (man 5 systemd.exec)
#EnvironmentFile=/etc/pg_isok_secrets

# The schema in which pg_isok is installed
Environment="ISOK_SCHEMA=isok"

# End of configuration

# The Uni*x user running the db engine
# (Expected to be the same as the PostgreSQL bootstrap superuser)
User=postgres
Type=oneshot
KillMode=process

PassEnvironment=CONNECTION_STRING ISOK_SCHEMA
ExecStart=/usr/local/bin/pg_isok_report

To perform actions before or after execution of run_isok_queries(), write a new function that takes run_isok_queries()'s arguments and returns run_isok_queries()'s results. And does what you wish before or afterward.

So, for example, to ensure a safe, consistent, value for search_path, you could write a function that executes SET search_path ...;, before itself calling run_isok_queries() and returning the result.

If you would like additional ways to classify the issues your queries discover, the IR_TYPES table may be extended.

Create your own table to do this, called, say, IR_TYPE_CLASSES.

The key of this table is that of the IR_TYPES table; may as well call it IRType. It is a foreign key, referencing IR_TYPES. So your new table has a one-to-one relationship with IR_TYPES.

Add as many columns as you like to your new table, a column for each (orthogonal) sub-category by which you would like to classify reported issues. Boolean columns behave as a tag, toggling classification. Other kinds of columns, possibly containing foreign keys to control the vocabulary used, allow richer classification schemes.

Indexing the JSONB ISOK_RESULTS.QR_Extra column improves performance.

If you know your JSONB keys, you can make a VIEW that exposes the value of those keys as the view's columns. Users of this view would not have to be familiar with querying JSONB.

More complex schemes involve putting a row-level BEFORE trigger on ISOK_RESULTS to distribute the various values appearing in ISOK_RESULTS into other tables. But doing so surely takes you past the point of diminishing returns. It is easier to modify the SQL that Isok installs. And doing so is probably also less of a long-term maintenance burden, which matters.

Should you want to make a modification like allowing the queries in ISOK_QUERIES.Query to return additional columns, you can do so by modifying the SQL that Isok loads.

Isok is pure SQL, so the SQL can be generated and then modified, in any way you like, before being loaded into a database's schema.

Isok uses the PGXN.org tools for building and distribution, which in turn uses parts of the PostgreSQL GNU make-based build system. So GNU make is required. In addition, the Isok documentation is done with the DocBook, as is PostgreSQL's, so the tooling required to build the documentation is the same as PostgreSQL, although Isok currently generates XHTML so may require a slightly different set of DTDs.

Aside from these requirements, the following additional tools are needed:

DBLatex also requires the installation of various TeX and LaTeX tooling, which your O/S's package manager is likely to install as a dependency.

When working with DocBook, the book DocBook XSL: The Complete Guide from Sagehill.net may also be useful.

Run make help for help on the Makefile targets.

Almost all the generated files are included in the distribution. This is so that the user, or the PGXN tooling, can use the Makefile for installation, and uninstallation, without having to have all the tooling required for development installed.

Funding and Support. We gratefully acknowledge the support of The National Institutes of Health grant R00HD057992 for the development of this system. We are also very grateful for the support given by The Leo S. Guthman Foundation, the Lincoln Park Zoo, Franklin & Marshall College, and The George Washington University.

Other Thanks. We would like to thank the myriad Free and Open Source communities, including those of PostgreSQL, the GNU Project, the Debian Project, Ubuntu, PhpPgAdmin, the Pyramid web framework, TeX and LaTeX, DBLatex, DocBook, Babase, and many others unmentioned, for giving, gratis, billions of dollars^[7] of work to the world, without which the Gombe-MI software and this book would not exist.

Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Institutes of Health, The Leo S. Guthman Foundation, the Lincoln Park Zoo, Franklin & Marshall College, The George Washington University, or any other organization which has supplied support for this work.

We gratefully acknowledge the support of the National Science Foundation for the supporting the collection of the majority of the data stored in the database; in the past decade in particular we acknowledge support from IBN 9985910, IBN 0322613, IBN 0322781, BCS 0323553, BCS 0323596, DEB 0846286, DEB 0846532 and DEB 0919200. We are also very grateful for support from the National Institute of Aging (R01AG034513-01 and P01AG031719) and the Princeton Center for the Demography of Aging (P30AG024361). We also thank the Chicago Zoological Society, the Max Planck Institute for Demographic Research, the L.S.B. Leakey Foundation and the National Geographic Society for support at various times over the years. In addition, we thank the National Institute of Aging (R03-AG045459-01) for supporting recent work extending the database to incorporate genetic and genomic data.

Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation, the National Institute of Aging, the Princeton Center for the Demography of Aging, the Chicago Zoological Society, the Max Planck Institute for Demographic Research, the L.S.B. Leakey Foundation, the National Geographic Society, or any other organization which has supplied support for this work.

At the time of this writing, there is no formal set of acknowledgments for SokweDB.

However, Microsoft provided funding for SokweDB and we would like to acknowledge and thank them for their support.

Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of Microsoft.

1. The work must carry prominent notices stating that you modified it, and giving a relevant date.

2. The work must carry prominent notices stating that it is released under this License and any conditions added under section 7. This requirement modifies the requirement in section 4 to “keep intact all notices”.

3. You must license the entire work, as a whole, under this License to anyone who comes into possession of a copy. This License will therefore apply, along with any applicable section 7 additional terms, to the whole of the work, and all its parts, regardless of how they are packaged. This License gives no permission to license the work in any other way, but it does not invalidate such permission if you have separately received it.

4. If the work has interactive user interfaces, each must display Appropriate Legal Notices; however, if the Program has interactive interfaces that do not display Appropriate Legal Notices, your work need not make them do so.

1. Convey the object code in, or embodied in, a physical product (including a physical distribution medium), accompanied by the Corresponding Source fixed on a durable physical medium customarily used for software interchange.

2. Convey the object code in, or embodied in, a physical product (including a physical distribution medium), accompanied by a written offer, valid for at least three years and valid for as long as you offer spare parts or customer support for that product model, to give anyone who possesses the object code either (1) a copy of the Corresponding Source for all the software in the product that is covered by this License, on a durable physical medium customarily used for software interchange, for a price no more than your reasonable cost of physically performing this conveying of source, or (2) access to copy the Corresponding Source from a network server at no charge.

3. Convey individual copies of the object code with a copy of the written offer to provide the Corresponding Source. This alternative is allowed only occasionally and noncommercially, and only if you received the object code with such an offer, in accord with subsection 6b.

4. Convey the object code by offering access from a designated place (gratis or for a charge), and offer equivalent access to the Corresponding Source in the same way through the same place at no further charge. You need not require recipients to copy the Corresponding Source along with the object code. If the place to copy the object code is a network server, the Corresponding Source may be on a different server (operated by you or a third party) that supports equivalent copying facilities, provided you maintain clear directions next to the object code saying where to find the Corresponding Source. Regardless of what server hosts the Corresponding Source, you remain obligated to ensure that it is available for as long as needed to satisfy these requirements.

5. Convey the object code using peer-to-peer transmission, provided you inform other peers where the object code and Corresponding Source of the work are being offered to the general public at no charge under subsection 6d.

1. Disclaiming warranty or limiting liability differently from the terms of sections 15 and 16 of this License; or

2. Requiring preservation of specified reasonable legal notices or author attributions in that material or in the Appropriate Legal Notices displayed by works containing it; or

3. Prohibiting misrepresentation of the origin of that material, or requiring that modified versions of such material be marked in reasonable ways as different from the original version; or

4. Limiting the use for publicity purposes of names of licensors or authors of the material; or

5. Declining to grant rights under trademark law for use of some trade names, trademarks, or service marks; or

6. Requiring indemnification of licensors and authors of that material by anyone who conveys the material (or modified versions of it) with contractual assumptions of liability to the recipient, for any liability that these contractual assumptions directly impose on those licensors and authors.

The person who associated a work with this deed has dedicated the work to the public domain by waiving all of his or her rights to the work worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.

You can copy, modify, distribute and perform the work, even for commercial purposes, all without asking permission. See Other Information below.

In no way are the patent or trademark rights of any person affected by CC0, nor are the rights that other persons may have in the work or in how the work is used, such as publicity or privacy rights.^[8]

Unless expressly stated otherwise, the person who associated a work with this deed makes no warranties about the work, and disclaims liability for all uses of the work, to the fullest extent permitted by applicable law.

