1 TODO list for PostgreSQL
2 ========================
3 Last updated: Fri Nov 1 17:52:40 EST 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 pg_backend_pid() 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
72 * Allow limits on per-db/user connections
73 * Have standalone backend read postgresql.conf
79 * -Add domain capability (Rod Taylor)
80 * Add IPv6 capability to INET/CIDR types
81 * Remove Money type, add money formatting for decimal type
82 * -SELECT cash_out(2) crashes because of opaque
83 * -Declare typein/out functions in pg_proc with a special "C string" data type
84 * -Functions returning sets do not totally work
85 * Change factorial to return a numeric
86 * Change NUMERIC data type to use base 10,000 internally
87 * Change NUMERIC to enforce the maximum precision, and increase it
88 * Add function to return compressed length of TOAST data values (Tom)
89 * -Add GUC parameter for DATESTYLE
90 * Allow INET subnet tests using non-constants
91 * -Allow bytea to handle LIKE with non-TEXT patterns
92 * -to_char(0,'FM999.99') returns a period, to_char(1,'FM999.99') doesn't (Karel)
93 * -Add floor(float8) and other missing functions
94 * Add now("transaction|statement|clock") functionality
97 o -Store binary-compatible type information in the system
98 o Allow better handling of numeric constants, type conversion
102 o Allow nulls in arrays
103 o Allow arrays to be ORDER'ed
104 o -Ensure we have array-eq operators for every built-in array type
105 o Support construction of array result values in expressions
108 o Improve vacuum of large objects, like /contrib/vacuumlo
109 o Add security checking for large objects
110 o Make file in/out interface for TOAST columns, similar to large object
111 interface (force out-of-line storage and no compression)
112 o Auto-delete large objects when referencing row is deleted
115 Multi-Language Support
116 ======================
118 * Add NCHAR (as distinguished from ordinary varchar),
119 * Allow LOCALE on a per-column basis, default to ASCII
120 * Support multiple simultaneous character sets, per SQL92
121 * Improve Unicode combined character handling
122 * Optimize locale to have minimal performance impact when not used (Peter E)
123 * Add octet_length_server() and octet_length_client() (Thomas, Tatsuo)
124 * Make octet_length_client the same as octet_length() (?)
125 * Prevent mismatch of frontend/backend encodings from converting bytea
126 data from being interpreted as encoded strings
127 * Remove Cyrillic recode support
133 * Automatically create rules on views so they are updateable, per SQL92 [view]
134 * Add the functionality for WITH CHECK OPTION clause of CREATE VIEW
135 * Allow NOTIFY in rules involving conditionals
136 * Have views on temporary tables exist in the temporary namespace
137 * Move psql backslash information into views
138 * Allow RULE recompilation
139 * -Remove brackets as multi-statement rule grouping, must use parens (Bruce)
140 * -Prevent aggregates from being used in rule WHERE clauses
146 * Allow CREATE INDEX zman_index ON test (date_trunc( 'day', zman ) datetime_ops)
147 fails index can't store constant parameters
148 * Order duplicate index entries by tid for faster heap lookups
149 * Allow inherited tables to inherit index, UNIQUE constraint, and primary
150 key, foreign key [inheritance]
151 * UNIQUE INDEX on base column not honored on inserts from inherited table
152 INSERT INTO inherit_table (unique_index_col) VALUES (dup) should fail
154 * -Allow UPDATE/DELETE on inherited table
155 * Add UNIQUE capability to non-btree indexes
156 * Add btree index support for reltime, tinterval, regproc
157 * Add rtree index support for line, lseg, path, point
158 * Certain indexes will not shrink, e.g. indexes on ever-increasing
159 columns and indexes with many duplicate keys
160 * Use indexes for min() and max() or convert to SELECT col FROM tab ORDER
