30.07.2021 14:10, Ivan Přenosil wrote:
Enabling replication of table by
ALTER TABLE tab ENABLE PUBLICATION;
is replicated, while
ALTER DATABASE INCLUDE TABLE tab TO PUBLICATION;
is not replicated.
Is it bug or by design?
ALTER DATABASE commands are not replicated, because they may include
20.07.2021 14:50, Mark Rotteveel пишет:
According to the Firebird Language Reference[1], the
RDB$RELATION_FIELDS.RDB$UPDATE_FLAG signals whether a column is a
computed column (0) or a regular column (1) (or I guess, updatable (1)
vs not-updatable (0)).
Updatable vs non-updatable is correct.
08.06.2021 11:03, Gabor Boros wrote:
I can send (to who and which address?) a test case privately which
demonstrate the problem. The generator's new value used at insert, the
new record replicated but the generator's value not.
Please send it to me: firebi...@yandex.ru
Dmitry
31.05.2021 22:21, Adriano dos Santos Fernandes wrote:
As the second, unrelated change, shorten EXECUTE
PROCEDURE/STATEMENT/what-else OBJECT to EXEC OBJECT and decide about
object type from context - provided it will not raise the hell in btyacc.
I do not understood what you mean.
31.05.2021 17:49, Dimitry Sibiryakov пишет:
31.05.2021 16:19, Alex Peshkoff via Firebird-devel wrote:
Dimitry, can you provide standard syntax for others to compare?
https://ronsavage.github.io/SQL/sql-2003-2.bnf.html#call%20statement
As far as I see, it's the same as EXECUTE PROCEDURE
31.05.2021 18:38, Mark Rotteveel wrote:
What is the real problem you're trying to solve? The current syntax is
20+ years old, and I think this is one of the first times (if not the
first time) I hear complaints about its verbosity, or a need to
selectively obtain output columns.
Blame me
22.05.2021 10:06, Mark Rotteveel wrote:
The majority of our users probably still use the old API, either
directly or indirectly. Given the undocumented state of the new API, I
suspect it does not see a lot of usage, nor will it unless that changes.
Introducing features that only benefit
22.05.2021 09:58, Mark Rotteveel wrote:
In any case, I think adding 1 to a length to obtain the actual length is
a bad idea even if it is not exposed to the outside. Doing that is a
great way to introduce increased risk of off-by-one errors and buffer
overflows; the risk of that increases
20.05.2021 18:54, Dimitry Sibiryakov wrote:
getConnectionID() returns signed 64 bits integer. Have anybody seen
negative values there?
Given that it's signed inside the header page, it may be theoretically
possible ;-)
Dmitry
Firebird-Devel mailing list, web interface at
19.05.2021 14:40, Mark Rotteveel wrote:
Dialect 1 was already deprecated with InterBase 6.0. I think people will
continue to use dialect 1 as long as Firebird supports it, though to be
honest, I would guess that most users of dialect 1 aren't even on recent
versions of Firebird. We should
19.05.2021 13:32, Adriano dos Santos Fernandes wrote:
v4 is not yet released. We may deprecate QLI in v4 already and remove in
master.
If it's just a matter of adding a paragraph to the release notes, then I
may agree to deprecate QLI in v4.
Dmitry
Firebird-Devel mailing list, web
19.05.2021 12:53, Dimitry Sibiryakov пишет:
It is what SET BIND is for.
Not really. It's OK to replace DECFLOAT with DOUBLE, but what to do with
simple arithmetics where 1/1 returns INT instead of DOUBLE? Do you want
to replace all returned INTs?
Dmitry
Firebird-Devel mailing list, web
19.05.2021 12:34, Dimitry Sibiryakov wrote:
AFAIR these discussions, people wanted new features in it (such as
BIGINT) and the only named advantage was double precision arithmetic.
This is covered by INT128/DECFLOAT arithmetic in v4 if I'm not mistaken
Correct.
so we can "support"
19.05.2021 12:27, Alex Peshkoff via Firebird-devel wrote:
UDF was been deprecated in FB 4.
Could we remove them from master for v5?
Doubt that. UDF is too actively used feature, lot of users will not be
happy.
