4 respuestas en este tema

[Rolphy Reyes]

Saludos.

Tal como dice el encabezado.

Enlace.

[javsolis3]

No la he usado pero empesare gracias nuevamente!!!

[Delphius]

Hoy he visto la publicación en facebook del aviso de esta liberación.
Por el aviso da a entender, que si bien es una subversión, que han habido varias mejoras muy importantes y recomiendan actualizar lo más pronto posible.
No he visto el informe de todos los cambios y mejoras pero seguramente deben ser muy destacables como para que "llamasen la atención".


Saludos,

[Marc]

Hola.

[noae][quote name="Delphius" post="56228" timestamp="1317778417"]
Hoy he visto la publicación en facebook del aviso de esta liberación.
Por el aviso da a entender, que si bien es una subversión, que han habido varias mejoras muy importantes y recomiendan actualizar lo más pronto posible.
No he visto el informe de todos los cambios y mejoras pero seguramente deben ser muy destacables como para que "llamasen la atención".


Saludos,
[/quote][/noae]

Básicamente tiene una lista enorme de correcciones de bugs (por eso es tan aconsejable actualizar), junto a unas pocas características nuevas de poca importancia.

Estas son las características nuevas :
[noae][quote]Firebird 2.5.1 Release (September 2011)                                                                                                      The months that have passed since the initial release have seen numerous Tracker issues                    tackled to address things that have been reported broken, bent or compromised in v.2.5.  As well                    as a long list of bug fixes, a few minor improvements have been made.                      In summary:                 
                    [list type=disc]
[*]                          The SQLSTATE code has been made available as a PSQL context variable, for use in                              WHEN .. exception blocks, in the same manner as GDSCODE and SQLCODE                         
[*]                          Now it is possible to write to global temporary tables in a read-only database
[*]                          Diagnostics for internal trace errors were improved
[*]                          The fbtracemgr utility will now do a periodic flush to output                         
[*]                          Performance of gbak restore at the data insertion stages                            has improved                         
[*]                          Conversions between BLOBs and other data types can now be effected                              in the API functions                         
[*]                          The Services API now supports the “metadata-only” restore                         
[*]                          A “silent install” switch has been implemented for                              make install on POSIX                         
[*]                          The unused bytes of VARCHAR values in the message buffer are now set to zero
[*]                          The actual record compression ratio is now estimated in the optimizer
[*]                          The MON$STATEMENT_ID value now stays constant among monitoring snapshots
[*]                          The SO_KEEPALIVE option on the client TCP socket will now be set, as a measure to                            guard against aggressive timing out of sockets by newer Windows systems                         
[*]                          Lock Manager can now cancel waits that become interminable
[*]                          A platform port of v.2.5.1 for HPPA has been done for both Linux and Alpha[/l][/l][/quote][/noae]
               
Y estas son las correcciones de bugs :