161 BY col DESC LIMIT 1 if appropriate index exists and WHERE clause acceptible
162 * Allow LIKE indexing optimization for non-ASCII locales
163 * Use index to restrict rows returned by multi-key index when used with
164 non-consecutive keys or OR clauses, so fewer heap accesses
165 * Be smarter about insertion of already-ordered data into btree index
166 * -Add deleted bit to index tuples to reduce heap access
167 * Prevent index uniqueness checks when UPDATE does not modifying column
168 * Use bitmaps to fetch heap pages in sequential order [performance]
169 * Use bitmaps to combine existing indexes [performance]
170 * Improve handling of index scans for NULL
171 * Allow SELECT * FROM tab WHERE int2col = 4 to use int2col index, int8,
172 float4, numeric/decimal too [optimizer]
173 * Add FILLFACTOR to btree index creation
174 * Add concurrency to GIST
175 * Improve concurrency of hash indexes (Neil)
176 * -Test hash index performance and discourage usage
182 * -Add SIMILAR TO to allow character classes, 'pg_[a-c]%'
183 * Add BETWEEN ASYMMETRIC/SYMMETRIC (Christopher)
184 * -Remove LIMIT #,# and force use LIMIT and OFFSET clauses in 7.3 (Bruce)
185 * Allow LIMIT/OFFSET to use expressions
186 * -Disallow TRUNCATE on tables that are involved in referential constraints
187 * -Add OR REPLACE clauses to non-FUNCTION object creation
188 * CREATE TABLE AS can not determine column lengths from expressions [atttypmod]
189 * Allow UPDATE to handle complex aggregates [update]
190 * -Prevent create/drop scripts from allowing extra args (Bruce)
191 * Allow command blocks to ignore certain types of errors
192 * Allow backslash handling in quoted strings to be disabled for portability
193 * Return proper effected tuple count from complex commands [return]
194 * Allow DELETE to handle table aliases for self-joins [delete]
195 * Add CORRESPONDING BY to UNION/INTERSECT/EXCEPT
196 * Allow REINDEX to rebuild all indexes, remove /contrib/reindex
197 * Make a transaction-safe TRUNCATE
200 o ALTER TABLE ADD COLUMN does not honor DEFAULT and non-CHECK CONSTRAINT
201 o -Add ALTER TABLE DROP COLUMN feature
202 o -Add ALTER TABLE DROP non-CHECK CONSTRAINT
203 o -ALTER TABLE ADD PRIMARY KEY (Tom)
204 o -ALTER TABLE ADD UNIQUE (Tom)
205 o -ALTER TABLE ALTER COLUMN SET/DROP NOT NULL (Christopher)
206 o ALTER TABLE ADD COLUMN column DEFAULT should fill existing
207 rows with DEFAULT value
208 o ALTER TABLE ADD COLUMN column SERIAL doesn't create sequence because
210 o -Have ALTER TABLE OWNER change all dependant objects like indexes
211 o Add ALTER TABLE tab SET WITHOUT OIDS
214 o -Cluster all tables at once using pg_index.indisclustered set during
216 o -Prevent loss of indexes, permissions, inheritance
217 o Automatically maintain clustering on a table
218 o Allow CLUSTER to cluster all tables, remove clusterdb
221 o -Allow specification of column names
222 o Allow psql \copy to specify column names
223 o Allow dump/load of CSV format
224 o -Change syntax to WITH DELIMITER, (keep old syntax around?)
225 o Allow COPY to report error lines and continue; optionally
226 allow error codes to be specified; requires savepoints or can
227 not be run in a multi-statement transaction
228 o -Generate failure on short COPY lines rather than pad NULLs
229 o Allow copy to understand \x as hex
232 o Allow BINARY option to SELECT, just like DECLARE
233 o MOVE 0 should not move to end of cursor
234 o Allow UPDATE/DELETE WHERE CURRENT OF cursor using per-cursor tid
235 stored in the backend
236 o Prevent DROP of table being referenced by our own open cursor
237 o Allow cursors outside transactions [cursor]
240 o Allow INSERT/UPDATE of system-generated oid value for a row
241 o Allow INSERT INTO tab (col1, ..) VALUES (val1, ..), (val2, ..)
242 o -Allow INSERT INTO my_table VALUES (a, b, c, DEFAULT, x, y, z, ...)
