2 * add export to SQLite database and SQLite support for the frontend
3 * switch to Python frontend as the default frontend
4 * update to osm2pgsql 1.11.0
5 * add support for new osm2pgsql middle table format
6 * simplify geometry for large polygon objects not used in addresses
7 * various performance tweaks for search in Python frontend
8 * fix regression in search with categories where it was confused with near
10 * partially roll back use of SQLAlchemy lambda statements due to bugs
12 * fix handling of timezones for timestamps from the database
13 * fix handling of full address searches in connection with a viewbox
14 * fix postcode computation of highway areas
15 * fix handling of timeout errors for Python <= 3.10
16 * fix address computation for postcode areas
17 * fix variable shadowing in osm2pgsql flex script, causing bugs with LuaJIT
18 * make sure extratags are always null when empty
19 * reduce importance of places without wikipedia reference
20 * improve performance of word count computations
21 * drop support for wikipedia tags with full URLs
22 * replace get_addressdata() SQL implementation with a Python function
23 * improve display name for non-address features
24 * fix postcode validation for postcodes with country code
25 (thanks @pawel-wroniszewski)
26 * add possibility to run imports without superuser database rights
27 (thanks @robbe-haesendonck)
28 * new CLI command for cleaning deleted relations (thanks @lujoh)
29 * add check for database version in the CLI check command
30 * updates to import styles ignoring more unused objects
31 * various typo fixes (thanks @kumarUjjawal)
34 * fix potential SQL injection issue for 'nominatim admin --collect-os-info'
35 * PHP frontend: fix on-the-fly lookup of postcode areas near boundaries
36 * Python frontend: improve handling of viewbox
37 * Python frontend: correct deployment instructions
40 * reintroduce result rematching
41 * improve search of multi-part names
42 * fix accidentally switched meaning of --reverse-only and --search-only in
46 * fix failing importance recalculation command
47 * fix merging of linked names into unnamed boundaries
48 * fix a number of corner cases with interpolation splitting resulting in
50 * fix failure in website generation when password contains curly brackets
51 * fix broken use of ST_Project in PostGIS 3.4
52 * new NOMINATIM_SEARCH_WITHIN_COUNTRIES setting to restrict reverse lookups
53 to known countries (thanks @alfmarcua)
54 * allow negative OSM IDs (thanks @alfmarcua)
55 * disallow import of Tiger data in a frozen DB
56 * avoid UPDATE to change settings to be compatible with r/o DBs (thanks @t-tomek)
57 * update bundled osm2pgsql to 1.9.2
58 * reorganise osm2pgsql flex style and make it the default
59 * exclude names ending in :wikipedia from indexing
60 * no longer accept comma as a list separator in name tags
61 * process forward dependencies on update to catch updates in geometries
63 * fix handling of isolated silent letters during transliteration
64 * no longer assign postcodes to large linear features like rivers
65 * introduce nominatim.paths module for finding data and libraries
66 * documentation layout changed to material theme
67 * new documentation section for library
68 * various smaller fixes to existing documentation
69 (thanks @woodpeck, @bloom256, @biswajit-k)
70 * updates to vagrant install scripts, drop support for Ubuntu 18
72 * removed obsolete configuration variables from env.defaults
73 * add script for generating a taginfo description (thanks @biswajit-k)
74 * modernize Python code around BDD test and add testing of Python frontend
75 * lots of new BDD tests for API output
79 * fix deletion handling for 'nominatim add-data'
80 * adapt place_force_delete() to new deletion handling
81 * flex style: avoid dropping of postcode areas
82 * fix update errors on address interpolation handling
86 * extend flex-style library to fully support all default styles
87 * fix handling of Hebrew aleph
88 * do not assign postcodes to rivers
89 * fix string matching in PHP code
90 * update osm2pgsql (various updates to flex)
91 * fix slow query when deleting places on update
92 * fix CLI details query
93 * fix recalculation of importance values
94 * fix polygon simplification in reverse results
95 * add class/type information to reverse geocodejson result
96 * minor improvements to default tokenizer configuration
97 * various smaller fixes to documentation
101 * fix XSS vulnerability in debug view
105 * add experimental support for osm2pgsql flex style
106 * introduce secondary importance value to be retrieved from a raster data file
107 (currently still unused, to replace address importance, thanks to @tareqpi)
108 * add new report tool `nominatim admin --collect-os-info`