Generally, it's OK to remove already deprecated features in the next
major
18.05.2021 14:36, Omacht András wrote:
Ok, let's see the following example:
Master:
CREATE SEQUENCE TMP_GEN_2;
commit;
CREATE TABLE TMP_TABLE (
VC VARCHAR(10)
);
commit;
CREATE OR ALTER TRIGGER TMP_TABLE_BIU0 FOR TMP_TABLE
ACTIVE BEFORE INSERT OR UPDATE POSITION 0
as
declare variable
18.05.2021 13:52, Omacht András wrote:
FB4 - 4.0.0.2489
We tried the following in both (sync, async) modes:
Master:
CREATE SEQUENCE TMP_GEN_1;
commit;
(The sequence was created fine in replicas.)
Master:
select gen_id(tmp_gen_1, 1) from rdb$database;
-- Result: 1 - OK
commit;
select
18.05.2021 12:38, Nils Bödeker wrote:
These settings are not necessary for synchronous-only setup. Moreover, removing
them will actually cause your issue to disappear ;-) But this way we'd never
find this bug ;-)
So that I get it right.
There is no bug in the synchronous replications.
17.05.2021 16:29, Nils Bödeker wrote:
Under parallel writing we get the following error in the replication log
(sometimes, not always):
FIREBIRD-4 (origin) Sat May 15 12:11:23 2021
Database: D:\NBS_DATABASE\MINICMS.GDB
ERROR: Log file
17.05.2021 12:47, Dimitry Sibiryakov wrote:
Yes, the first wins. I'm wondering, however, wouldn't it be better to
raise an error for multiple configurations detected for the current
database? Or maybe just log a warning and keep using the first entry.
Synchronous replication can work with
17.05.2021 12:19, Mark Rotteveel wrote:
This is an impossible configuration: you're defining two sets of
configuration for `C:\R\DB\ORIGINAL.FDB`. You should have only one entry
for a database in databases.conf.
Correct.
I'm not sure if the first or the last one wins, but based on your
17.05.2021 12:02, Gabor Boros wrote:
Only one mode works in the same time per database?
You may specify both sync and async settings inside the same (single)
database entry.
Dmitry
Firebird-Devel mailing list, web interface at
https://lists.sourceforge.net/lists/listinfo/firebird-devel
09.05.2021 20:20, Adriano dos Santos Fernandes wrote:
What I want to ask about if it's clear/good is that DefaultTimeZone set
in client does not implicitly pass isc_dpb_session_time_zone in the
attachment.
For me such behavior looks suspicious (i.e. a bug to be precise).
DefaultTimeZone set
09.05.2021 18:25, Roman Simakov wrote:
I don't remember why I did so but maybe it's better to allow drop sql
security for any object type and remove the wrong statement at all.
What is the point in dropping, if AFAIU the legacy (lacking sql
security) mode is actually the same as "definer"?
05.05.2021 16:55, Adriano dos Santos Fernandes wrote:
As I understand, 255 limit for table version is due to record format
number embeded in records.
We had the same situation with 32-bit transaction numbers and it's now
possible to have 64-bit transactions using rhd_long_tranum flag.
Which
02.05.2021 18:42, Mark Rotteveel wrote:
I attempted to reduce the time to run the Jaybird tests, but I noticed -
against a recent Firebird 4 snapshot - that disabling forced write on
database creation did not make a huge difference (30s less, from 4m to
3m30s), while explicitly enabling
01.05.2021 14:21, Nils Bödeker wrote:
Do you have any plans for a Firebird 4 final release date?
I hope to have the codebase tagged during the next week. After that
we'll need to compose the release notes and make the builds, it will
take some more days.
We changed for FB 4 from pagesize
30.04.2021 17:25, Dimitry Sibiryakov wrote:
Currently replication drops blob type.
Sounds like an oversight, stream blobs are surely not deprecated.
Dmitry
Firebird-Devel mailing list, web interface at
https://lists.sourceforge.net/lists/listinfo/firebird-devel
13.04.2021 12:22, Dimitry Sibiryakov wrote:
I'd like for example to set the database from read/write mode to read
modus and vs.
You can get current database mode via API isc_database_info() call
and set it using DPB tag on attach.
Or, alternatively, via Services API.
Dmitry
08.04.2021 13:43, Pro Turm wrote:
how would you communicate with a fb 4.0 server with e.g. an external
component, which would like to control the FB 4.0 server replication
behavior ? Are there interfaces for it, protocols (e.g. tcp etc.) ?
