01.04.2011 11:51, Dimitry Sibiryakov wrote:
I don't understand why this job is deferred. If DDL is impossible, user must
be informed
about this fact ASAP.
In order to perform the validation, you have to lock the table from
concurrent modifications. If it's done ASAP and the lock is then
01.04.2011 12:24, Alex Peshkoff wrote:
Dmitry, where have you found a reference in standard tat string column
can be altered to wider one type?
What Is see in SQL2008 is:
Your copy is outdated :-) My version of 2010-10-14 says:
alter table statement ::=
ALTER TABLE table name alter table
01.04.2011 12:53, Dimitry Sibiryakov wrote:
May I suggest to add non-standard optional clause NOVALIDATE for users who
are sure
that validation is unnecessary and want to save time on commit?..
I'm against this. But feel free to keep arguing :-)
Dmitry
02.04.2011 12:53, Thomas Steinmaurer wrote:
in respect to the sub-release policy, I guess that forthcoming V2.5
sub-releases also won't include new features. But, would it be
*technical* possible to add the Window functions to the V2.5 branch or
is there too much needed from the V3 code base?
11.04.2011 3:53, Adriano dos Santos Fernandes wrote:
Our functions dealing with XSQLDA/XSQLVAR has a dialect parameter.
When dialect is 0, the structures are interpreted as SQLDA/SQLVAR. These
structs are a legacy version with less items.
They declarations are not present in ibase.h at
12.04.2011 2:09, Adriano dos Santos Fernandes wrote:
I suggest we fix namespace usages in an incremental manner. How?
[skip]
This way subsystems are children of Firebird namespace and don't require
Firebird:: everywhere. We also avoid using namespace Firebird in
implementation files.
It
13.04.2011 15:13, Alex Peshkoff wrote:
But, all things considered, I feel that the feature is too complicated
and has too many security implications to be worth it. If anything, I
might suggest that the alias config that's already present could be
augmented with a flag for each alias
17.04.2011 1:21, Leyne, Sean wrote:
Am I reading this correctly; any non-zero value would be considered true?
If so, then I do not agree with this logic. True values must be specific,
all other values must be false.
FWIW, this logic is used since v1.5 and nobody so far complained.
Dmitry
17.04.2011 22:03, Adriano dos Santos Fernandes wrote:
It's allowed to use XSQLVAR (or blr messages) between different types
like number-varchar (or vice-versa) but it's not allowed to use
different types when one of them is a blob.
DSQL moves data using MOVD_move, which do it like a client
19.04.2011 9:49, Kovalenko Dmitry wrote:
Hi
The system table RDB$FUNCTION_ARGUMENTS (in ODS12) still exists
without RDB$DESCRIPTION field.
It was surely missed on my part, to be added.
RDB$PROCEDURE_PARAMETERS contains this (and RDB$SYSTEM_FLAG also) field.
I'm not so sure about
05.05.2011 14:34, marius adrian popa wrote:
Feature request related to this thread
http://tech.groups.yahoo.com/group/firebird-support/message/113203
Maybe the api should show it somewhere like in the the engine_version
example http://firebirdsql.su/doku.php?id=rdb_get_context
SELECT
09.05.2011 14:06, Vlad Khorsun wrote:
All metadata objects moved into Attachment. Metadata syncronization is guarded
by attachment's mutex now. Database::SyncGuard and company are replaced by
corresponding Attachment::XXX classes.
To make AST's work we need to release attachment mutex
13.05.2011 9:43, Russell R wrote:
This post is related to the following 2 reported incidents:
http://tracker.firebirdsql.org/browse/CORE-3121
http://firebird.1100200.n4.nabble.com/Error-Restoring-v2-1-Database-to-v2-5-td2994314.html
The error is caused when you are running a 2.5 server, but
29.06.2011 19:12, Paul Reeves wrote:
I've been studying the output of the MON$IO_STATS table but something doesn't
make sense. I've built a test rig that
a) catches the mon$io_stats to disk at the attachment level on disconnect
and
b) modifies a million records randomly during a
05.07.2011 10:23, Paul Reeves wrote:
All tests are run from a freshly restored backup containing 100,000 rows in
the test table. Forced Writes are ON and hard drive disc caches are disabled.
