1 TODO list for PostgreSQL
2 ========================
3 Last updated: Sat Oct 5 00:00:16 EDT 2002
5 Current maintainer: Bruce Momjian (pgman@candle.pha.pa.us)
7 The most recent version of this document can be viewed at
8 the PostgreSQL web site, http://www.PostgreSQL.org.
10 A dash (-) marks changes that will appear in the upcoming 7.3 release.
12 Bracketed items "[]" have more detailed.
18 * Add replication of distributed databases [replication]
21 o master/slave replication
22 o multi-master replication
23 o partition data across servers
24 o sample implementation in contrib/rserv
25 o queries across databases or servers (two-phase commit)
26 o allow replication over unreliable or non-persistent links
27 o http://gborg.postgresql.org/project/pgreplication/projdisplay.php
28 * Point-in-time data recovery using backup and write-ahead log
34 * Allow elog() to return error codes, module name, file name, line
35 number, not just messages (Peter E)
36 * Add error codes (Peter E)
37 * Make error messages more consistent [error]
38 * -Change DEBUG startup tag to LOG (Bruce)
39 * Show location of syntax error in query [yacc]
40 * -Add getpid() function to backend
41 * -Allow logging of query durations
47 * -Improve control over user privileges, including table creation
48 * -Allow user/group names to be specified directly in pg_hba.conf (Bruce)
49 * -Add ~/.pgpass to store passwords with user/host/password combinations
50 * -Allow permissions for functions (Peter E)
51 * -Allow object creation to be disabled for specific users
58 * -Make it easier to create a database owned by someone who can't createdb,
59 perhaps CREATE DATABASE dbname WITH OWNER = "user" (Gavin)
60 * -Make equals sign optional in CREATE DATABASE WITH param = 'val'
61 * Remove unreferenced table files and temp tables during database vacuum
62 or postmaster startup (Bruce)
63 * Remove behavior of postmaster -o after making postmaster/postgres
65 * -Prevent SIGHUP and 'pg_ctl reload' from changing command line
66 specified parameters to postgresql.conf defaults (Peter E)
67 * Allow easy display of usernames in a group
68 * Allow configuration files to be specified in a different directory
69 * -Reserve last few process slots for super-user if max_connections reached
70 * -Add GUC parameter to print queries that generate errors
71 * Add start time to pg_stat_activity
77 * -Add domain capability (Rod Taylor)
78 * Add IPv6 capability to INET/CIDR types
79 * Remove Money type, add money formatting for decimal type
80 * -SELECT cash_out(2) crashes because of opaque
81 * -Declare typein/out functions in pg_proc with a special "C string" data type
82 * -Functions returning sets do not totally work
83 * Change factorial to return a numeric
84 * Change NUMERIC data type to use base 10,000 internally
85 * Change NUMERIC to enforce the maximum precision, and increase it
86 * Add function to return compressed length of TOAST data values (Tom)
87 * -Add GUC parameter for DATESTYLE
88 * Allow INET subnet tests using non-constants
89 * -Allow bytea to handle LIKE with non-TEXT patterns
90 * -to_char(0,'FM999.99') returns a period, to_char(1,'FM999.99') doesn't (Karel)
91 * Add floor(float8) and other missing functions
94 o -Store binary-compatible type information in the system
95 o Allow better handling of numeric constants, type conversion
99 o Allow nulls in arrays
100 o Allow arrays to be ORDER'ed
101 o -Ensure we have array-eq operators for every built-in array type
102 o Support construction of array result values in expressions
105 o Improve vacuum of large objects, like /contrib/vacuumlo
106 o Add security checking for large objects
107 o Make file in/out interface for TOAST columns, similar to large object
108 interface (force out-of-line storage and no compression)
109 o Auto-delete large objects when referencing row is deleted
112 Multi-Language Support
113 ======================
115 * Add NCHAR (as distinguished from ordinary varchar),
116 * Allow LOCALE on a per-column basis, default to ASCII
117 * Support multiple simultaneous character sets, per SQL92
118 * Improve Unicode combined character handling
119 * Optimize locale to have minimal performance impact when not used (Peter E)
120 * Add octet_length_server() and octet_length_client() (Thomas, Tatsuo)
121 * Make octet_length_client the same as octet_length() (?)