When using or citing the work, you should not imply endorsement by the author or the affirmer.^[9]

The Commons Deed is not a legal instrument. It is simply a handy reference for understanding the CC0 Legal Code, a human-readable expression of some of its key terms. Think of it as the user-friendly interface to the CC0 Legal Code beneath. This Deed itself has no legal value, and its contents do not appear in CC0.

Creative Commons is not a law firm and does not provide legal services. Distributing, displaying, or linking to this Commons Deed does not create an attorney-client relationship.

Creative Commons has not verified the copyright status of any work to which CC0 has been applied. CC makes no warranties about any work or its copyright status in any jurisdiction, and disclaims all liability for all uses of any work.

Creative Commons Legal Code

CC0 1.0 Universal

    CREATIVE COMMONS CORPORATION IS NOT A LAW FIRM AND DOES NOT
    PROVIDE LEGAL SERVICES. DISTRIBUTION OF THIS DOCUMENT DOES
    NOT CREATE AN ATTORNEY-CLIENT RELATIONSHIP. CREATIVE COMMONS
    PROVIDES THIS INFORMATION ON AN "AS-IS" BASIS. CREATIVE
    COMMONS MAKES NO WARRANTIES REGARDING THE USE OF THIS
    DOCUMENT OR THE INFORMATION OR WORKS PROVIDED HEREUNDER, AND
    DISCLAIMS LIABILITY FOR DAMAGES RESULTING FROM THE USE OF
    THIS DOCUMENT OR THE INFORMATION OR WORKS PROVIDED
    HEREUNDER.

Statement of Purpose

The laws of most jurisdictions throughout the world
automatically confer exclusive Copyright and Related Rights
(defined below) upon the creator and subsequent owner(s) (each
and all, an "owner") of an original work of authorship and/or a
database (each, a "Work").

Certain owners wish to permanently relinquish those rights to a
Work for the purpose of contributing to a commons of creative,
cultural and scientific works ("Commons") that the public can
reliably and without fear of later claims of infringement build
upon, modify, incorporate in other works, reuse and redistribute
as freely as possible in any form whatsoever and for any
purposes, including without limitation commercial purposes.
These owners may contribute to the Commons to promote the ideal
of a free culture and the further production of creative,
cultural and scientific works, or to gain reputation or greater
distribution for their Work in part through the use and efforts
of others.

For these and/or other purposes and motivations, and without any
expectation of additional consideration or compensation, the
person associating CC0 with a Work (the "Affirmer"), to the
extent that he or she is an owner of Copyright and Related
Rights in the Work, voluntarily elects to apply CC0 to the Work
and publicly distribute the Work under its terms, with knowledge
of his or her Copyright and Related Rights in the Work and the
meaning and intended legal effect of CC0 on those rights.

1. Copyright and Related Rights. A Work made available under CC0
may be protected by copyright and related or neighboring rights
("Copyright and Related Rights"). Copyright and Related Rights
include, but are not limited to, the following:

  i. the right to reproduce, adapt, distribute, perform,
     display, communicate, and translate a Work;
 ii. moral rights retained by the original author(s) and/or
     performer(s);
iii. publicity and privacy rights pertaining to a person's image
     or likeness depicted in a Work;
 iv. rights protecting against unfair competition in regards to
     a Work, subject to the limitations in paragraph 4(a), below;
  v. rights protecting the extraction, dissemination, use and
     reuse of data in a Work;
 vi. database rights (such as those arising under Directive
     96/9/EC of the European Parliament and of the Council of 11
     March 1996 on the legal protection of databases, and under
     any national implementation thereof, including any amended
     or successor version of such directive); and
vii. other similar, equivalent or corresponding rights
     throughout the world based on applicable law or treaty, and
     any national implementations thereof.

2. Waiver. To the greatest extent permitted by, but not in
contravention of, applicable law, Affirmer hereby overtly,
fully, permanently, irrevocably and unconditionally waives,
abandons, and surrenders all of Affirmer's Copyright and Related
Rights and associated claims and causes of action, whether now
known or unknown (including existing as well as future claims
and causes of action), in the Work (i) in all territories
worldwide, (ii) for the maximum duration provided by applicable
law or treaty (including future time extensions), (iii) in any
current or future medium and for any number of copies, and (iv)
for any purpose whatsoever, including without limitation
commercial, advertising or promotional purposes (the
"Waiver"). Affirmer makes the Waiver for the benefit of each
member of the public at large and to the detriment of Affirmer's
heirs and successors, fully intending that such Waiver shall not
be subject to revocation, rescission, cancellation, termination,
or any other legal or equitable action to disrupt the quiet
enjoyment of the Work by the public as contemplated by
Affirmer's express Statement of Purpose.