109 (thanks @micahcochran, @tareqpi)
110 * reorganise index to improve lookup performance and size
111 * run index creation after import in parallel
112 * run ANALYZE more selectively to speed up continuation of indexing
113 * fix crash on update when addr:interpolation receives an illegal value
114 * fix minimum number of retrieved results to be at least 10
115 * fix search for combinations of special term + name (e.g Hotel Bellevue)
116 * do not return interpolations without a parent street on reverse search
117 * improve invalidation of linked places on updates
118 * fix address parsing for interpolation lines
119 * make sure socket timeouts are respected during replication
120 (working around a bug in some versions of pyosmium)
121 * update bundled osm2pgsql to 1.7.1
122 * add support for PostgreSQL 15
123 * typing fixes to work with latest type annotations from typeshed
124 * smaller improvements to documentation (thanks to @mausch)
128 * fix XSS vulnerability in debug view
132 * switch to ICU tokenizer as default
133 * add housenumber normalization and support optional spaces during search
134 * add postcode format checking and support optional spaces during search
135 * add function for cleaning housenumbers in word table
136 * add updates/deletion of country names imported from OSM
137 * linked places no longer overwrite names from a place permanently
138 * move default country name configuration into yaml file (thanks @tareqpi)
139 * more compact layout for interpolation and TIGER tables
140 * introduce mutations to ICU tokenizer (used for German umlauts)
141 * support reinitializing a full project directory with refresh --website
142 * fix various issues with linked places on updates
143 * add support for external sanitizers and token analyzers
144 * add CLI commands for forced indexing
145 * add CLI command for version report
146 * add offline import mode
147 * change geocodejson to return a feature class in the 'type' field
148 * add ISO3166-2 to address output (thanks @I70l0teN4ik)
149 * improve parsing and matching of addr: tags
150 * support relations as street members of associatedStreet
151 * better ranking for address results from TIGER data
152 * adapt rank classification to changed tag usage in OSM
153 * update bundled osm2pgsql to 1.6.0
154 * add typing information to Python code
155 * improve unit test coverage
156 * reorganise and speed up code for BDD tests, drop support for scenes
157 * move PHP unit tests to PHP 9.5
158 * extensive typo fixes in documentation (thanks @woodpeck,@StephanGeorg,
159 @amandasaurus, @nslxndr, @stefkiourk, @Luflosi, @kianmeng)
160 * drop official support for installation on CentOS
161 * add installation instructions for Ubuntu 22.04
162 * add support for PHP8
163 * add setup instructions for updates and systemd
164 * drop support for PostgreSQL 9.5
168 * fix XSS vulnerability in debug view
172 * fix initialisation error in replication script
173 * ICU tokenizer: avoid any special characters in word tokens
174 * better error message when API php script does not exist
175 * fix quoting of house numbers in SQL queries
176 * small fixes and improvements in search query parsing
177 * add documentation for moving the database to a different machine
181 * refactor name token computation and introduce ICU tokenizer
182 * name processing now happens in the indexer outside the DB
183 * reorganizes abbreviation handling and moves it to the indexing phases
184 * adds preprocessing of names
185 * add country-specific ranking for Spain, Slovakia
186 * partially switch to using SP-GIST indexes
187 * better updating of dependent addresses for name changes in streets
188 * remove unused/broken tables for external housenumbers
189 * move external postcodes to CSV format and no longer save them in tables
190 (adds support for postcodes for arbitrary countries)
191 * remove postcode helper entries from placex (thanks @AntoJvlt)
192 * change required format for TIGER data to CSV
193 * move configuration of default languages from wiki into config file
194 * expect customized configuration files in project directory by default
195 * disable search API for reverse-only import (thanks @darkshredder)
196 * port most of maintenance/import code to Python and remove PHP utils
197 * add catch-up mode for replication
198 * add updating of special phrases (thanks @AntoJvlt)
199 * add support for special phrases in CSV files (thanks @AntoJvlt)
200 * switch to case-independent matching between place and boundary names
201 * remove disabling of reverse query parsing
202 * minor tweaks to search algorithm to avoid more false positives
203 * major overhaul of the administrator and developer documentation
204 * add security disclosure policy