122 * Prevent mismatch of frontend/backend encodings from converting bytea
123 data from being interpreted as encoded strings
124 * Remove Cyrillic recode support
130 * Automatically create rules on views so they are updateable, per SQL92 [view]
131 * Add the functionality for WITH CHECK OPTION clause of CREATE VIEW
132 * Allow NOTIFY in rules involving conditionals
133 * Have views on temporary tables exist in the temporary namespace
134 * Move psql backslash information into views
135 * Allow RULE recompilation
136 * -Remove brackets as multi-statement rule grouping, must use parens (Bruce)
137 * Prevent aggregates from being used in rule WHERE clauses
143 * Allow CREATE INDEX zman_index ON test (date_trunc( 'day', zman ) datetime_ops)
144 fails index can't store constant parameters
145 * Order duplicate index entries by tid for faster heap lookups
146 * Allow inherited tables to inherit index, UNIQUE constraint, and primary
147 key, foreign key [inheritance]
148 * UNIQUE INDEX on base column not honored on inserts from inherited table
149 INSERT INTO inherit_table (unique_index_col) VALUES (dup) should fail
151 * -Allow UPDATE/DELETE on inherited table
152 * Add UNIQUE capability to non-btree indexes
153 * Add btree index support for reltime, tinterval, regproc
154 * Add rtree index support for line, lseg, path, point
155 * Certain indexes will not shrink, e.g. indexes on ever-increasing
156 columns and indexes with many duplicate keys
157 * Use indexes for min() and max() or convert to SELECT col FROM tab ORDER
158 BY col DESC LIMIT 1 if appropriate index exists and WHERE clause acceptible
159 * Allow LIKE indexing optimization for non-ASCII locales
160 * Use index to restrict rows returned by multi-key index when used with
161 non-consecutive keys or OR clauses, so fewer heap accesses
162 * Be smarter about insertion of already-ordered data into btree index
163 * -Add deleted bit to index tuples to reduce heap access
164 * Prevent index uniqueness checks when UPDATE does not modifying column
165 * Use bitmaps to fetch heap pages in sequential order [performance]
166 * Use bitmaps to combine existing indexes [performance]
167 * Improve handling of index scans for NULL
168 * Allow SELECT * FROM tab WHERE int2col = 4 to use int2col index, int8,
169 float4, numeric/decimal too [optimizer]
170 * Add FILLFACTOR to btree index creation
171 * Improve concurrency in GIST
172 * Improve concurrency of hash indexes (Neil Conway)
173 * -Test hash index performance and discourage usage
179 * -Add SIMILAR TO to allow character classes, 'pg_[a-c]%'
180 * Add BETWEEN ASYMMETRIC/SYMMETRIC (Christopher)
181 * -Remove LIMIT #,# and force use LIMIT and OFFSET clauses in 7.3 (Bruce)
182 * Allow LIMIT/OFFSET to use expressions
183 * -Disallow TRUNCATE on tables that are involved in referential constraints
184 * -Add OR REPLACE clauses to non-FUNCTION object creation
185 * CREATE TABLE AS can not determine column lengths from expressions [atttypmod]
186 * Allow UPDATE to handle complex aggregates [update]
187 * -Prevent create/drop scripts from allowing extra args (Bruce)
188 * Allow command blocks to ignore certain types of errors
189 * Allow backslash handling in quoted strings to be disabled for portability
190 * Return proper effected tuple count from complex commands [return]
191 * Allow DELETE to handle table aliases for self-joins [delete]
192 * Add CORRESPONDING BY to UNION/INTERSECT/EXCEPT
193 * Allow REINDEX to rebuild all indexes, remove /contrib/reindex
196 o ALTER TABLE ADD COLUMN does not honor DEFAULT and non-CHECK CONSTRAINT
197 o -Add ALTER TABLE DROP COLUMN feature
198 o -Add ALTER TABLE DROP non-CHECK CONSTRAINT
199 o -ALTER TABLE ADD PRIMARY KEY (Tom)
200 o -ALTER TABLE ADD UNIQUE (Tom)
201 o -ALTER TABLE ALTER COLUMN SET/DROP NOT NULL (Christopher)
202 o ALTER TABLE ADD COLUMN column DEFAULT should fill existing
203 rows with DEFAULT value
204 o ALTER TABLE ADD COLUMN column SERIAL doesn't create sequence because
206 o -Have ALTER TABLE OWNER change all dependant objects like indexes
207 o Add ALTER TABLE tab SET WITHOUT OIDS
210 o -Cluster all tables at once using pg_index.indisclustered set during
212 o -Prevent loss of indexes, permissions, inheritance
213 o Automatically maintain clustering on a table
214 o Allow CLUSTER to cluster all tables, remove clusterdb
217 o -Allow specification of column names
218 o Allow dump/load of CSV format
219 o -Change syntax to WITH DELIMITER, (keep old syntax around?)