Firebird 2.1.4 is the server.
FW and HDD settings don't affect the page I/O stats.
The attachment
05.07.2011 21:32, Paul Reeves wrote:
Are you running SuperServer, Classic or SuperClassic?
It is super server, although this should not make any difference.
Classic doesn't have any background I/O activity, so there will be a
difference.
Dmitry
All,
I've faced problems while adding subj into trunk.
For procedure parameters, the syntax is:
COMMENT ON PARAMETER proc name.param name IS text
However, procedures and functions have different namespaces, so there
may be a function and its input parameter with exactly the same names.
Thus
I'm now convincing myself to the third option:
Extend the syntax but preserve backward compatibility via default:
COMMENT ON [{PROCEDURE | FUNCTION}] PARAMETER ...
but if the object type specifier is omitted, then the engine tries to be
smart and determine whether it's a procedure or a
26.07.2011 16:00, Alessandro Fiorino wrote:
Can someone point me to where can I found in the source how a DECIMAL
field is encoded when saved on disk ?
I need it to manually recover a broken database.
It depends on its declared precision and database dialect.
Also, you're unlikely to detect
26.07.2011 16:30, Alessandro Fiorino wrote:
I've been at least a bit lucky: the record structure of the data I
want to recover is simple and I already found how to detect a record
and decode all the other fields (a varchar which always store a 17
chars string and two integers).
The
26.07.2011 16:46, Alessandro Fiorino wrote:
Already tried it says it can't recover anything. Also tried ibrecovery
and it only recovers 20% of the data, less that what I have from my
last backup.
The IBSurgeon team also offers manual database recovery and they're
experienced in that. Perhaps
01.08.2011 18:24, Elmar Haneke wrote:
On FB 2.5 invalid data in Timestamp fields does raise an error (even on
reading).
On FB 1.5 the invalid data was stored without any problems.
How can I avoid that error on reading data which has been stored by FB 1.5?
You should correct those values in
07.08.2011 18:24, Dimitry Sibiryakov wrote:
07.08.2011 16:04, Ajay Singh wrote:
I need some information about the undo/redo logs which firebird keeps to
support
concurrent transactions.
Firebird is not Oracle, it doesn't use logs to support concurrent
transactions.
RTFM:
18.08.2011 13:55, Alex Peshkoff wrote:
Hi, who knows what was a reason to comment out following constants in
consts_pub.h in trunk?
//#define isc_info_db_SQL_dialect 62
//#define isc_dpb_SQL_dialect 63
//#define isc_dpb_set_db_SQL_dialect65
Without
25.08.2011 10:36, Kjell Rilbe wrote:
Any rough estimate when the next RC or final release will be ready? I
assume it's dependent on feedback from this RC, but based on experience etc?
Usually the next RC or final release appears after a month of public
testing.
Dmitry
30.08.2011 0:44, Kjell Rilbe wrote:
Has this been improved since FB 2.1?
Nope.
Any future plans?
In FB3, perhaps.
Dmitry
--
Special Offer -- Download ArcSight Logger for FREE!
Finally, a world-class log management
10.10.2011 12:26, Vlad Khorsun wrote:
If such clause is inside trigger code then it not affects security checks and
*whole* trigger code is checked. SQL standard says something about security
checks for conditional triggers ?
I believe the rules are the same. In general, it's impossible to
10.10.2011 14:43, Dimitry Sibiryakov wrote:
Wrong reason, as Oracle doesn't support external triggers.
Oracle SQL Reference for version 10 has other opinion:
-- quote ---
CREATE TRIGGER salary_check
BEFORE INSERT OR UPDATE OF salary, job_id ON
25.10.2011 14:03, Adriano dos Santos Fernandes wrote:
2) FW=OFF, and use fsync on COMMIT - pages will not be reordered, and
when COMMIT happens they will be written to disk in order
I believe this is wrong assumption. Nobody guarantees that OS will be
flushing the dirty pages from the
25.10.2011 14:45, Paul Reeves wrote:
The documentation in firebird.conf (v2.5) indicates that this is disabled for
posix.