205 * add testing of installation scripts via CI
206 * drop support for Python < 3.6 and Postgresql < 9.5
210 * fix XSS vulnerability in debug view
214 * fix database check for reverse-only imports
215 * do not error out in status API result when import date is missing
216 * add array_key_last function for PHP < 7.3 (thanks to @woodpeck)
217 * fix more url when server name is unknown (thanks to @mogita)
218 * commit changes to replication log table
222 * fix smaller issues with special phrases import (thanks @AntoJvlt)
223 * add index to speed up continued indexing during import
224 * fix index on location_property_tiger(parent_place_id) (thanks @changpingc)
225 * make sure Python code is backward-compatible with Python 3.5
226 * various documentation fixes
230 * switch to dotenv for configuration file
231 * introduce 'make install' (reorganising most of the code)
232 * introduce nominatim tool as replacement for various php scripts
233 * introduce project directories and allow multiple installations from same build
234 * clean up BDD tests: drop nose, reorganise step code
235 * simplify test database for API BDD tests and autoinstall database
236 * port most of the code for command-line tools to Python
237 (thanks to @darkshredder and @AntoJvlt)
238 * add tests for all tooling
239 * replace pyosmium-get-changes with custom internal implementation using
241 * improve search for queries with housenumber and partial terms
242 * add database versioning
243 * use jinja2 for preprocessing SQL files
244 * introduce automatic migrations
245 * reverse fix preference of interpolations over housenumbers
246 * parallelize indexing of postcodes
247 * add non-key indexes to speed up housenumber + street searches
248 * switch housenumber field in placex to save transliterated names
252 * add full support for searching by and displaying of addr:* tags
253 * improve address output for large-area objects
254 * better use of country names from OSM data for search and display
255 * better debug output for reverse call
256 * add support for addr:place links without an place equivalent in OSM
257 * improve finding postcodes with normalisation artefacts
258 * batch object to index for rank 30, avoiding a wrap-around of transaction
260 * introduce dynamic address rank computation for administrative boundaries
261 depending on linked objects and their place in the admin level hierarchy
262 * add country-specific address ranking for Indonesia, Russia, Belgium and
263 the Netherlands (thanks @hendrikmoree)
264 * make sure wikidata/wikipedia tags are imported for all styles
265 * make POIs searchable by name and housenumber (thanks @joy-yyd)
266 * reverse geocoding now ignores places without an address rank (rivers etc.)
267 * installation of a webserver is no longer mandatory, for development
268 use the php internal webserver via 'make serve
269 * reduce the influence of place nodes in addresses
270 * drop support for the unspecific is_in tag
271 * various minor tweaks to supplied styles
272 * move HTML web frontend into its own project
273 * move scripts for processing external data sources into separate directories
274 * introduce separate configuration for website (thanks @krahulreddy)
275 * update documentation, in particular, clean up development docs
276 * update osm2pgsql to 1.4.0
280 * ensure that wikipedia tags are imported for all styles
281 * reinstate verbosity for indexing during updates
282 * make house number reappear in display name on named POIs
283 * introduce batch processing in indexer to avoid transaction ID overrun
284 * increase splitting for large geometries to improve indexing speed
285 * remove deprecated get_magic_quotes_gpc() function
286 * make sure that all postcodes have an entry in word and are thus searchable
287 * remove use of ST_Covers in conjunction with ST_Intersects,
288 causes bad query planning and slow updates in Postgis3
293 * disable jit and parallel processing in PostgreSQL for osm2pgsql
294 * update libosmium to 2.15.6 (fixes an issue with processing hanging
295 on large multipolygons)
299 * structured select on HTML search page
300 * new PHP Nominatim\Shell class to wrap shell escaping
301 * remove polygon parameter from all API calls
302 * improve handling of postcode areas
303 * reorganise place linking algorithm, now using wikidata tag as well
304 * remove linkees from search_name and larger_area tables
305 * introduce country-specific address ranks
306 * reorganise rank address computation
307 * cleanup of partition function
308 * improve parenting for large POIs
309 * add support for Postgresql 12 and Postgis 3
310 * add earlier cleanup when --drop is given, to reduce memory usage
311 * remove use of place_id in URLs