[noae][quote]Firebird 2.5.1 Release                                                                                                                          Core Engine/APIServer CrashesData Manipulation LanguageCommand-line UtilitiesDatabase Monitoring/AdministrationServices ManagerPOSIX-Only BugsWindows-Only BugsRemote Interface/API                                  The following improvements and bug fixes were reported as fixed prior to the                    v.2.5.1 release:               
                                                                                                  Core Engine/API                                                                                                                      (CORE-3537)                                Improvement :: The                        “undoing” on transaction rollback of changes made to global temporary tables created with the                        ON COMMIT DELETE ROWS option was unnecessary and was removed.                   
implemented by V. Khorsun
    ~ ~ ~   
(CORE-3536)                                Improvement :: Garbage                      collection in global temporary tables was being delayed unnecessarily by active transactions in                        other attachments.                   
implemented by V. Khorsun
    ~ ~ ~   
(CORE-3457)                                Improvement :: The                        temporary space manager has been made more optimal with regard to small chunk allocations.                   
implemented by D. Yemanov
    ~ ~ ~   
(CORE-3323)                                                      Improvement :: Lock Manager was provided with the                      capability to cancel waiting.                   
Considering the following example:
[pre]  tx1: update table t ... where id = 1    tx2: update table t ... where id = 1          [/pre]If transaction tx2 is in WAIT mode, it will wait for the end of tx1 forever and                      this wait cannot be broken using either a DELETE FROM MON$xxx or a                      fb_cancel_operation request.                   
The improvement provides Lock Manager with the ability to break such                      interminable waits.                   
implemented by V. Khorsun
    ~ ~ ~   
(CORE-3295)                                                      Improvement :: Estimate the actual record compression ratio in                        the optimizer, thus allowing a more precise guess about table cardinalities (number of stored                        records).                   
implemented by D. Yemanov
    ~ ~ ~   
(CORE-3560)                                The V.2.5 Classic Server was using                      more memory than V.2.1.5 when caching metadata.                   
fixed by A. Peshkov
    ~ ~ ~   
(CORE-3549)                                Database corruption could occur at                      the end of a session, throwing the error “page xxx is of wrong type expected 4                          found 7”                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3547)                                A floating-point negative zero was not                      the same as a positive zero in indexes.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3535)                                The write target of a dirty page could                      be undefined if an error occurred when the nbackup state was changed.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3533)                                On SuperServer, a memory leak would                        occur if a connection was terminated without explicitly releasing the handles for all the                        statements prepared and cached by the client application.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3532)                                The server would hang on starting                      a new session when a trace was running.                   
fixed by A. Peshkov
    ~ ~ ~   
(CORE-3525)                                Autonomous transactions were wrongly                      inheriting the run-time flags of the "parent" transaction.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3515)                                Index corruption could occur under                      a complex set of conditions that was causing index updates to get out of synch and lose                      entries. Validation would pick it up and write the "missing entries" message into                      the firebird.log                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3512)                               
fixed by A. PeshkovThe server could hang when a trace was                      running.                   
    ~ ~ ~   
(CORE-3509)                                ALTER PROCEDURE was permitting the                      addition of a parameter with the same name as an existing one.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3502)                                DROP VIEW was ignoring the existing                      non-column dependencies.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3494)                                Attach would fail anyway, after                      a shutdown had been rejected by the handler installed in fb_shutdown_callback().                   
fixed by A. Peshkov
    ~ ~ ~   
(CORE-3491)                                Altering a TYPE OF COLUMN parameter                      in a PSQL module would affect the original column.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3461)                                DDL operations would fail after                      a backup and restore.                   
fixed by A. Peshkov
    ~ ~ ~   
(CORE-3443)                                Race conditions could occur during                      UDF library lookup.                   
fixed by A. Peshkov
    ~ ~ ~   
(CORE-3418)                                Database trigger created as INACTIVE was                      coming up active.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3398)                                GRANT ADMIN ROLE was not being                      accepted.                   
fixed by A. Peshkov
    ~ ~ ~   
(CORE-3397)                                Unresolved symbols were found in intl                      and trace libraries.                   
fixed by A. Peshkov
    ~ ~ ~   
(CORE-3394)                                Failed attempt to violate a unique                        constraint could leave an unnecessary “lock conflict” error in the status                        vector.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3341)                                Events posted from inside an autonomous                        transaction could get lost and never be delivered to the listening client.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3340)                                In an autonomous transaction with an                      empty exception handler, duplicate values could be inserted into a primary of unique key column,                      leading to an unrestorable backup.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3327)                                The thread pool in a network server                      could create more threads than were necessary.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3326)                                A fast mutex could be left in locked                      state by a dead process.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3325)                                Under high load it was possible that                      a new process would fail to map shared memory.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3315)                                The audit plugin would record                      a second *FAILED* EXECUTE_STATEMENT_FINISH after a “normal” one.