243 o -Disallow missing columns in INSERT ... (col) VALUES, per ANSI
244 o Allow INSERT/UPDATE ... RETURNING new.col or old.col; handle
248 o -Add command to display locks
249 o -Add SET or BEGIN timeout parameter to cancel query
250 o Add SET REAL_FORMAT and SET DOUBLE_PRECISION_FORMAT using printf args
251 o -Remove SET KSQO option now that OR processing is improved (Bruce)
252 o Add SET PERFORMANCE_TIPS option to suggest INDEX, VACUUM, VACUUM
254 o -Add SHOW command to see locale
255 o -Allow SHOW to output as a query result, like EXPLAIN
256 o -Abort all SET changes made in an aborted transaction
258 o Allow EXPLAIN EXECUTE to see prepared plans
259 o Allow SHOW of non-modifiable variables, like pg_controldata
260 o Add GUC parameter to control the maximum number of rewrite cycles
262 * SERVER-SIDE LANGUAGES
263 o Allow PL/PgSQL's RAISE function to take expressions
264 o -Fix PL/PgSQL to handle quoted mixed-case identifiers
265 o Change PL/PgSQL to use palloc() instead of malloc()
266 o Add untrusted version of plpython
267 o Allow Java server-side programming, http://pljava.sourceforge.net
269 o Fix problems with complex temporary table creation/destruction
270 without using PL/PgSQL EXECUTE, needs cache prevention/invalidation
271 o Fix PL/pgSQL RENAME to work on variables other than OLD/NEW
272 o Improve PL/PgSQL exception handling
273 o Allow parameters to be specified by name and type during
275 o Allow function parameters to be passed by name,
276 get_employee_salary(emp_id => 12345, tax_year => 2001)
277 o Add PL/PgSQL packages
278 o Allow array declarations and other data types in PL/PgSQL DECLARE
279 o Add PL/PgSQL PROCEDURES that can return multiple values
280 o Add table function support to pltcl, plperl, plpython
281 o Make PL/PgSQL %TYPE schema-aware
287 * -Have pg_dump use LEFT OUTER JOIN in multi-table SELECTs
288 or multiple SELECTS to avoid bad system catalog entries
289 * -Have pg_dump -C dump database location and encoding information
290 * -Allow psql \d to show foreign keys
291 * -Allow psql \d to show temporary table structure (Tom)
292 * Allow psql to show transaction status if backend protocol changes made
293 * Add XML interface: psql, pg_dump, COPY, separate server (?)
294 * -Have pg_dump use ADD PRIMARY KEY after COPY, for performance (Neil)
295 * Add schema, cast, and conversion backslash commands to psql
296 * Allow pg_dump to dump a specific schema
299 o Comprehensive test suite. This may be available already.
300 o -Updateable resultSet
301 o JDBC-standard BLOB support
302 o Error Codes (pending backend implementation)
303 o Support both 'make' and 'ant'
304 o Fix LargeObject API to handle OIDs as unsigned ints
305 o -Implement cancel() method on Statement
306 o Use cursors implicitly to avoid large results (see setCursorName())
307 o -Add support for CallableStatements
308 o Add LISTEN/NOTIFY support to the JDBC driver (Barry)
309 o -Compile under jdk 1.4
312 o Implement set descriptor, using descriptor
313 o Make casts work in variable initializations
315 o Allow multi-threaded use of SQLCA
316 o Solve cardinality > 1 for input descriptors / variables
317 o Understand structure definitions outside a declare section
318 o sqlwarn[6] should be 'W' if the PRECISION or SCALE value specified
319 o Improve error handling
320 o Allow :var[:index] or :var[<integer>] as cvariable for an array var
321 o Add a semantic check level, e.g. check if a table really exists
322 o Fix nested C comments
324 o fix handling of DB attributes that are arrays
327 Referential Integrity
328 =====================
330 * Add MATCH PARTIAL referential integrity [foreign]
331 * Add deferred trigger queue file (Jan)
332 * -Allow oid to act as a foreign key
333 * Implement dirty reads and use them in RI triggers
334 * Enforce referential integrity for system tables
335 * -Allow user to control trigger firing order (Tom)
336 * -Add ALTER TRIGGER ... RENAME
337 * Change foreign key constraint for array -> element to mean element
339 * -Fix foreign key constraints to not error on intermediate db states (Stephan)
340 * Allow DEFERRABLE UNIQUE constraints
341 * Allow triggers to be disabled [trigger]
347 * -Add pg_depend table for dependency recording; use sysrelid, oid,
348 depend_sysrelid, depend_oid, name
349 * -Auto-destroy sequence on DROP of table with SERIAL; perhaps a separate
351 * -Prevent column dropping if column is used by foreign key
