Re: [Firebird-devel] Code readability

On 10/13/16 13:50, Dimitry Sibiryakov wrote: > Hello, All. > > Is here a person, who can read and understand this line: > >Arg::Num(index) << (string("IMessageMetadata::") + method)).copyTo(status); > > but cannot understand this line: > >Arg::Num(index) <<

Re: [Firebird-devel] FB 3.0.1 - Different concatenation result to FB 2.5.6

On 10/10/16 18:21, Dimitry Sibiryakov wrote: 10.10.2016 17:04, Alex Peshkoff wrote: To be precise the result in FB3 is: T2 == T1 _SUFIX T1234567890_SUFIX where _SUFIX starts exactly at position == 31 (see

Re: [Firebird-devel] FB 3.0.1 - Different concatenation result to FB 2.5.6

On 10/10/16 13:10, Slavomir Skopalik wrote: > Hi, > > I found different behavior of concatenation between FB2.5.6 and FB 3.0.1. > > FB 2.5.6: > > EXECUTE BLOCK RETURNS(T2 VARCHAR(70)) AS > DECLARE T1 RDB$RELATION_NAME = 'T1'; > BEGIN > T2 = T1 || '_SUFIX'; > SUSPEND; > T1 =

Re: [Firebird-devel] Regression service API for isc_spb_sts_table?

On 03.10.2016 21:19, Mark Rotteveel wrote: > On 3-10-2016 11:42, Alex Peshkoff wrote: >> On 01.10.2016 16:47, Mark Rotteveel wrote: >>> After upgrading to Firebird 3.0.1 a test that uses the services API to >>> obtain table statistics nows fails with an exception: &

Re: [Firebird-devel] Regression service API for isc_spb_sts_table?

On 01.10.2016 16:47, Mark Rotteveel wrote: > After upgrading to Firebird 3.0.1 a test that uses the services API to > obtain table statistics nows fails with an exception: > > unrecognized service parameter block; All services except for getting > server log require switches [SQLState:42000, ISC

Re: [Firebird-devel] Firebird 3.0.1 still don't build under Fedora 24 with gcc6

On 30.09.2016 12:04, Philippe Makowski wrote: > Hi, > > I tried to build latest stable 3.0.1 under Fedora 24 without success. > Fedora 24 use gcc 6.2.1 > > > log is here : > https://copr-be.cloud.fedoraproject.org/results/makowski/firebird/fedora-24-x86_64/00459581-firebird/build.log.gz > > I

Re: [Firebird-devel] Allowed C++11 feature

On 23.09.2016 02:27, Adriano dos Santos Fernandes wrote: > I would like to add the following well-supported C++11 feature to our > allowed list: > > Strongly-typed enum - > http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2007/n2347.pdf > > Comments? +1

Re: [Firebird-devel] Necessity of prepare in FB3

On 22.09.2016 17:28, Dmitry Yemanov wrote: > 22.09.2016 17:09, Alex Peshkoff wrote: >> Just one note - let's finish with batch API interface first. Looks like >> required functionality may be present in that interfaces in much more >> logical way. > I'd say these features

Re: [Firebird-devel] Necessity of prepare in FB3

On 22.09.2016 16:42, Dmitry Yemanov wrote: > 20.09.2016 17:02, Alex Peshkoff wrote: >> I just wantedto say that modifying behaviour of openCursor() >> and letting IResultSetfetch from non-cursor object is hardly >> correct solution. > While this approach is surely not eleg

Re: [Firebird-devel] Is IXpbBuilder broken?

On 22.09.2016 13:50, Gabor Boros wrote: > Hi All, > > Wrote in august to the support list but no answer: > > https://groups.yahoo.com/neo/groups/firebird-support/conversations/topics/129585 > > > > So, I want to try the new API (looks more easier to use than the old) > but got an exception with

Re: [Firebird-devel] Necessity of prepare in FB3

On 20.09.2016 18:35, Dimitry Sibiryakov wrote: > 20.09.2016 17:38, Martin Schreiber wrote: >> Additionally I prefer robust and simple solutions without guessing. > Then you have no choice but call prepare. > > With current API - yes, but we discuss possible API change.