Please describe what replication behaviour you'd
: Bug
Components: Engine
Affects Versions: 4.0 RC 1, 4.0 Beta 2
Reporter: Dmitry Yemanov
Test case:
select 1 from rdb$database where timestamp '01.01.2021 13:00:00 +03:00' =
timestamp '01.01.2021 12:00:00 +02:00';
-- Values are surely equal:
CONSTANT
05.04.2021 14:36, Mark Rotteveel wrote:
As asked on firebird-support by Luciano Rodrigues Nunes Mendes, the
Firebird 3.0 snapshots on Windows are missing at
http://web.firebirdsql.org/download/snapshot_builds/win/3.0/
Can someone investigate why and fix the issue?
Windows build is currently
29.03.2021 22:47, Ivan Přenosil wrote:
I have FB4 database created using Backup/Restore from FB2.5 database
where attempt to drop column will raise this error:
ALTER TABLE TAB DROP STAV;
Statement failed, SQLSTATE = 22000
[335544351] unsuccessful metadata update
[336397287] -ALTER TABLE TAB
22.03.2021 11:14, Mark Rotteveel wrote:
It turns out that RETURNING is not a reserved word, which allows
statements like:
```
SQL> create table returning (returning boolean);
SQL> insert into returning (returning) values (true) returning returning
returning;
RETURNING
=
SQL>
06.03.2021 14:53, Mark Rotteveel wrote:
Now I wonder, what would be necessary to introduce this alternative?
Would this be a matter of adding the right alternative and adding the
order by and offset/fetch info to the RseNode returned by the query
expression, or would more be needed?
I'm
25.02.2021 20:37, Lucas Schatz wrote:
So I guess that for now the "safest" option is async (related to network
instability) , correct?
Right.
Just a curiosity, is it possible right now to enable a multiple replica
system in async mode, by defining multiple log_archive_directory?
If not, I
25.02.2021 17:10, Lucas Schatz wrote:
I im testing for a week, and I think that the current Synchronous mode
is awesome but too fragile to a daily use as a security feature, if I'm
not mistaken at any network failure or replica restart you'll be prone
to inconsistency in the replica data,
25.02.2021 17:27, Dimitry Sibiryakov wrote:
AFAIK replication server resend transactions using control files,
replication sequence and DB UUID received from INF call. INF call for
remote database is ok, control files can be in local FS as normally and
replication packet is sent through
25.02.2021 15:40, Dimitry Sibiryakov wrote:
I mean write journals on master side and then apply them from master
side directly to remote replica database when it is available.
Again, we'll need to track the oldest segment marker -- from where to
resend (and what can be deleted).
Dmitry
25.02.2021 15:07, Dimitry Sibiryakov wrote:
25.02.2021 13:04, Dmitry Yemanov wrote:
25.02.2021 13:39, Dimitry Sibiryakov wrote:
Is there option to set "database = inet://remote:database" in
replication.conf for a replica?
Nope. Something like that is considered for futur
25.02.2021 13:39, Dimitry Sibiryakov wrote:
Is there option to set "database = inet://remote:database" in
replication.conf for a replica?
Nope. Something like that is considered for future versions.
Dmitry
Firebird-Devel mailing list, web interface at
25.02.2021 09:27, Peter Gore wrote:
Your suggestion still did not work. What _does _seem to work is if I
make log_source_directory = log_directory (NOT log_archive_directory as
advised). *Please comment*.
This is wrong.
I am getting errors however in the replication log
TOSHPJG (replica)
24.02.2021 18:14, Lucas Schatz wrote:
Can someone reproduce it too? If yes, I'll open a report
I can, thanks. Please post to the tracker.
Dmitry
Firebird-Devel mailing list, web interface at
https://lists.sourceforge.net/lists/listinfo/firebird-devel
24.02.2021 15:55, Lucas Schatz wrote:
Hi, I was experiencing this hanging here too, when using local
synchronous replication the server hangs during the disconnection, maybe
open a bug report?
If you have a reproducible scenario - sure, feel free.