It's disabled by default (setting = -1). But it does work if reconfigured.
Dmitry
--
The
25.10.2011 14:39, Philippe Makowski wrote:
in fact with theses settings, FW=OFF is safer than before
safe enough to be the default ?
Even in the paranoid mode (MaxUnflushedWrites = 1) they still don't
guarantee the write order. So, if the crash happens while the
transaction is being
31.10.2011 22:14, Michal Kubecek wrote:
when building current trunk from SVN (revision 53601) on Linux (x86_64
OpenSuSE Factory but the same was observed on i586 SLES 11 SP1 as well),
the command
gfix -write async security.tmp
sometimes segfaults (not always, in about half of the
06.11.2011 3:01, Vlad Khorsun wrote:
Running CORE-3650 test-case in different isql embedded sessions with
SharedCache and SharedDatabase, I had (one time) a hang and a crash. The
crash happened after the first command. I ran it in session1, session2,
session1, etc (or some variant).
This is
07.11.2011 11:23, Thomas Steinmaurer wrote:
With all the discussion, is it worth to create a JIRA issue for this
feature request?
I don't mind :-)
Dmitry
--
RSA(R) Conference 2012
Save $700 by Nov 18
Register now
14.11.2011 23:15, Damyan Ivanov wrote:
# session 1
$ FIREBIRD=. FIREBIRD_LOCK=. isql-fb dbd-firebird-test.fdb
Database: dbd-firebird-test.fdb
SQL
# session 2, from another terminal, running in the same working
# directory
FIREBIRD=. FIREBIRD_LOCK=. isql-fb dbd-firebird-test.fdb
# hangs
15.11.2011 2:11, Thomas Steinmaurer wrote:
I guess the same applies to MON$IO_STATS.MON$PAGE_FETCHES? If so, isn't
comparing MON$PAGE_READS with MON$PAGE_FETCHES a bit misleading if one
wants to check to possibly increase the database page buffers?
I bet you're thinking about the cache hit
17.11.2011 11:26, Alex Peshkoff wrote:
And one more interesting line:
open(./fb_init, O_RDWR|O_CREAT|O_LARGEFILE, 0600) = 7
Why is _SHARED_ file opened in CWD? It should go to /tmp/firebird.
Because he explicitly sets FIREBIRD_LOCK to CWD before running ISQL :-)
Dmitry
17.11.2011 4:39, Frank Ingermann wrote:
a) [ ] file it as a bug - this _should_ work!
This would get my vote, as it works fine (and returns rows in the
expected order) in PGSQL.
Dmitry
--
All the data continuously
17.11.2011 12:59, Philippe Makowski wrote:
I really wonder if we can't get MaxUnflushedWrites or something like
that under Linux too
I would be curious to do same test with this parameter under Linux
It works for all platforms, you just need to uncomment and set it up in
firebird.conf.
17.11.2011 13:49, Thomas Steinmaurer wrote:
the previously reported bug in the Trace API is a bit nasty for
customers how are extensively using the Trace stuff.
I wonder if there is some kind of time frame for releasing 2.5.2
available, because for people coming from 2.1, they can't download
17.11.2011 14:14, Adriano dos Santos Fernandes wrote:
Or to flush group of independent pages at once, like I proposed.
Sure. I was speaking about the current code only.
Dmitry
--
All the data continuously generated
17.11.2011 15:37, Vlad Khorsun wrote:
And, note, most heavy case (flush on commit) is already optimized.
With FW=ON, it's optimized only for better write order (less HDD head
movements). But AFAIU we're talking here about batched writes instead of
singleton writes.
Dmitry
28.11.2011 23:51, Helen Borrie wrote:
Here's another one: the COMMENT syntax for loading varchar text into
RDB$DESCRIPTION.