3. Public License Fallback. Should any part of the Waiver for
any reason be judged legally invalid or ineffective under
applicable law, then the Waiver shall be preserved to the
maximum extent permitted taking into account Affirmer's express
Statement of Purpose. In addition, to the extent the Waiver is
so judged Affirmer hereby grants to each affected person a
royalty-free, non transferable, non sublicensable, non
exclusive, irrevocable and unconditional license to exercise
Affirmer's Copyright and Related Rights in the Work (i) in all
territories worldwide, (ii) for the maximum duration provided by
applicable law or treaty (including future time extensions),
(iii) in any current or future medium and for any number of
copies, and (iv) for any purpose whatsoever, including without
limitation commercial, advertising or promotional purposes (the
"License"). The License shall be deemed effective as of the date
CC0 was applied by Affirmer to the Work. Should any part of the
License for any reason be judged legally invalid or ineffective
under applicable law, such partial invalidity or ineffectiveness
shall not invalidate the remainder of the License, and in such
case Affirmer hereby affirms that he or she will not (i)
exercise any of his or her remaining Copyright and Related
Rights in the Work or (ii) assert any associated claims and
causes of action with respect to the Work, in either case
contrary to Affirmer's express Statement of Purpose.

4. Limitations and Disclaimers.

a.  No trademark or patent rights held by Affirmer are waived,
    abandoned, surrendered, licensed or otherwise affected by
    this document.

b.  Affirmer offers the Work as-is and makes no representations
    or warranties of any kind concerning the Work, express,
    implied, statutory or otherwise, including without
    limitation warranties of title, merchantability, fitness for
    a particular purpose, non infringement, or the absence of
    latent or other defects, accuracy, or the present or absence
    of errors, whether or not discoverable, all to the greatest
    extent permissible under applicable law.

c.  Affirmer disclaims responsibility for clearing rights of
    other persons that may apply to the Work or any use thereof,
    including without limitation any person's Copyright and
    Related Rights in the Work.  Further, Affirmer disclaims
    responsibility for obtaining any necessary consents,
    permissions or other rights required for any use of the
    Work.

d.  Affirmer understands and acknowledges that Creative Commons
    is not a party to this document and has no duty or
    obligation with respect to this CC0 or use of the Work.
  

^[1] The downside to installing in a schema is that when writing SQL you must either set your search_path or qualify names, by prepending the object name with the schema name and a period. For example, without setting a search path, if you installed into a schema named isok, instead of SELECT * FROM run_isok_queries();, you would have to write SELECT * FROM isok.run_isok_queries();.

^[2] Possibly, the GNU m4 implementation is required. This is what PostgreSQL requires, and alternatives have not been tested.

^[3] The unspecified PL/pgSQL behavior being, that the target in a PL/pgSQL statement of the form “FOR target IN EXECUTE text_expression [ USING expression [, ... ] ] LOOP” is allowed to contain more variables than the test_expression returns columns, in which case the extra variables are assigned the NULL value. Because this behavior is undocumented, it is subject to change. Should this behavior change, returning a third column will be required, not optional.

^[4] To remove an ISOK_RESULTS.Deferred_To value and add a ISOK_RESULTS.Resolved value without raising an error either update both values in the same UPDATE statement or first set the Deferred_To value to NULL and then the Resolved value to something non-NULL.

^[5] The time the current transaction started, which, if a transaction was not explicitly started, is the time the database engine received the current SQL statement from the client and began execution.

^[6] Yes, this is true of all objects. Only the owner has access to any newly-created object. But having a single point of access, the schema dedicated to Isok, that grants access to all of Isok, provides a very useful point of control that serves as an easily audited gateway to Isok's functionality.

^[7] See: Estimating the Total Development Cost of a Linux Distribution.

^[8] publicity or privacy — The use of a work free of known copyright restrictions may be otherwise regulated or limited. The work or its use may be subject to personal data protection laws, publicity, image, or privacy rights that allow a person to control how their voice, image or likeness is used, or other restrictions or limitations under applicable law.

^[9] endorsement — In some jurisdictions, wrongfully implying that an author, publisher or anyone else endorses your use of a work may be unlawful.