Dmitry
Firebird-Devel mailing list, web
24.02.2021 11:13, Martin Wong wrote:
I am trying to test the new Firebird 4 synchronous replication on
windows 7 64bit machine with the following configurations:
#databases.conf
master = d:\fb4\master.fdb
replica = d:\fb4\replica.fdb
#replication.conf
database = master
{
sync_replica =
24.02.2021 09:52, Peter Gore wrote:
database = D:\Users\Peter\Documents\FB4Data\FB4TEST.FDB
{
log_directory = D:\Apps\FB4\FB4Test\chlog
log_archive_directory = D:\Apps\FB4\FB4Test\archlog
log_archive_timeout = 10
}
database = D:\Users\Peter\Documents\FB4Data\FB4TESTReplica.FDB
{
23.02.2021 12:30, Mark Rotteveel wrote:
Dmitry, your push with the change was a bit messy
Yes, I've noticed that :-(
could you rebase
before pushing if your local master diverges from the remote master.
Forgot about that, sorry.
Dmitry
Firebird-Devel mailing list, web interface at
23.02.2021 11:57, Peter Gore wrote:
Here's my new config
database = D:\Users\Peter\Documents\FB4Data\FB4TEST.FDB
{
log_directory = D:\Apps\FB4\FB4Test\chlog
log_archive_directory = D:\Apps\FB4\FB4Test\archlog
log_archive_timeout = 10
verbose_logging = true
}
database =
23.02.2021 09:31, Peter Gore wrote:
As a test I have tried to get asynchronous replication working. Is this
even possible when Master and Replica is on the same Server?
Sure. Even using a single FB instance for both.
I have two
databases; FB4Test.fdb and FB4TestReplica.fdb o the same
22.02.2021 13:40, Dimitry Sibiryakov wrote:
Currently Firebird uses rule "CHAR(X)+CHAR(Y)=>CHAR(MAX(X,Y))".
Theoretically it can be changed into "CHAR(X)+CHAR(Y)=>CHAR(X) if
X==Y and CHAR(X)+CHAR(Y)=>VARCHAR(MAX(X,Y)) if X!=Y".
But what SQL standard says about it?
IIRC, current
21.02.2021 04:57, Lucas Schatz wrote:
async replication is working fine and created 5 chlog journal files
When I run rsync, Firebird successfully created/excluded everything, and
I noted that Firebird deletes the journal files... and created a "UUID
control file"
So, how I was curious I
20.02.2021 00:43, Lucas Schatz wrote:
The async replication works correctly when both servers are running, the
journaling files are created and moved from log_directory to
log_archive_directory and the data is replicated almost instantly to
the replica server. But when I stop the replica
18.02.2021 19:46, Ivan Přenosil wrote:
When I have this in fbtrace configuration:
log_warnings = 1
log_errors = 1
include_gds_codes = 335544778
then I never get warnings. It seems that include_gds_codes-filter is
applied to warnings too.
And it sounds logical, as
15.02.2021 10:26, Mark Rotteveel wrote:
The Firebird 4.0 RC 1 version in tracker is still marked as unreleased.
Can someone with sufficient rights mark it as released (on February 1st)?
Done, thanks.
Dmitry
Firebird-Devel mailing list, web interface at
08.02.2021 17:02, Vlad Khorsun wrote:
I agree, except of usage of first entry of TempDirectories for
GTT\temporary TS -
it was never used in this way and I see no good reason to do it now.
OK.
Dmitry
Firebird-Devel mailing list, web interface at
03.02.2021 16:29, Dmitry Starodubov wrote:
As it's described in CORE-1241, configuration parameter
"TempDirectories" does not affect a place where some temporary files are
stored. Instead they are stored in the path given by the FIREBIRD_TMP
environment variable and it's not very obvious.
02.02.2021 16:47, Mark Rotteveel wrote:
A policy
that says for example that Firebird 4.0 is the last to provide x86-32
builds might be good enough.
I'd suggest to keep x86 builds for the v4.x series and remove them from
v5 onwards (but keep the codebase compilable on 32-bits). Whether v5
01.02.2021 17:51, Dmitry Yemanov wrote:
is there a release date for Firebird 4.0 RC and the Final Version ?
RC1 is expected to be out in a couple of days.