Frank has shown a somewhat similar one: it's a more or less common
practice to empty/nullify the RDB$***_SOURCE columns.
So perhaps, unless some good alternative can
29.11.2011 12:35, Vlad Khorsun wrote:
While i agree with you about SUBSTRING, i think the much better solution
will be to allow to index COMPUTED BY columns. It allows to not compare
potentially complex expressions at optimize time and to not write exactly
same expressions in every query
30.11.2011 11:37, Dmitry Yemanov wrote:
So the reason is elsewhere.
Found it. SF has an ability to set the default download for the every
major platform (Win/Lin/Mac/etc). This option was left empty for the
server binaries but set for the .NET provider packages.
I have it fixed now
30.11.2011 12:08, Thomas Steinmaurer wrote:
Or more object type bound, e.g.:
ALTER TABLE ... CHANGE OWNER TO ...
ALTER VIEW ... CHANGE OWNER TO ...
+1, although I'd prefer SET instead of CHANGE. And I believe that CREATE
OR ALTER should miss this clause.
Dmitry
30.11.2011 12:24, Mark Rotteveel wrote:
Is it able to discern between 32 and 64 bit?
Unfortunately, not.
On Windows Vista 64 bit (using Firefox 32 bit though)
I get the suggestion to download the Firebird 32 bit package.
This is the lesser evil. You know that 32-bit builds can work in both
30.11.2011 12:44, Alex Peshkoff wrote:
Who knows which evil is lesser...
If one downloads 32-bit build for 64-bit OS, he can always stay with it
even not knowing that 64 builds exist.
But at least it will work fine for him, even being possibly a bit slower
than the x64 build. And those who
14.12.2011 20:02, Carlos H. Cantu wrote:
Vlad, 06/12/2010 16:24:51:
could you edit firebird.conf at both client and server hosts and set
TcpRemoteBufferSize = 32700
?
ok, much better (but still worse than MySQL):
[06/12/2010 16:27:47] Opening FB Query and fetching all data ()
[06/12/2010
14.12.2011 21:09, Dimitry Sibiryakov wrote:
Only for massive fetches. Performance of big number of small queries cannot
be improved
this way.
They won't perform worse either. And I was replying to the example where
MySQL was faster in fetching rows.
Dmitry
31.12.2011 17:57, Alexander Peshkov wrote:
This will make each version of the record (not the record- but EACH
version of it) 4 bytes longer.
Not strictly necessary. We could use a variable-length encoding for txn
ids longer than 32 bits and mark such records with a new flag. It would
add
01.01.2012 14:34, Kjell Rilbe wrote:
Are there unused flag bits available in the current record format, that
could be used for this purpose? Or how are such flag bits encoded?
We have 7 or 8 bits (out of 16) currently available.
Dmitry
01.01.2012 16:48, Jesus Garcia wrote:
Is not better use 64 bits integer?. In the near future computers speed, hdd
speeds, etc. will increase, and transaction id of 64 bits overhead will
affect less and less to performance.
Disk speed was always the issue, it doesn't increase that fast as
03.01.2012 16:16, Dimitry Sibiryakov wrote:
03.01.2012 13:11, Dmitry Yemanov wrote:
I believe that the internal UDR queries should work exactly like
EXECUTE STATEMENT WITH CALLER PRIVILEGES.
I would prefer fine-grained rights instead of bunch...
Could you elaborate?
Dmitry
03.01.2012 17:25, Alex Peshkoff wrote:
I believe that the internal UDR queries should work exactly like
EXECUTE STATEMENT WITH CALLER PRIVILEGES. Dunno however whether its
already the case or still in the pipeline.
Why only caller privileges? If we have external procedure, what's wrong
if
03.01.2012 17:40, Alex Peshkoff wrote:
In EXECUTE STATEMENT certainly yes. But I do not understand why should
external procedure behave like EXECUTE STATEMENT, not like any other
stored procedure.