220 o Allow COPY to report error lines and continue; optionally
221 allow error codes to be specified; requires savepoints or can
222 not be run in a multi-statement transaction
223 o -Generate failure on short COPY lines rather than pad NULLs
226 o Allow BINARY option to SELECT, just like DECLARE
227 o MOVE 0 should not move to end of cursor
228 o Allow UPDATE/DELETE WHERE CURRENT OF cursor using per-cursor tid
229 stored in the backend
230 o Prevent DROP of table being referenced by our own open cursor
231 o Allow cursors outside transactions [cursor]
234 o Allow INSERT/UPDATE of system-generated oid value for a row
235 o Allow INSERT INTO tab (col1, ..) VALUES (val1, ..), (val2, ..)
236 o -Allow INSERT INTO my_table VALUES (a, b, c, DEFAULT, x, y, z, ...)
237 o -Disallow missing columns in INSERT ... (col) VALUES, per ANSI
238 o Allow INSERT/UPDATE ... RETURNING new.col or old.col; handle
242 o -Add command to display locks
243 o -Add SET or BEGIN timeout parameter to cancel query
244 o Add SET REAL_FORMAT and SET DOUBLE_PRECISION_FORMAT using printf args
245 o -Remove SET KSQO option now that OR processing is improved (Bruce)
246 o Add SET PERFORMANCE_TIPS option to suggest INDEX, VACUUM, VACUUM
248 o -Add SHOW command to see locale
249 o -Allow SHOW to output as a query result, like EXPLAIN
250 o -Abort all SET changes made in an aborted transaction
252 o Allow EXPLAIN EXECUTE to see prepared plans
254 * SERVER-SIDE LANGUAGES
255 o Allow PL/PgSQL's RAISE function to take expressions
256 o -Fix PL/PgSQL to handle quoted mixed-case identifiers
257 o Change PL/PgSQL to use palloc() instead of malloc()
258 o Add untrusted version of plpython
259 o Allow Java server-side programming, http://pljava.sourceforge.net
261 o Fix problems with complex temporary table creation/destruction
262 without using PL/PgSQL EXECUTE, needs cache prevention/invalidation
263 o Fix PL/pgSQL RENAME to work on variables other than OLD/NEW
264 o Improve PL/PgSQL exception handling
265 o Allow parameters to be specified by name and type during
267 o Allow function parameters to be passed by name,
268 get_employee_salary(emp_id => 12345, tax_year => 2001)
269 o Add PL/PgSQL packages
270 o Allow array declarations and other data types in PL/PgSQl DECLARE
271 o Add PL/PgSQL PROCEDURES that can return multiple values
272 o Add table function support to pltcl, plperl, plpython
273 o Make PL/PgSQL %TYPE schema-aware
279 * -Have pg_dump use LEFT OUTER JOIN in multi-table SELECTs
280 or multiple SELECTS to avoid bad system catalog entries
281 * -Have pg_dump -C dump database location and encoding information
282 * -Allow psql \d to show foreign keys
283 * -Allow psql \d to show temporary table structure (Tom)
284 * Allow psql to show transaction status if backend protocol changes made
285 * Add XML interface: psql, pg_dump, COPY, separate server (?)