312 * replace C nominatim indexer with a simpler Python implementation
313 * split up the huge sql/functions.sql file
314 * move osm2pgsql tests to osm2pgsql
315 * add new extratags style which imports all tags from OSM
316 * add new script for checking the import after completion
317 * update osm2pgsql, reducing memory usage
318 * use new wikipedia importance and add processing of wikidata tags
319 * add search form for details page
320 * use ExtraDataPath for country_grid table
321 * remove short_name from list of names to be displayed
322 * split up CMakeFile, so that all parts can be built separately
323 * update installation instructions for CentOS and Ubuntu
324 * add script for importing/updating multiple country extracts
325 * various documentation improvements
329 * fix security bug in /details endpoint where user input was not
334 * update osm2pgsql to fix hans during updates and lost address numbers
339 * increase required version for PostgreSQL(9.3), PostGIS(2.2) and PHP(7.0)
340 * better error reporting for out-of-memory errors
341 * exclude postcode ranges separated by colon from centre point calculation
342 * update osm2pgsql, better handling of imports without flatnode file
343 * switch to more efficient algorithm for word set computation
344 * use only boundaries for country and state parts of addresses
345 * improve updates of addresses with housenumbers and interpolations
346 * remove country from place_addressline table and use country_code instead
347 * optimise indexes on search_name partition tables
348 * improve searching of attached streets for large objects like airports
349 * drop support for python 2
350 * new scripts for importing Wikidata for importance
351 * create and drop indexes concurrently to not clash with auto vacuum
352 * various documentation improvements
357 * zoom 17 in reverse now zooms in on minor streets
358 * fix use of postcode relations in address
359 * support for housenumber 0 on interpolations
360 * replace database abstraction DB with PDO and switch to using exceptions
361 * exclude line features at rank 30 from reverse geocoding
362 * remove self-reference and country from place_addressline
363 * make json output more readable (less escaping)
364 * update conversion scripts for postcodes
365 * scripts in utils/ are no longer executable (always use scripts in build dir)
366 * remove Natural Earth country fallback (OSM is complete enough)
367 * make rank assignments configurable
368 * allow accept languages with underscore
369 * new reverse-only import mode (without search index table)
370 * rely on boundaries only for states and countries
371 * update osm2pgsql, now using a configurable style
372 * provide multiple import styles
373 * improve search when house number and postcodes are dropped
374 * overhaul of setup code
375 * add support for PHPUnit 6
376 * update test database
377 * various documentation improvements
381 * complete rewrite of reverse search algorithm
382 * add new geojson and geocodejson output formats
383 * add simple export script to export addresses to CSV
384 * remove is_in terms from address computation
385 * remove unused search_name_country tables
386 * various smaller fixes to query parsing
387 * convert Tokens and token types to class types
388 * correctly handle update when boundary object type is changed
389 * improve debug output for /search endpoint
390 * update to latest osm2pgsql and leaflet.js
391 * overhaul of /details endpoint:
392 * new class parameter when using osmtype/osmid parameters
393 * permalink to instance-independent osmtype/osmid parameter format
394 * new json output format
395 * update CentOS vagrant machine to use SELinux
396 * add vagrant scripts for Ubuntu 18.04
397 * fix build process for BSD
398 * enable running the database on a different host than the setup scripts
399 * allow to configure use of custom PHP binaries (PHP_BIN)
400 * extensive coding style improvements to PHP code
401 * more PHP unit tests for new classes
402 * increase coverage for API tests
403 * add documentation for API
407 * rework postcode handling and introduce location_postcode table
408 * make setup less verbose and print a summary in the end
409 * setup: error out when web site user does not exist
410 * add more API tests to complete code coverage
411 * reinstate key-value amenity search (special term [key=value])
412 * fix detection of coordinates in query
413 * various smaller tweaks to ranking of search interpretations
414 * complete overhaul of PHP frontend code using OOP
415 * add address rank to details page
416 * update Tiger scripts for 2017 data and clean up unused code
417 * various bug fixes and improvements to UI
418 * improve reverse geocoding performance close to coasts
419 * more PHP style cleanup (quoting)