We seem to have some misunderstanding :-)
Both UDRs and EXECUTE STATEMENTs deal with dynamic
03.01.2012 18:09, Alex Peshkoff wrote:
Definitely. I was talking about ability to grant privileges to the whole
external procedure as an object :-)
Me too :-) I was describing how it's going to work. Caller is an UDR, so
caller privileges are privileges granted to that UDR.
Dmitry
03.01.2012 18:06, Vlad Khorsun wrote:
Who will be the CALLER if UDR function is used in SELECT statement issued by
client ?
CALLER is UDR itself, so this question is meaningless for me, sorry.
Dmitry
--
Write once.
03.01.2012 18:53, Vlad Khorsun wrote:
If we speak about statements that issued by UDR - certainly.
I was speaking only about statements issued by UDR, as it was the
subject initiated by Dimitry Sibiryakov.
But if we speak about
UDR itself - what set of privileges should it have ? So far
03.01.2012 18:20, Alex Peshkoff wrote:
How deep in call stack we going to propagate CALLER privileges (UDR call
some SQL statement which have UDR calls embedded and so on) ?
May be unlimited?
No recursion is required at all. If UDP MY_PROC selects UDF MY_FUNC from
table T1 and MY_FUNC in
All,
Would there be any objections if we remove the support for protocol
versions less than 10?
It means that fbclient would not connect to pre-v6 versions of InterBase
and fbserver would not accept connections from pre-v6 versions of gds32.
All Firebird versions would still be supported, as
18.01.2012 15:20, Alex Peshkoff wrote:
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
31.01.2012 2:12, Adriano dos Santos Fernandes wrote:
There is two obscure SQL commands called READ BLOB and WRITE BLOB.
The idea behind these commands is to make blob read/write using the
cursor (fetch/insert) API. Why they exist is not something I know, maybe
it predates blob API functions?
03.02.2012 14:10, Thomas Steinmaurer wrote:
what's the meaning of 2 for MON$STATEMENTS.MON$STATE?
STALLED. I.e. an interval between client fetches from the open cursor.
Priorly this state was reported as ACTIVE because the query is alive and
running but the CPU time is not spent by the engine
09.02.2012 17:58, Thomas Steinmaurer wrote:
Splitting temp destinations into e.g. RAM disk and if this is exhausted
to regular disk is getting more and more popular, but due to CORE-2422,
So far nobody has proved that this approach is better than adjusting
TempCacheLimit.
this doesn't seem
09.02.2012 20:58, Leyne, Sean wrote:
What Thomas is looking to define is a setup where an RAM would be the first
device for the temp results (to enable much faster processing -- RAM disk has
100x IO of even an SSD!!), with any excess falling to an SSD or HDD. The
approach/setup is
13.02.2012 11:39, Alex Peshkoff wrote:
The simplest part of the task. Just make them offset_t - very logical
name for offsets :-)
It depends on what headers you include. For example, our own offset_t
defined in File.h is always 64-bit :-)
Certainly, a lot of places where lock size is
13.02.2012 11:51, Nikolay Samofatov wrote:
Taking into an account that most of users do not need2Gb of lock
table, 64-bit offsets (at least for 2.5) should better remain tunable
build parameter, turned off by default.
Great idea, BTW. I haven't thought of a build parameter.
I can agree with
13.02.2012 14:13, Vlad Khorsun wrote:
Imagine x64 CS server and x32 embedded apps at the same time.
They must share the same lock table...
It was never possible. 64-bit lock table is incompatible with 32-bit one
and it has different version number in the header.
Dmitry
13.02.2012 13:45, Alex Peshkoff wrote:
I'm not really sure this is strictly required for v2.x. We surely must
add an overflow protection, here I agree.
A check in shared memory grow code should do a trick.
Sure.