Re: [Firebird-devel] Necessity of prepare in FB3

On 20.09.2016 18:32, Martin Schreiber wrote: > On Tuesday 20 September 2016 16:02:26 Alex Peshkoff wrote: >> On 20.09.2016 12:25, Martin Schreiber wrote: >>> On Tuesday 20 September 2016 10:21:54 Alex Peshkoff wrote: >>>>> A convenient solution of the dilem

Re: [Firebird-devel] Necessity of prepare in FB3

On 20.09.2016 12:25, Martin Schreiber wrote: > On Tuesday 20 September 2016 10:21:54 Alex Peshkoff wrote: >>> A convenient solution of the dilemma would be if openCursor() would work >>> with any SQL statement. >>> It seems that a part of that solution alr

Re: [Firebird-devel] CNCT_user and CNCT_host

On 20.09.2016 14:47, Jiří Činčura wrote: >> Pay attention - _removes_. Does not matter was there something useful in >> CNCT or not. > OK, I missed that. I though it's like a merge of values. Unfortunately not. Certainly that can be fixed for fb4 and later but clients should better work with

Re: [Firebird-devel] Necessity of prepare in FB3

On 18.09.2016 10:24, Martin Schreiber wrote: > Hi, > Please bear with me if I am completely wrong, I am a simple toolmaker and no > database specialist. > > MSEide+MSEgui includes a fork of the Free Pascal SQLdb framework. There users > provide arbitrary SQL-statements and MSEgui provides the

Re: [Firebird-devel] CNCT_user and CNCT_host

On 19.09.2016 21:05, Jiří Činčura wrote: >> things. But for os user name - no, not ok. > Can you elaborate? > I've already did it before ;) > Existing server code removes them from DPB, takes values loaded > earlier from connect packet and puts into DPB. Pay attention - _removes_. Does not

Re: [Firebird-devel] CNCT_user and CNCT_host

On 19.09.2016 20:14, Jiří Činčura wrote: >> Historical reasons. >> This values were passed in CNCT since pre-firebird times therefore >> current solution was OK for old clients. > But there's inherently nothing wrong with using DPB, right? Because i.e. > isc_dpb_process_id and

Re: [Firebird-devel] CNCT_user and CNCT_host

On 16.09.2016 23:22, Jiří Činčura wrote: > I was thinking quite the opposite. Dropping the CNCT values. In DPB > there's more (like process ID etc.) so I thought I would logically > group it there. Historical reasons. This values were passed in CNCT since pre-firebird times therefore current

Re: [Firebird-devel] CNCT_user and CNCT_host

On 16.09.2016 22:48, Jiří Činčura wrote: > Thanks. Not sure why I'm sending it in CNCT and DPB as too. Maybe some > history code. > > Time to clean it up! :) And save few bytes. > Better drop them from DPB. Existing server code removes them from DPB, takes values loaded earlier from connect

Re: [Firebird-devel] CNCT_user_verification

On 16.09.2016 19:01, Jiří Činčura wrote: > Hi, > > what exactly this flag controls? I can't see much explanation in > protocol.h. > It's unused today but sooner of all needed for protocol 10 (sorry, no desire to review interbase 6 sources).

Re: [Firebird-devel] Fwd: [Firebird-checkins] [FirebirdSQL/firebird] 1d2990: Implemented CORE-5204: create an option in the ins...

On 15.09.2016 16:59, Adriano dos Santos Fernandes wrote: > On 15/09/2016 10:56, Alex Peshkoff wrote: >> On 15.09.2016 16:41, Adriano dos Santos Fernandes wrote: >>> On 15/09/2016 09:56, Alex Peshkoff wrote: >>>> Option from subj is called -path, i.e. now one can: >

Re: [Firebird-devel] Fwd: [Firebird-checkins] [FirebirdSQL/firebird] 1d2990: Implemented CORE-5204: create an option in the ins...

On 15.09.2016 16:41, Adriano dos Santos Fernandes wrote: > On 15/09/2016 09:56, Alex Peshkoff wrote: >> Option from subj is called -path, i.e. now one can: >> >> ./install.sh -path /opt/fb.4 >> >> and firebird will be installed in fb.4 instead firebird. >>

Re: [Firebird-devel] Fwd: [Firebird-checkins] [FirebirdSQL/firebird] 1d2990: Implemented CORE-5204: create an option in the ins...