420 * allow unnamed road in reverse geocoding to avoid too far off results
421 * add function to recalculate counts for full-word search term
422 * add function to check if new updates are available
423 * update documentation and switch to mkdocs for generating HTML
427 * fix bug in geometry building algorithm in osm2pgsql
428 * fix typos in build instructions
432 * move to cmake build system
433 * various fixes to HTML output
434 * reverse endpoint now can return geometries
435 * add support for PHP7
436 * move to on-the-fly computation of interpolations
437 * improve handling of linked places (updating)
438 * improve test framework:
439 * replace lettuce with behave
440 * use smaller database for API tests
441 * drop support for postgres < 9.1, postgis < 2.0 and PHP < 5.4
442 * make external data use optional (useful for imports without US)
443 * detect postgres and postgis versions automatically
444 * clean up query logging and remove unused tables
445 * move installation documentation into this repo
446 * add self-documenting vagrant scripts
447 * remove --create-website, recommend to use website directory in build
448 * add accessor functions for URL parameters and improve error checking
449 * remove IP blocking and rate-limiting code
450 * enable CI via travis
451 * reformatting for more consistent coding style
452 * make country search term creation part of setup
453 * update country names and country grid
454 * handle roads that cross boundaries better
455 * keep full information on address tags
456 * update to refactored osm2pgsql which use libosmium based types
457 * switch from osmosis to pyosmium for updates
458 * be more strict when matching against special search terms
459 * handle postcode entries with multiple values correctly
463 * reverse geocoding includes looking up housenumbers from Tiger data
464 * added parameter to return simplified geometries
465 * new lookup call for getting address information for OSM objects
466 * new namedetails and extratags parameters that expose the name and extratags
467 fields of the placex table
473 * drop support for postgres 8.4
474 * rewrite address interpolation
475 * switch to C++ version of osm2pgsql and rewrite tag filtering
476 * support for bridge:name and tunnel:name, man_made, junction
477 * drop way-node index (reduces database size by about 15%)
478 * add support for configuring tablespaces and webserver user
479 * better evaluation of search queries in right-to-left notation
480 * improve house number search for streets with many duplicate entries
481 * code cleanup (remove unused functions and tables)
485 * further improve ordering of results
486 * support for more lat/lon formats in search-as-reverse
487 * fix handling of GB postcodes
488 * new functional test suite
489 * support for waterway relations
490 * inherit postcodes from street to poi
491 * fix housenumber normalisation to find non-latin house numbers
492 * take viewbox into account for ordering of results
493 * pois may now inherit address tags from surrounding buildings
494 * improve what objects may participate in an address
495 * clean up handled class/type combinations to current OSM usage
500 * correct database rights for www-data
501 * add timestamps for update output
502 * load postgis via extension for postgis >= 2.0
503 * remove non-admin boundaries from addresses
504 * further improve ordering of results with same importance
505 * merge addr:postcode tags into object addresses
506 * include rank and importance in reverse geocode output
507 * replace ST_Line_Interpolate_Point with ST_LineInterpolatePoint
509 * update osm2pgsql to latest version
510 * properly detect changes of admin_level
511 * remove landuses when name is removed
516 * update to openlayers 2.12 (now custom built)
517 * update fallback OSM boundaries
518 * add support for postgresql 9.2/9.3 and postgis 2.x
519 * add structured queries
520 * add options for polygon output in various formats
521 (geojson, svg, kml, postgis text)
522 * maintenance functions for deleting objects and updating regions
523 (plcae_force_update/place_force_delete)
524 * web view for polygons that need deleting
525 * rate limiting using memcache
526 * improve layout of details page
527 * add support for boundary:postal_code
529 * improve parenting of POIs
530 * support for extract daily diffs from Geofabrik
531 * support for addresses without a street
532 (addr:place and conscription number house numbers)
533 * improve layout of word and search_name_* tables
534 * support for US ZIP+4 codes
535 * refactoring of front-end PHP code
536 * lots of smaller bug fixes
540 * delete outdated entries from location_area_country
541 * remove remaining uses of INTEGER, to allow node ids larger than 2^31