352 * -Propagate column or table renaming to foreign key constraints
353 * -Automatically drop constraints/functions when object is dropped
354 * -Make foreign key constraints clearer in dump file
355 * -Make other constraints clearer in dump file
356 * -Make foreign keys easier to identify
357 * Flush cached query plans when their underlying catalog data changes
358 * Use dependency information to dump data in proper order
364 * -Allow autocommit so always in a transaction block
365 * Overhaul bufmgr/lockmgr/transaction manager
366 * Allow savepoints / nested transactions [transactions]
372 * Add sql3 recursive unions
373 * Add the concept of dataspaces/tablespaces [tablespaces]
374 * -Add SQL92 schemas (Tom)
375 * Allow queries across multiple databases [crossdb]
376 * Add pre-parsing phase that converts non-ANSI features to supported features
377 * Allow plug-in modules to emulate features from other databases
378 * SQL*Net listener that makes PostgreSQL appear as an Oracle database
389 * Delay fsync() when other backends are about to commit too [fsync]
390 o Determine optimal commit_delay value
391 * Determine optimal fdatasync/fsync, O_SYNC/O_DSYNC options
392 o Allow multiple blocks to be written to WAL with one write()
397 * -Cache most recent query plan(s) (Neil) [prepare]
398 * Shared catalog cache, reduce lseek()'s by caching table size in shared area
399 * Add free-behind capability for large sequential scans (Bruce)
400 * Allow binding query args over FE/BE protocol
401 * Consider use of open/fcntl(O_DIRECT) to minimize OS caching
402 * Make blind writes go through the file descriptor cache
408 * Improve speed with indexes (perhaps recreate index instead) [vacuum]
409 * Reduce lock time by moving tuples with read lock, then write
410 lock and truncate table [vacuum]
411 * Provide automatic running of vacuum in the background (Tom) [vacuum]
412 * Allow free space map to be auto-sized or warn when it is too small
418 * Make locking of shared data structures more fine-grained
419 * Add code to detect an SMP machine and handle spinlocks accordingly
420 from distributted.net, http://www1.distributed.net/source,
421 in client/common/cpucheck.cpp
422 * Research use of sched_yield() for spinlock acquisition failure
428 * Experiment with multi-threaded backend [thread]
429 * Add connection pooling [pool]
430 * Allow persistent backends [persistent]
431 * Create a transaction processor to aid in persistent connections and
433 * Do listen() in postmaster and accept() in pre-forked backend
434 * Have pre-forked backend pre-connect to last requested database or pass
435 file descriptor to backend pre-forked for matching database
436 * -Cache system catalog information in per-database files (Tom)
442 * Have after-change WAL write()'s write only modified data to kernel
443 * Reduce number of after-change WAL writes; they exist only to gaurd against
444 partial page writes [wal]
445 * Turn off after-change writes if fsync is disabled (?)
446 * Add WAL index reliability improvement to non-btree indexes
447 * -Reorder postgresql.conf WAL items in order of importance (Bruce)
448 * -Remove wal_files postgresql.conf option because WAL files are now recycled
449 * Find proper defaults for postgresql.conf WAL entries
450 * Add checkpoint_min_warning postgresql.conf option to warn about checkpoints
451 that are too frequent
452 * Allow xlog directory location to be specified during initdb, perhaps
454 * Allow pg_xlog to be moved without symlinks
460 * Improve Subplan list handling
461 * Allow Subplans to use efficient joins(hash, merge) with upper variable
462 * -Improve dynamic memory allocation by introducing tuple-context memory
464 * Add hash for evaluating GROUP BY aggregates
465 * -Nested FULL OUTER JOINs don't work (Tom)
466 * Allow merge and hash joins on expressions not just simple variables (Tom)
467 * -Add new pg_proc cachable settings to specify whether function can be
468 evaluated only once or once per query
469 * -Change FIXED_CHAR_SEL to 0.20 from 0.04 to give better selectivity (Bruce)
470 * Make IN/NOT IN have similar performance to EXISTS/NOT EXISTS [exists]
471 * Missing optimizer selectivities for date, r-tree, etc. [optimizer]
472 * Allow ORDER BY ... LIMIT to select top values without sort or index
473 using a sequential scan for highest/lowest values (Oleg)
474 * Inline simple SQL functions to avoid overhead (Tom)