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3314)                                Dependencies were not being removed                      after a procedure and the table it depended on were dropped in the same transaction.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3312)                                The join plan was sub-optimal when the                      slave table depended on the master one by means of the OR predicate.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3306)                                An invariant sub-query was                      being treated as variant,  causing multiple invocations of a nested stored procedure.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3283)                                A bad plan was produced with a query                      using a LEFT OUTER JOIN in a sub-query.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3282)                                EXECUTE STATEMENT was using the wrong                      character set when parsing the SQL text argument.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3266)                                A race condition could occur                      between the asynchronous service detach request and a running user trace service.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3256)                                Error “request depth exceeded”                      could appear while preparing a select query against a view with explicit plan.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3237)                                Stored procedures were being compiled                      too slowly.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3207)                                AccessViolationException inside                      FB_965910463_Class.isc_start_multiple on beginning transaction.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3205)                                isc_dsql_exec_immed2() was not                      returning error codes isc_stream_eof and isc_sing_select_err when they occurred.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3188)                                Error “Page 0 is of wrong type                          (expected 6, found 1)” was occurring unexpectedly.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3176)                                A view with a column derived from a                      sub-query would join the sub-query output to the table without using an index.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3168)                                The exclude_filter was not working for                      the <services> section of the trace facility.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3157)                                Adding a parameter description to a                      stored procedure could lead to heavy memory consumption and protracted execution time.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3151)                                Memory allocated in sqlda_sup could be                      left unreleased in some cases.                   
fixed by A. Peshkov
    ~ ~ ~   
(CORE-3148)                                Dangerous code was discovered in                      SQZ_apply_differences.                   
fixed by D. Kovalenko, A. Peshkov
    ~ ~ ~   
(CORE-3140)                                Comments for parameters were not being                      preserved after altering procedures.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3137)                                Partial rollback was possible for a                      selectable procedure that was modifying data.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3131)                                WIN1257_LV (Latvian) collation was wrong                      for the four letters A, E, I and U.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3125)                                An access violation was occurring                      in the routine Worker::shutdown().                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3058)                                New generators were being created with                      wrong values when more than 32,767 generators had been created previously.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3029)                                Bugcheck “Too many                          savepoints (287)” at rollback after an exception from EXECUTE BLOCK that contained                      an exception handler.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3024)                                Error “no current record for                          fetch operation” would occur after ALTER VIEW.                   
fixed by A. Peshkov
    ~ ~ ~   
(CORE-2835)                                Natural was being used to select, when the                      primary key index should have been used.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-2827)                                Prepare was very slow for complex                      interrelated metadata when the operation being prepared indirectly involved many triggers.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-2709)                                Indexed reads in                      a compound index with NULLs were excessive.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-1752)                                Results of a join with different collations                      exhibited variations according to the execution plan used.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-1274)                                Results would be wrong when                      PLAN MERGE was chosen and the data types of the equality predicate arguments were different.                   
fixed by D. Yemanov
    ~ ~ ~   
                                                                                                                    Server Crashes                                                                                                                      (CORE-3554)                                Passing an empty SQL query                        remotely could crash the server during the prepare or throw an                      incorrect parsing error.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3557)                                The server could crash while                        preparing a query against a table that was in the process of being dropped.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3524)                                The server would crash while recompiling                      a stored procedure that was in use.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3503)                                ALTER VIEW would crash the server                      if the new version had an artificial (aggregate or union) stream at the position of a regular                      context in the older version.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3477)                                The server would crash whenever                      non-existent SQL parameters were passed to it.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3440)                                The server would crash if                      isc_que_events() had queued 0 events.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3247)                                The server was crashing with                      BLOBs in character set UTF8.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3419)                                Server could hang or crash when an                      autonomous transaction was rolled back.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3400)                                Server crashes were occurring                        frequently on FreeBSD8.2R.                   