On 15.09.2016 16:02, Dimitry Sibiryakov wrote: >> Before porting it to FB3 > > Ist'n release branch supposed for bugfixes only?.. For porting efforts too. Ability to install to arbitrary path exist in windows for along time, now it's added to linux port. But I'm open to others minds - if people

[Firebird-devel] Fwd: [Firebird-checkins] [FirebirdSQL/firebird] 1d2990: Implemented CORE-5204: create an option in the ins...

Option from subj is called -path, i.e. now one can: ./install.sh -path /opt/fb.4 and firebird will be installed in fb.4 instead firebird. Before porting it to FB3 I ask everyone interested to try on favourite linux distro cause changes are highly distro-specific, please pay special attention

Re: [Firebird-devel] Foreign key error even there should not be one (maybe, Weird problem in one database)

On 15.09.2016 10:14, Tommi Prami wrote: > Whole story here : > > https://plus.google.com/105626236962310493174/posts/3UZB71NjUhA > > Copy of message > - - - - - - - - - - - - - - - - > Is there a tool (or way) to debug database step by step. > > Now we have weird situation where delete from

Re: [Firebird-devel] Minimal buffer size for IUtil.formatStatus() (FB3)

On 09/05/2016 12:54 PM, Martin Schreiber wrote: > Hi, > It seems that IUtil.formatStatus() needs a minimal buffer size of 50 because > of src/yvalve/gds.cpp:749 > " > static SLONG safe_interpret(char* const s, const FB_SIZE_T bufsize, > const ISC_STATUS** const vector, bool legacy) > { >

Re: [Firebird-devel] Compiling Firebird.pas on Free Pascal

On 09/04/2016 06:26 PM, Tony Whyman wrote: > {$ifdef fpc} {$mode delphi} {$endif} Yes -Mdelphi is required switch. And where should be that line added? In the most beginning of firebird.pas? -- Firebird-Devel mailing

Re: [Firebird-devel] isc_dpb_trusted_auth in FB3

On 09/01/2016 05:40 PM, Jiří Činčura wrote: > Nice, nailed it. > > Another question. For Srp I send the auth data as a HEX string in UTF8 > bytes (and that works), while in Sspi the auth data is sent (at least > only that works) as bytes directly. In both cases in > isc_dpb_specific_auth_data. > >

Re: [Firebird-devel] isc_dpb_trusted_auth in FB3

On 09/01/2016 04:54 PM, Jiří Činčura wrote: > So where the client auth data should be sent? And which tag? > client auth data may be sent: 1. In op_connect, in user_id, tag CNCT_specific_data, together with CNCT_login, CNCT_plugin_name and CNCT_plugin_list (used first time). 2. In DPB, tag

Re: [Firebird-devel] isc_dpb_trusted_auth in FB3

On 09/01/2016 01:39 PM, Jiří Činčura wrote: > Looks like I moved somewhere. In op_connect+op_accept I send as plugin > Win_Sspi and I send the client token. Then I get back response as > op_accept_data, there I find the auth data from server and I can use > there to create client authentication

Re: [Firebird-devel] Response to op_get_segment under with compression on

On 08/22/2016 11:53 AM, Jiří Činčura wrote: with some debugging code turned on. Can I turn that on? Please :) diff --git a/src/remote/protocol.cpp b/src/remote/protocol.cpp index 042d879..7f0648b 100644 --- a/src/remote/protocol.cpp +++ b/src/remote/protocol.cpp @@ -285,6 +285,10 @@ bool_t

Re: [Firebird-devel] Response to op_get_segment under with compression on

On 08/22/2016 11:53 AM, Jiří Činčura wrote: >> with some debugging code turned on. > Can I turn that on? > >> So must say I see no problems with protocol. > Oh no, sure. I'm not saying something is wrong with the protocol. It > works fine when I turn compression off or if the data is all 0 (see my

Re: [Firebird-devel] Response to op_get_segment under with compression on

On 08/22/2016 11:44 AM, Jiří Činčura wrote: > OK, I was able to strip it down to single select on a fresh empty > database. > > If I do "select >

Re: [Firebird-devel] Stopping firebird.exe

On 08/21/2016 09:43 AM, Jiří Činčura wrote: > Hi *, > > if I start firebird using "firebird.exe -a", is there a way to stop it? > In a scripted way. > I know programmatical way to stop it. It's ugly but should work. Query services manager with undocumented(!!!, it's in svc_undoc.h)