And it's just been published:
https://firebirdsql.org/en/firebird-4-0-0-rc1/
Dmitry
Firebird-Devel mailing list, web interface
01.02.2021 17:39, Nils Bödeker wrote:
is there a release date for Firebird 4.0 RC and the Final Version ?
RC1 is expected to be out in a couple of days.
Dmitry
Firebird-Devel mailing list, web interface at
https://lists.sourceforge.net/lists/listinfo/firebird-devel
29.01.2021 19:55, Dimitry Sibiryakov wrote:
Reset of replica mode to NONE after backup-restore is a feature of a
bug?
Bug, methinks. Please create a ticket.
Dmitry
Firebird-Devel mailing list, web interface at
https://lists.sourceforge.net/lists/listinfo/firebird-devel
23.01.2021 14:20, Dimitry Sibiryakov wrote:
Is there a way to create a local, non-replicated table in a
replicated database?
I.e. neither "create table" itself nor following DML must be replicated.
CREATE TABLE is currently always replicated. The table data may be
excluded though. And
23.01.2021 12:31, Mark Rotteveel wrote:
Going over the Firebird 4 release notes, I was wondering if there is a
isc_database_info / Attachment::getInfo() call to determine if a
database is standalone (no replication), a primary, or a replica, and if
a replica, what its replication mode
20.01.2021 16:03, Dimitry Sibiryakov wrote:
I agree that config parameters hardly can be not in ASCII but I
cannot imagine them to be longer that 10-20 characters. It would be very
hard to type such long settings without errors and my imagination is not
enough to suggest a meaningful name of
30.12.2020 19:44, Ivan Přenosil wrote:
Perhaps problem with Windows version only?
Does everything work OK if you uncomment (without changing) the
AuthClient setting in the v4 firebird.conf ?
Yes, with this change the connection works.
Thanks for confirmation. The fix was already
30.12.2020 18:16, Ivan Přenosil wrote:
Perhaps problem with Windows version only?
Does everything work OK if you uncomment (without changing) the
AuthClient setting in the v4 firebird.conf ?
Dmitry
Firebird-Devel mailing list, web interface at
30.12.2020 18:16, Ivan Přenosil wrote:
Perhaps problem with Windows version only?
Confirmed for Windows. Builds <= 2272 are OK, >= 2276 are broken. I'm
investigating.
Dmitry
Firebird-Devel mailing list, web interface at
https://lists.sourceforge.net/lists/listinfo/firebird-devel
22.12.2020 16:57, Ivan Přenosil wrote:
Is RC1 already frozen?
Or is there any chance that
http://tracker.firebirdsql.org/browse/CORE-6382 will go there?
You did great job by fixing that problem and it helps a lot, but I found
one special case where it still fails.
Not frozen yet, but the
17.12.2020 13:51, Jiří Činčura wrote:
what are the release plans for Firebird 4 RC1/2? Is the official RC1 going to
happen anytime soon?
Yes, RC1 is being prepared for release.
Dmitry
Firebird-Devel mailing list, web interface at
13.11.2020 01:33, Adriano dos Santos Fernandes wrote:
I labeled it (and RDB$BLOB_UTIL) for FB 5 (or next version after 4) as
For sure it's worth targeting at v4.1, not v5. Minor ODS changes are
allowed in minor releases.
FB 4 was considered frozen for new development, specially for ODS
09.11.2020 17:43, Nils Bödeker wrote:
is there a release date for Firebird 4.0 RC?
Not a particular date, but likely in December.
Dmitry
Firebird-Devel mailing list, web interface at
https://lists.sourceforge.net/lists/listinfo/firebird-devel
20.10.2020 13:58, Vlad Khorsun via Firebird-devel wrote:
AFAIU, it was discussed here in February 2018, thread "Valid date or
not".
I've re-read that thread quickly and I saw nor final decision, nor
proposition
to change (or break) rules for traditional (legacy) date\time types
(without
20.10.2020 13:40, Roman Simakov wrote:
Because documentantion here
(https://firebirdsql.org/en/firebird-date-literals/) says they are valid
separators.
Keep in mind, that link is an excerpt from Helen Borrie's Firebird Book
from 2004 (Firebird 1.5 era). It is descriptive of what worked at
05.10.2020 22:23, Max Dittrich wrote:
Are there any updates on the release schedule for Firebird 4?