286 * -Have pg_dump use ADD PRIMARY KEY after COPY, for performance (Neil Conway)
287 * Add schema, cast, and conversion backslash commands to psql
290 o Comprehensive test suite. This may be available already.
291 o -Updateable resultSet
292 o JDBC-standard BLOB support
293 o Error Codes (pending backend implementation)
294 o Support both 'make' and 'ant'
295 o Fix LargeObject API to handle OIDs as unsigned ints
296 o -Implement cancel() method on Statement
297 o Use cursors implicitly to avoid large results (see setCursorName())
298 o -Add support for CallableStatements
299 o Add LISTEN/NOTIFY support to the JDBC driver (Barry)
300 o -Compile under jdk 1.4
303 o Implement set descriptor, using descriptor
304 o Make casts work in variable initializations
306 o Allow multi-threaded use of SQLCA
307 o Solve cardinality > 1 for input descriptors / variables
308 o Understand structure definitions outside a declare section
309 o sqlwarn[6] should be 'W' if the PRECISION or SCALE value specified
310 o Improve error handling
311 o Allow :var[:index] or :var[<integer>] as cvariable for an array var
312 o Add a semantic check level, e.g. check if a table really exists
313 o Fix nested C comments
315 o fix handling of DB attributes that are arrays
318 Referential Integrity
319 =====================
321 * Add MATCH PARTIAL referential integrity [foreign]
322 * Add deferred trigger queue file (Jan)
323 * -Allow oid to act as a foreign key
324 * Implement dirty reads and use them in RI triggers
325 * Enforce referential integrity for system tables
326 * -Allow user to control trigger firing order (Tom)
327 * -Add ALTER TRIGGER ... RENAME
328 * Change foreign key constraint for array -> element to mean element
330 * -Fix foreign key constraints to not error on intermediate db states (Stephan)
331 * Allow DEFERRABLE UNIQUE constraints
332 * Allow triggers to be disabled [trigger]
338 * -Add pg_depend table for dependency recording; use sysrelid, oid,
339 depend_sysrelid, depend_oid, name
340 * -Auto-destroy sequence on DROP of table with SERIAL; perhaps a separate
342 * -Prevent column dropping if column is used by foreign key
343 * -Propagate column or table renaming to foreign key constraints
344 * -Automatically drop constraints/functions when object is dropped
345 * -Make foreign key constraints clearer in dump file
346 * -Make other constraints clearer in dump file
347 * -Make foreign keys easier to identify
348 * Flush cached query plans when their underlying catalog data changes
349 * Use dependency information to dump data in proper order
355 * -Allow autocommit so always in a transaction block
356 * Overhaul bufmgr/lockmgr/transaction manager
357 * Allow savepoints / nested transactions [transactions]
363 * Add sql3 recursive unions
364 * Add the concept of dataspaces/tablespaces [tablespaces]
365 * -Add SQL92 schemas (Tom)
366 * Allow queries across multiple databases [crossdb]
367 * Add pre-parsing phase that converts non-ANSI features to supported features
368 * Allow plug-in modules to emulate features from other databases
369 * SQL*Net listener that makes PostgreSQL appear as an Oracle database
380 * Delay fsync() when other backends are about to commit too [fsync]
381 o Determine optimal commit_delay value
382 * Determine optimal fdatasync/fsync, O_SYNC/O_DSYNC options
383 o Allow multiple blocks to be written to WAL with one write()
388 * -Cache most recent query plan(s) (Neil) [prepare]
389 * Shared catalog cache, reduce lseek()'s by caching table size in shared area
390 * Add free-behind capability for large sequential scans (Bruce)
391 * Allow binding query args over FE/BE protocol
