EXECUTE BLOCK parameters definitions are not respected and may cause wrong
behavior related to character sets
-
Key: CORE-3737
URL: http://tracker.firebird
> I have a table:
> recreate table test (
> base numeric(15,4),
> test1 computed by (base+base),
> test2 computed by (cast(base+base as numeric(15,4))),
> foo bigint,
> bar numeric(15,4)
> );
>
> If I run the query:
> select rf.RDB$FIELD_NAME, f.RDB$FIELD_SCALE, f.RDB$FIELD_TYPE,
> f.RDB$FIEL
18.01.2012 13:18, Alex Peshkoff wrote:
> But I do not like an idea of making SQL
> command (GRANT CREATE DATABASE TO SomeUser) dependent upon
> authentication plugins behavior.
But you don't mind SQL command CREATE USER to be dependent on plugin, do
you?..
--
SY, SD.
On 01/18/12 16:06, Dimitry Sibiryakov wrote:
> 18.01.2012 12:20, Alex Peshkoff wrote:
>> This table is checked every time when
>> createDatabase() is called - even if authentication is not related to
>> security database. Other suggestions?
>Why security plugin cannot decide this? If plugin us
On Wed, 18 Jan 2012 13:06:18 +0100, Dimitry Sibiryakov
wrote:
> 18.01.2012 12:20, Alex Peshkoff wrote:
>> This table is checked every time when
>> createDatabase() is called - even if authentication is not related to
>> security database. Other suggestions?
>
>Why security plugin cannot decid
18.01.2012 12:20, Alex Peshkoff wrote:
> This table is checked every time when
> createDatabase() is called - even if authentication is not related to
> security database. Other suggestions?
Why security plugin cannot decide this? If plugin uses security database -
it can check
the table. Oth
We have plans to make it possible to grant and revoke rights to create
new database objects (tables, procedures, generators, etc.). The
implementation is more or less clear - it may be new system table or new
type of ACL. But there is one DDL which does stand separately - CREATE
DATABASE. We do no
user without any rights can delete from RDB$DATABASE, RDB$COLLATIONS,
RDB$CHARACTER_SETS
Key: CORE-3735
URL: http://tracker.firebirdsql.org/browse/CORE-3735
Proj
user with read_only right on some table can block others from updating this
table
-
Key: CORE-3736
URL: http://tracker.firebirdsql.org/browse/CORE-3736
Project: Firebird
On Wed, 18 Jan 2012 09:55:45 +0400, Alex Peshkoff
wrote:
> But getting serious - certainly it's possible to set affinity on linux,
> but there is no sense in doing that. Windows make thread 'jump' from one
> core to another too often, therefore cache is used not efficiently. As
> far as I remember
On Wed, 18 Jan 2012 08:58:09 +0100, Jiri Cincura
wrote:
> Hi *,
>
> I have a table:
> recreate table test (
> base numeric(15,4),
> test1 computed by (base+base),
> test2 computed by (cast(base+base as numeric(15,4))),
> foo bigint,
> bar numeric(15,4)
> );
>
> If I run the query:
> se
On 01/18/12 12:37, Philippe Makowski wrote:
> thanks
>
> For the fedora package, I will just add a patch
>
The problem that this patch will be needed not for fedora only (you see
yourself, this is not gcc specific version issue), but for every build
of 3.3.0 with current ibase.h. May be something
Alex Peshkoff [2012-01-18 09:24] :
> On 01/18/12 12:07, Philippe Makowski wrote:
>> Alex Peshkoff [2012-01-18 09:01] :
>>> What branch are you compiling? In trunk I see lines with errors
>>> commented. Do not think that gcc 4.7 stopped to understand comments :)
>>> BTW, it's time to remove that
On 01/18/12 12:07, Philippe Makowski wrote:
> Alex Peshkoff [2012-01-18 09:01] :
>> What branch are you compiling? In trunk I see lines with errors
>> commented. Do not think that gcc 4.7 stopped to understand comments :)
>> BTW, it's time to remove that lines at all.
>>
> the official tar.bz2 ki
Hi *,
I have a table:
recreate table test (
base numeric(15,4),
test1 computed by (base+base),
test2 computed by (cast(base+base as numeric(15,4))),
foo bigint,
bar numeric(15,4)
);
If I run the query:
select rf.RDB$FIELD_NAME, f.RDB$FIELD_SCALE, f.RDB$FIELD_TYPE,
f.RDB$FIELD_SUB_TYPE
f
15 matches
Mail list logo