But if it would be possible to
have ULONG offsets instead of SLONG, it would
11.02.2012 20:49, Mark Rotteveel wrote:
* Is the size being allocated for SQL_TEXT right? It is allocating
sqllen + 1 and null-terminating it, while looking at the Interbase 6
ApiGuide they are always allocating sqllen and not null-terminating
You don't have to null-terminate it, sqllen is
18.02.2012 16:54, Björn Reimer wrote:
Current memory = -507.368.239
Max memory = -466.997.327
Memory buffers = 20.480
Fetches from cache = -1.256.613.653
Historically, all performance counters were 32-bit. Later, they were
enhanced to represent 64-bit values. Firebird 2.5 could report
18.02.2012 23:33, Jiri Cincura wrote:
can I somehow ask for isc_info_firebird_version without database
handle? I'd like to know it before processing op_attach.
You can use the service manager and ask it for isc_info_svc_server_version.
Dmitry
15.02.2012 20:01, Mark Rotteveel wrote:
Does SQL_VARYING require the sqllen to be updated as well, or can that
be left as it was originally set by isc_dsql_describe_bind?
I'd say it should correspond to the real data. You're not required to
use isc_dsql_describe_bind(), after all.
Dmitry
22.02.2012 6:27, Claudio Valderrama C. wrote:
I think that playing here with streamList vs this-streamList is a recipe
for misreading the code and making mistakes:
This is why I prefer to prefix all member variables with m_, e.g.
m_streamList. Others may surely disagree with the chosen syntax
24.02.2012 11:26, Martijn Tonies wrote:
The message is I get using Windows authentication is:
I/O error during CreateFile (open) operation for file
E:\TEMP\FB3TEST.FDBError while trying to open fileThe process cannot
access the file because it is being used by another process.
I'm using
06.03.2012 10:46, Alex Peshkoff wrote:
And taking into an account that we do not support ODS 12 in current
engine - do we need to keep support of old BLR versions? I see 2 ways
for old BLR to arrive in ODS12 database. First of all - when old
database is restored. May be we should learn gbak
06.03.2012 15:05, Alex Peshkoff wrote:
Also want to agree, but I remember there were some real reasons
(unfortunately do not remember them) making people stay with dialect1.
May be somebody remembers that reasons?
Lack of precisions bigger than 18 in NUMERIC/DECIMAL and, even more
important,
The Firebird Project team is happy to announce that the v2.0.7 release
candidate kits for Linux, Windows and Mac OS X platforms are ready for
testing.
The download page:
http://www.firebirdsql.org/en/firebird-2-0-7-rc1/
Enjoy the testing and please don't hesitate to report the found
Claudio,
People, what's the difference between
const int csb_modify = 64; // we are processing a
modify
combined with
const int csb_erase = 256; // we are processing
an erase
and the single flag
const int csb_update = 1024;
Claudio,
Hello, can someone comment on the usefulness of this flag, please?
class dsql_req : public pool_allocdsql_type_req
{
static const unsigned FLAG_EMBEDDED = 0x02;
It's only used in dsql.cpp, inside DSQL_execute()
if ((SSHORT) in_msg_type ==
17.03.2012 19:44, marius adrian popa wrote:
I have announced here for the comments, Maybe Adriano or others can
share their comments about the core improvements :)
It was mostly an info for end users (those who's interested). Once it
gets closer to the implementation, it will be discussed
18.03.2012 11:59, Kjell Rilbe wrote:
I'd like to subscribe to your blog, being a FB user (app/system devel).
I'm a bit dumb as to blogs/rss, but I see no subscription link on your
blog site. How can I get an RSS feed from your blog?
The links has been added, sorry for inconvenience.
Atom
20.03.2012 12:12, Thomas Steinmaurer wrote:
Is this a known issue?
http://firebird.1100200.n4.nabble.com/Fb-3-cannot-get-additional-connections-td4413326.html
Dmitry
--
This SF email is sponsosred by:
Try Windows
20.03.2012 13:22, Dmitry Yemanov wrote:
Not for me (in both modes).
It seems I see it now, sorry.
Will have a look and report back.