392 * Consider use of open/fcntl(O_DIRECT) to minimize OS caching
393 * Make blind writes go through the file descriptor cache
399 * Improve speed with indexes (perhaps recreate index instead) [vacuum]
400 * Reduce lock time by moving tuples with read lock, then write
401 lock and truncate table [vacuum]
402 * Provide automatic running of vacuum in the background (Tom) [vacuum]
403 * Allow free space map to be auto-sized or warn when it is too small
409 * Make locking of shared data structures more fine-grained
410 * Add code to detect an SMP machine and handle spinlocks accordingly
411 from distributted.net, http://www1.distributed.net/source,
412 in client/common/cpucheck.cpp
413 * Research use of sched_yield() for spinlock acquisition failure
419 * Experiment with multi-threaded backend [thread]
420 * Add connection pooling [pool]
421 * Allow persistent backends [persistent]
422 * Create a transaction processor to aid in persistent connections and
424 * Do listen() in postmaster and accept() in pre-forked backend
425 * Have pre-forked backend pre-connect to last requested database or pass
426 file descriptor to backend pre-forked for matching database
427 * -Cache system catalog information in per-database files (Tom)
433 * Have after-change WAL write()'s write only modified data to kernel
434 * Reduce number of after-change WAL writes; they exist only to gaurd against
435 partial page writes [wal]
436 * Turn off after-change writes if fsync is disabled (?)
437 * Add WAL index reliability improvement to non-btree indexes
438 * -Reorder postgresql.conf WAL items in order of importance (Bruce)
439 * -Remove wal_files postgresql.conf option because WAL files are now recycled
440 * Find proper defaults for postgresql.conf WAL entries
441 * Add checkpoint_min_warning postgresql.conf option to warn about checkpoints
442 that are too frequent
443 * Allow xlog directory location to be specified during initdb, perhaps
445 * Allow pg_xlog to be moved without symlinks
451 * Improve Subplan list handling
452 * Allow Subplans to use efficient joins(hash, merge) with upper variable
453 * -Improve dynamic memory allocation by introducing tuple-context memory
455 * Add hash for evaluating GROUP BY aggregates
456 * -Nested FULL OUTER JOINs don't work (Tom)
457 * Allow merge and hash joins on expressions not just simple variables (Tom)
458 * -Add new pg_proc cachable settings to specify whether function can be
459 evaluated only once or once per query
460 * -Change FIXED_CHAR_SEL to 0.20 from 0.04 to give better selectivity (Bruce)
461 * Make IN/NOT IN have similar performance to EXISTS/NOT EXISTS [exists]
462 * Missing optimizer selectivities for date, r-tree, etc. [optimizer]
463 * Allow ORDER BY ... LIMIT to select top values without sort or index
464 using a sequential scan for highest/lowest values (Oleg)
465 * Inline simple SQL functions to avoid overhead (Tom)
466 * Precompile SQL functions to avoid overhead (Neil Conway)
467 * Add utility to compute accurate random_page_cost value
468 * Improve ability to display optimizer analysis using OPTIMIZER_DEBUG
469 * Use CHECK constraints to improve optimizer decisions
470 * Check GUC geqo_threshold to see if it is still accurate
471 * Allow sorting, temp files, temp tables to use multiple work directories
477 * Do async I/O for faster random read-ahead of data
478 * Get faster regex() code from Henry Spencer <henry@zoo.utoronto.ca>
480 * Use mmap() rather than SYSV shared memory or to write WAL files (?) [mmap]
481 * Improve caching of attribute offsets when NULLs exist in the row
482 * -Add Intimate Shared Memory(ISM) for Solaris