For RC1 I see only fixed issues, so when do you think will you release
RC1 officially to get a greater audience for final testing and what is
your rough estimation when do we could expect the
28.09.2020 20:44, Adriano dos Santos Fernandes wrote:
I have thought on that possibility since the start and the conflict
points I see are:
- Name of RDB$PROFILE_STATS - perhaps RDB$PROFILE_PSQL?
Maybe we may leave it as is. And add RDB$PROFILE_CURSOR_STATS.
The question is how they should
28.09.2020 13:27, Adriano dos Santos Fernandes wrote:
I would like something about cursors too and I think in the future it
can be added.
I believe this should be a top-priority thing. I doubt many users
implement heavy logic in PSQL without queries, so per-statement stats is
mostly usable
28.09.2020 03:11, Adriano dos Santos Fernandes wrote:
I have a doubt about permissions.
If user can call procedure P1 but cannot himself call procedure P2, but
procedure P1 can call P2, if profiler should not collect data from P2.
Current it does.
Given that user can see other statistics
28.09.2020 09:41, Simonov Denis via Firebird-devel wrote:
2. The minimum and maximum time is certainly interesting, but it would
be nice to display the average execution time.
RDB$ACCUMULATED_TIME / RDB$COUNTER
Dmitry
Firebird-Devel mailing list, web interface at
28.09.2020 09:41, Simonov Denis via Firebird-devel wrote:
1. Why is statistics limited to procedure / function / trigger and block
only? But what about queries and cursors inside blocks. Their execution
times are much more interesting.
RDB$PROFILE_STATS contains statement ID and
25.09.2020 01:08, Dimitry Sibiryakov wrote:
Is impossibility to use IReplicator interface the only practical
difference between REPLICA_NONE and REPLICA_READ_WRITE modes?
Currently, yes (IIRC). Although there may be others in the future, I
suspect.
Dmitry
Firebird-Devel mailing list,
19.09.2020 09:41, Dmitry Yemanov wrote:
Setup your Windows to generate crash dumps, repeat the crash, find the
created memory dump, upload it somewhere and share a link.
But before that, please wait for the new build (will appear over the
weekend). The current build 2203 is buggy in regard
19.09.2020 06:09, Eliezer Riani wrote:
My FB4.0x64 Server (WI-V4.0.0.2187) has crashed and started an
automatic "service stop"-"service start" infinite loop, as you can
watch on this video:
"https://www.awesomescreenshot.com/video/1029749?key=5f3d9d5ac5e6d100e9e77756860c6f6d;.
The problem
16.09.2020 13:48, Dimitry Sibiryakov wrote:
A long time ago there were talks about a new record format:
self-descriptive, length encoded.
Do anybody still sleep on it?
AFAIK, nope. I don't even remember such talks ;-)
You'd better start this thread with the problem you're willing to solve
12.09.2020 14:03, Mark Rotteveel wrote:
I just noticed that SEC$DB_CREATORS (in Firebird 3) does not restrict
visibility. Any user can see all users granted the CREATE DATABASE
privilege. I would expect only SYSDBA (or users with RDB$ADMIN active)
to be able to see all users.
Is that
02.09.2020 00:50, Dimitry Sibiryakov wrote:
Why is the matcher created for each attachment separately? Isn't one
per database enough?..
By design, it belongs to the replication manager (i.e. per database).
But something went wrong ;-) IIRC, I was worried about synchronization
overhead.
Alex et al,
I believe that many of you are all too narrowly linking HASH and
whether it is appropriate for cryptography.
What about particular CRC32 - it produces output value with length just
4 bytes, even using SHA1 with 20 bytes returned was accepted as security
risk in some places (like
06.08.2020 18:12, Dimitry Sibiryakov wrote:
Is it ok that replication merge all BLOB segments into one and cannot
work with BLOBs larger than 4G?
Not really. This is a limitation of the original implementation and
apparently nobody stepped on it yet ;-) Worth improving for sure.
Dmitry
02.08.2020 13:14, Mark Rotteveel wrote:
Firebird 4 introduces the function CRC32 which generates a CRC32 hash,
while Firebird 4 also introduces an extension to the HASH function to
add (cryptographic) hashes.