fixed by A. Peshkov
    ~ ~ ~   
(CORE-3374)                                The server could crash or corrupt data                      if SELECT WITH LOCK was issued against records not in the latest format.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3320)                                A certain MERGE syntax was able to crash                      the server. Details are not published.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3255)                                The server could crash using views                      with GROUP BY.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3219)                                Trace manager would crash the server                      with DSQL_unprepare.                   
fixed by V. Khorsun
    ~ ~ ~   
(CORE-3217)                                The server would crash inside the Lock                      Manager when multiple connections were attaching or detaching simultaneously.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3202)                                execute_immediate API call family could                      crash the remote server.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3180)                                ALTER VIEW with non-matched columns in                      the declaration and selection would crash the server.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3138)                                An internal error or a crash would occur                      when accessing any MON$ table after altering its structure.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3064)                                Using both the identifier of a procedure                      and its alias in an explicit plan would crash the server.                   
fixed by D. Yemanov
    ~ ~ ~   
                                                                                                                    Data Manipulation Language                                                                                                                      (CORE-3523)                                SIMILAR TO was giving false matches                      on descending ranges.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3493)                                Adding a value to a timestamp earlier                      than '16.11.1858 00:00:01' would throw the error "value exceeds the range for valid                      timestamp".                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3489)                                Blob transliteration sometimes failed to                      happen inside a union.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3479)                                The ASCII_VAL() function would raise                      an error for empty strings instead of returning 0.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3355)                                Comparison of DATE and TIMESTAMP would                      be wrong if an index was used.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3353)                                Predicate (blob_field LIKE ?) was                      describing the parameter as VARCHAR(30) rather than as BLOB.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3335)                                Internal wrapping was occurring                      for the multi-byte blob SUBSTRING function and its boundary arguments, causing wrong results.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3311)                                The error “data type unknown”                      would be thrown while preparing an UPDATE or DELETE statement with a parameterized ROWS                      clause.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3302)                                DISTINCT aggregates were returning wrong                      (duplicated) data.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3277)                                The RIGHT() function was giving a                      wrong result for varchars of character set UTF8.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3245)                                A SUBSTRING() operation on a long text                      BLOB would return a BLOB containing only 32767 characters if the optional third argument was                      not present.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3244)                                The result of POSITION() was wrong for an                      empty string ('') if the third argument was present.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3233)                                LIKE, STARTING and CONTAINING would                      fail if the second operand was 32KB or more.                   
fixed by A. dos Santos Fernandes, D. Yemanov
    ~ ~ ~   
(CORE-3228)                                The RIGHT() function would fail                      with multi-byte text blobs more than 1024 bytes in length.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3227)                                The ASCII_VAL() function would fail                      if the argument contained multi-byte characters anywhere.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3222)                                A view with “WITH CHECK OPTION”                      did not like a TRIM() function call in the WHERE clause.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3211)                                String truncation would occur when                      selecting from a view that contained a NOT IN condition.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3210)                                Error “no current record for fetch                          operation” would occur unexpectedly in a SELECT query.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3208)                                Recursive queries were exhibiting                      significant memory leaks.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3203)                                UPDATE OR INSERT with RETURNING                      would cause an “Invalid Cursor” error.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3173)                                Selecting from a stored procedure that                      contained two Common Table Expressions, the second with a GROUP BY clause, as well as an inner                      join, would wrongly return nothing in the result.                   
fixed by D. Yemanov
    ~ ~ ~   
(CORE-3164)                                Parameterized requests involving BLOB fields would                      fail when the client was connected using character set UTF8.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3141)                                The last column in a view was being returned                      as NULL, even when it should contain a value.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
(CORE-3091)                                The built-in function POWER(X, Y) did not                      work when the X argument was negative and the Y value was a scaled numeric of scale 0.                   
fixed by A. dos Santos Fernandes
    ~ ~ ~   
                                                                                                                    Command-line Utilities                                                                                                                                                                                                                                    gbak                                                                  &

[Delphius]

¡Con razón!


Y tal parece que también han actualizado la documentación para ajustarse a esta nueva sub versión.


Saludos,