483 * -Use faster flex flags for performance improvement (Peter E)
484 * -Add BSD-licensed qsort() for Solaris
490 * Add use of 'const' for variables in source tree
491 * -Fix problems with libpq non-blocking/async code
492 * -Make sure all block numbers are unsigned to increase maximum table size
493 * -Merge LockMethodCtl and LockMethodTable into one shared structure (Bruce)
494 * -HOLDER/HOLDERTAB rename to PROCLOCK/PROCLOCKTAG (Bruce)
495 * -Remove LockMethodTable.prio field, not used (Bruce)
496 * Rename some /contrib modules from pg* to pg_*
497 * Move some things from /contrib into main tree
498 * Remove warnings created by -Wcast-align
499 * Move platform-specific ps status display info from ps_status.c to ports
500 * -Make one version of simple_prompt() in code (Bruce, Tom)
501 * -Compile in syslog functionaility by default (Tatsuo)
502 * Modify regression tests to prevent failures do to minor numeric rounding
503 * Add OpenBSD's getpeereid() call for local socket authentication (Bruce)
504 * Improve access-permissions check on data directory in Cygwin (Tom)
505 * -Report failure to find readline or zlib at end of configure run
506 * Add --port flag to regression tests
507 * -Increase identifier length (NAMEDATALEN) if small performance hit,
508 * -Increase maximum number of function parameters if little wasted space
509 * Add documentation for perl, including mention of DBI/DBD perl location
510 * Add optional CRC checksum to heap and index pages
511 * Change representation of whole-tuple parameters to functions
512 * Clarify use of 'application' and 'command' tags in SGML docs
513 * Better document ability to build only certain interfaces (Marc)
514 * Remove or relicense modules that are not under the BSD license, if possible
515 * Remove memory/file descriptor freeing befor elog(ERROR) (Bruce)
516 * Create native Win32 port [win32]
517 * Fix glibc's mktime() to handle pre-1970's dates
520 ---------------------------------------------------------------------------
523 Developers who have claimed items are:
524 --------------------------------------
525 * Barry is Barry Lind <barry@xythos.com>
526 * Billy is Billy G. Allie <Bill.Allie@mug.org>
527 * Bruce is Bruce Momjian <pgman@candle.pha.pa.us>
528 * Christopher is Christopher Kings-Lynne <chriskl@familyhealth.com.au>
529 * D'Arcy is D'Arcy J.M. Cain <darcy@druid.net>
530 * Dave is Dave Cramer <dave@fastcrypt.com>
531 * Edmund is Edmund Mergl <E.Mergl@bawue.de>
532 * Gavin Sherry <swm@linuxworld.com.au>
533 * Hiroshi is Hiroshi Inoue <Inoue@tpf.co.jp>
534 * Karel is Karel Zak <zakkr@zf.jcu.cz>
535 * Jan is Jan Wieck <wieck@sapserv.debis.de>
536 * Liam is Liam Stewart <liams@redhat.com>
537 * Marc is Marc Fournier <scrappy@hub.org>
538 * Mark is Mark Hollomon <mhh@mindspring.com>
539 * Marko is Marko Kreen <marko@l-t.ee>
540 * Michael is Michael Meskes <meskes@postgresql.org>
541 * Neil is Neil Conway <nconway@klamath.dyndns.org>
542 * Oleg is Oleg Bartunov <oleg@sai.msu.su>
543 * Peter M is Peter T Mount <peter@retep.org.uk>
544 * Peter E is Peter Eisentraut <peter_e@gmx.net>
545 * Philip is Philip Warner <pjw@rhyme.com.au>
546 * Rod is Rod Taylor <rbt@zort.ca>
547 * Ross is Ross J. Reedstrom <reedstrm@wallace.ece.rice.edu>
548 * Ryan is Ryan Bradetich <rbrad@hpb50023.boi.hp.com>
549 * Stephan is Stephan Szabo <sszabo@megazone23.bigpanda.com>
550 * Tatsuo is Tatsuo Ishii <t-ishii@sra.co.jp>
551 * Thomas is Thomas Lockhart <lockhart@fourpalms.org>
552 * Tom is Tom Lane <tgl@sss.pgh.pa.us>
553 * TomH is Tom I Helbekkmo <tih@Hamartun.Priv.no>
554 * Vadim is Vadim B. Mikheev <vadim4o@email.com>