Would it make sense to remove CRC32 as a standalone function and instead
add it to
17.07.2020 19:24, Vlad Khorsun wrote:
For this particular case the following solution may be used instead:
memset((void*)&*impure, 0, sizeof(*impure));
Looks even more ugly than before :(
Agreed.
Also, I don't think we should make mass cleanups in stable (v3) and,
probably, in
10.07.2020 12:21, Karol Bieniaszewski wrote:
I know that you wait for some fidback about testing. And you wait for
some bugs to be reported especially about new features.
But i suppose it is also valuable to you to have info that old
functionality work ok and server do not crash.
Sure ;-)
24.06.2020 21:14, marius adrian popa wrote:
so now we have int128 with this commit
https://github.com/FirebirdSQL/firebird/commit/c2692c6ceec2a434a23be920afc2d430a85e27cb?
We already had it in Beta 2 as NUMERIC(38). Only the alias name was
added in this commit.
Dmitry
Firebird-Devel
18.06.2020 13:33, Dimitry Sibiryakov wrote:
18.06.2020 12:12, Karol Bieniaszewski wrote:
In my humble opinion this discussion is usless.
Yes. But someone has too much free time because of the pandemy and
this someone want to command others to do a lot of pointless work.
Given that our
17.06.2020 21:34, Adriano dos Santos Fernandes wrote:
So it's about this, correct?
-- Uses two index segments
select registro_sec_roest from dp_registro_oest where registro_pri_roest
= 1 and registro_sec_roest = 1;
-- Uses one index segment - more generalization of IN problem
select
17.06.2020 19:42, Adriano dos Santos Fernandes wrote:
Unfortunately, not exactly so. ConditionalStream is good only when
choice is between NATURAL and INDEX plans. It cannot be used when
choosing between good and bad index matches.
Currently, your PR shows a regression for CORE-5236, subquery
17.06.2020 14:33, Dimitry Sibiryakov wrote:
What is coding style rules about indents? Only tabs are allowed?
https://firebirdsql.org/en/coding-style/
#38. Tabs are allowed and must be used for indentation.
Dmitry
Firebird-Devel mailing list, web interface at
Components: Engine
Affects Versions: 4.0 Beta 2, 3.0.5, 2.5.9, 4.0 Beta 1, 3.0.4, 3.0.3,
2.5.8, 4.0 Alpha 1, 3.0.2, 2.5.7, 3.0.1, 2.5.6, 3.0.0, 4.0 Initial, 2.5.5,
2.5.4, 2.5.3 Update 1, 2.5.3, 2.5.2 Update 1, 2.5.2, 2.5.1
Reporter: Dmitry Yemanov
Priority: Minor
While UPDATE
03.06.2020 20:29, Adriano dos Santos Fernandes wrote:
I ran Mark's test for CORE-6322 (and variant with index) and TCS. Would
like to have the full test suite ran on it too.
In the first commit there is only the necessary changes (and would be
backported to v3 if approved). The second commit
16.06.2020 10:49, Alex Peshkoff via Firebird-devel wrote:
Is "constexpr" allowed for usage?
IMO it should.
I'm also playing with some containers I think we should allow:
shared_ptr (with adaptor for our pools), unique_ptr.
I plan to talk more on this in the future.
Do we have some
[
http://tracker.firebirdsql.org/browse/CORE-4985?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Dmitry Yemanov reopened CORE-4985:
--
Fix Version/s: (was: 4.0 Alpha 1)
Re-opened due to regression found, fix is rolled back
: Windows only
Reporter: Dmitry Yemanov
Firebird uses file-system ID of the open database file to uniquely name its
dependent shared memory files - lock table, event table, etc. File-system ID is
retrieved using GetFileInformationByHandle() API and composed from the three
fields
19.05.2020 13:22, Adriano dos Santos Fernandes wrote:
I have made small changes to jrt\cvt.cpp and common\cvt.cpp to correct
the behavior of varchars not enforcing size limits when there are
trailing spaces in the string and uses UTF-8 in FB 3.0 and 4.0.
Why do you say it's about UTF-8/MBCS
29.04.2020 18:50, Dmitry Yemanov wrote:
Client side cursor appears not too complex at the first glance, but
not sure do we want to have client or server side cursor.
Scrollable cursor is already materialized at the server side, so I see
no point doubling efforts on the client. It should
101 - 200 of 1223 matches
Mail list logo