Dmitry
--
This SF email is sponsosred by:
Try Windows Azure free for 90 days Click
22.03.2012 12:29, Thomas Steinmaurer wrote:
= Does that mean: We have a single executable in Firebird 3. When -m is
used, this means a single server process and the SharedCache,
SharedDatabase parameters controll whether it's SS or SC. When -m is not
used, then it's one process per
22.03.2012 12:34, Alex Peshkoff wrote:
Probably small change in help text - SS or SC - will be useful.
Before doing that, we need to decide whether the former SuperClassic
will be the documented option or just some internal stuff for easier
debugging etc, as FB3 is intended to make it mostly
22.03.2012 13:00, Thomas Steinmaurer wrote:
* Under the assumption that SS as a single-process model with it's
shared cache supports SMP (I think this was a primarly goal) in 3.0 like
SC in 2.5, I don't see a reason to offer SC in 3.0 at all (at least the
SC terminology; which already shows
23.03.2012 11:13, Alex Peshkoff wrote:
In trunk - definitely. What about 2.5 - not sure. May be having patch
for high-end systems is better approach than commit in svn?
I'd say we need to see the patch before deciding.
Dmitry
28.03.2012 17:23, Jiri Cincura wrote:
The 2.7.5 version of ADO.NET provider for Firebird is ready for download.
http://blog.cincura.net/232764-ado-net-provider-for-firebird-2-7-5-released/
Your blog refers to the Firebird website for download, but the website
does not contain the links for
03.04.2012 11:55, Alex Peshkoff wrote:
To work with encrypted database file we need a tool to encrypt database.
I see 3 possible solutions for it. In all 3 cases some plugin dependent
parameter may be passed to plugin. In all cases one may use decrypt
instead encrypt to make
1. ALTER
Claudio Valderrama C. wrote:
Would it call the INF functions to get that information? I ask because
obviously gfix cannot work like gstat in local mode, IE we don't expect it
to be able to read the page header directly.
Yes, isc_database_info() could be a solution here (if we really need a
07.04.2012 19:45, Adriano dos Santos Fernandes wrote:
It really don't expose all the engine internals.
You can get it running SQL on this method.
Perhaps it would be a better idea to write a SQL trigger calling the
external procedure and passing the required context through.
Dmitry
09.04.2012 14:19, Mark Rotteveel wrote:
Are events available on embedded, and if so are there any special things
you need to take care of to get them working?
Of course they're available and working out of the box, without any
trickery.
Dmitry
09.04.2012 19:26, Mark Rotteveel wrote:
I notice in /jrd/why.cpp (2.5 branch) that a similar named array does
not have the isc_info_sql_relation_alias entry:
static const SCHAR sql_prepare_info2[] =
{
isc_info_sql_stmt_type,
// describe_select_info
isc_info_sql_select,
09.04.2012 20:22, Dimitry Sibiryakov wrote:
Isn't aliasname field of XSQLVAR for the relation alias name?..
It's the field alias, IIRC.
Dmitry
--
For Developers, A Lot Can Happen In A Second.
Boundary is the first to
11.04.2012 6:14, Claudio Valderrama C. wrote:
Example of an API call?
isc_compile_request, isc_transact_request, isc_dsql_*_m family.
I assume that if someone wants to talk BLR v6, he should indicate that.
Sure, any BLR stream starts with its version, be it on disk or at the
API level. I
12.04.2012 14:50, Alex Peshkoff wrote:
I do not like use of sed script for that.
Me neither. And IIRC, it's not required for Windows builds now, unless
you build the installer.
If we anyway maintain own copy of btyacc - why not change it
Agreed.
and suggest patch to them?
It's pointless,
11.04.2012 17:41, Adriano dos Santos Fernandes wrote:
In the past, I thought about this problem, and a possibility I saw was
to create a BLR prefix, saying the next BLR-rse verb would use two bytes
for stream numbers instead of one.
I do not know if this is a good solution, it's just a
13.04.2012 18:34, marius adrian popa wrote:
In Firebird 3.x you can write your own protocol, this is what i
understand from the pluggable architecture
In your wishes, perhaps. While it might be doable, we are not going to
replace everything with plugins.
Dmitry
1 - 100 of 1223 matches
Mail list logo