475 * Precompile SQL functions to avoid overhead (Neil)
476 * Add utility to compute accurate random_page_cost value
477 * Improve ability to display optimizer analysis using OPTIMIZER_DEBUG
478 * Use CHECK constraints to improve optimizer decisions
479 * Check GUC geqo_threshold to see if it is still accurate
480 * Allow sorting, temp files, temp tables to use multiple work directories
486 * Do async I/O for faster random read-ahead of data
487 * Get faster regex() code from Henry Spencer <henry@zoo.utoronto.ca>
489 * Use mmap() rather than SYSV shared memory or to write WAL files (?) [mmap]
490 * Improve caching of attribute offsets when NULLs exist in the row
491 * -Add Intimate Shared Memory(ISM) for Solaris
492 * -Use faster flex flags for performance improvement (Peter E)
493 * -Add BSD-licensed qsort() for Solaris
499 * Add use of 'const' for variables in source tree
500 * -Fix problems with libpq non-blocking/async code
501 * -Make sure all block numbers are unsigned to increase maximum table size
502 * -Merge LockMethodCtl and LockMethodTable into one shared structure (Bruce)
503 * -HOLDER/HOLDERTAB rename to PROCLOCK/PROCLOCKTAG (Bruce)
504 * -Remove LockMethodTable.prio field, not used (Bruce)
505 * Rename some /contrib modules from pg* to pg_*
506 * Move some things from /contrib into main tree
507 * Remove warnings created by -Wcast-align
508 * Move platform-specific ps status display info from ps_status.c to ports
509 * -Make one version of simple_prompt() in code (Bruce, Tom)
510 * -Compile in syslog functionaility by default (Tatsuo)
511 * Modify regression tests to prevent failures do to minor numeric rounding
512 * Add OpenBSD's getpeereid() call for local socket authentication (Bruce)
513 * Improve access-permissions check on data directory in Cygwin (Tom)
514 * -Report failure to find readline or zlib at end of configure run
515 * Add --port flag to regression tests
516 * -Increase identifier length (NAMEDATALEN) if small performance hit,
517 * -Increase maximum number of function parameters if little wasted space
518 * Add documentation for perl, including mention of DBI/DBD perl location
519 * Add optional CRC checksum to heap and index pages
520 * Change representation of whole-tuple parameters to functions
521 * Clarify use of 'application' and 'command' tags in SGML docs
522 * Better document ability to build only certain interfaces (Marc)
523 * Remove or relicense modules that are not under the BSD license, if possible
524 * Remove memory/file descriptor freeing befor elog(ERROR) (Bruce)
525 * Create native Win32 port [win32]
526 * -Fix glibc's mktime() to handle pre-1970's dates
527 * -Move /contrib/retep to gborg.postgresql.org
529 ---------------------------------------------------------------------------
532 Developers who have claimed items are:
533 --------------------------------------
534 * Barry is Barry Lind <barry@xythos.com>
535 * Billy is Billy G. Allie <Bill.Allie@mug.org>
536 * Bruce is Bruce Momjian <pgman@candle.pha.pa.us>
537 * Christopher is Christopher Kings-Lynne <chriskl@familyhealth.com.au>
538 * D'Arcy is D'Arcy J.M. Cain <darcy@druid.net>
539 * Dave is Dave Cramer <dave@fastcrypt.com>
540 * Edmund is Edmund Mergl <E.Mergl@bawue.de>
541 * Gavin Sherry <swm@linuxworld.com.au>
542 * Hiroshi is Hiroshi Inoue <Inoue@tpf.co.jp>
543 * Karel is Karel Zak <zakkr@zf.jcu.cz>
544 * Jan is Jan Wieck <wieck@sapserv.debis.de>
545 * Liam is Liam Stewart <liams@redhat.com>
546 * Marc is Marc Fournier <scrappy@hub.org>
547 * Mark is Mark Hollomon <mhh@mindspring.com>
548 * Marko is Marko Kreen <marko@l-t.ee>
549 * Michael is Michael Meskes <meskes@postgresql.org>
550 * Neil is Neil Conway <neilc@samurai.com>
551 * Oleg is Oleg Bartunov <oleg@sai.msu.su>
552 * Peter M is Peter T Mount <peter@retep.org.uk>
553 * Peter E is Peter Eisentraut <peter_e@gmx.net>
554 * Philip is Philip Warner <pjw@rhyme.com.au>
555 * Rod is Rod Taylor <rbt@zort.ca>
556 * Ross is Ross J. Reedstrom <reedstrm@wallace.ece.rice.edu>
557 * Ryan is Ryan Bradetich <rbrad@hpb50023.boi.hp.com>
558 * Stephan is Stephan Szabo <sszabo@megazone23.bigpanda.com>
559 * Tatsuo is Tatsuo Ishii <t-ishii@sra.co.jp>
560 * Thomas is Thomas Lockhart <lockhart@fourpalms.org>
561 * Tom is Tom Lane <tgl@sss.pgh.pa.us>
562 * TomH is Tom I Helbekkmo <tih@Hamartun.Priv.no>
563 * Vadim is Vadim B. Mikheev <vadim4o@email.com>