Re: [Firebird-devel] Response to op_get_segment under with compression on

On 08/21/2016 12:51 PM, Jiří Činčura wrote: > Hi *, > > when I try to read response to op_get_segment (which works fine without > compression), after flush I only get op_response (9) code and then > nothing is in the stream. The next piece I'm trying to read is 4 > bytes/integer for object handle

Re: [Firebird-devel] [SPAM] Re: RFC: Timeouts

On 08/18/2016 06:26 PM, Vlad Khorsun wrote: > 18.08.2016 18:08, Dimitry Sibiryakov wrote: >> 18.08.2016 16:55, Vlad Khorsun wrote: >>> Global timeout is a last line of defense for DBA against bad apps, >>> wrong queries, >>> developer mistakes, unlucky days (dropped some indices last week but

Re: [Firebird-devel] POSIX os_utils::open() and O_CLOEXEC

On 08/11/2016 03:13 PM, Dimitry Sibiryakov wrote: > Hello, All. > > If posix/os_utils.cpp I see following code: > >> // force file descriptor to have O_CLOEXEC set >> int open(const char* pathname, int flags, mode_t mode) >> { >> int fd; >> fd = openFile(pathname, flags |

Re: [Firebird-devel] isc_dpb_auth_block format

On 08/05/2016 07:40 PM, Dimitry Sibiryakov wrote: > 05.08.2016 18:18, Alex Peshkoff wrote: >> On 08/05/2016 07:01 PM, Dimitry Sibiryakov wrote: >>> 05.08.2016 17:48, Alex Peshkoff wrote: >>>> That's priority byte. If 2 different clumplets of data from auth

Re: [Firebird-devel] isc_dpb_auth_block format

On 08/05/2016 07:13 PM, Dimitry Sibiryakov wrote: > 05.08.2016 17:48, Alex Peshkoff wrote: >> In the end of ClumpletReader.h you can find AuthReader class that helps >> to parse auth block. > I see the class but it gives me no idea how to interpret data from auth >

Re: [Firebird-devel] isc_dpb_auth_block format

On 08/05/2016 07:01 PM, Dimitry Sibiryakov wrote: > 05.08.2016 17:48, Alex Peshkoff wrote: >> That's priority byte. If 2 different clumplets of data from auth block >> both map into same authentication type (role for example) but with >> different values wins one

Re: [Firebird-devel] isc_dpb_auth_block format

On 08/05/2016 05:18 PM, Dimitry Sibiryakov wrote: > Hello, All. > > I've dumped dpb and got following sequence: > > 79 - isc_dpb_auth_block, ok > 90, 0, 0, 0 - its length is 90, ok, next 90 bytes is supposedly auth > block itself in > wide untagged format > 0 - it must be auth

Re: [Firebird-devel] [FB-Tracker] Resolved: (CORE-5323) Don't delete isc_dpb_user_name clumplet from remote connection dpb

On 08/05/2016 04:28 PM, Dimitry Sibiryakov wrote: > 05.08.2016 15:26, Alexander Peshkov (JIRA) wrote: >> That's not improvement but step back compared with use of authentication >> block in provider. > In this case inconsistent transfer of auth parameters into a provider is > a bug. Should >

Re: [Firebird-devel] Firebird 3 crash on detach

On 08/01/2016 05:45 PM, Dimitry Sibiryakov wrote: > 01.08.2016 16:39, Alex Peshkoff wrote: >> Dmitry, if it crashes better provide backtrace. > Here it is: > >>> fbclient.dll!Why::YEntry::~YEntry() >>> Line 1188 C++ >> fbclien

Re: [Firebird-devel] Firebird 3 crash on detach

On 08/01/2016 05:08 PM, Dimitry Sibiryakov wrote: > Hello, All. > > In YAttachment::detach() I see following code: > >> YEntry entry(status, this, 0); >> >> if (entry.next()) >> entry.next()->detach(status); >> >> if

Re: [Firebird-devel] IStatement and externally provided message metadata

On 08/01/2016 12:34 PM, Dimitry Sibiryakov wrote: > 01.08.2016 11:26, Alex Peshkoff wrote: >> IResutSet:getMetadata() should always return actual format that will be >> used when fetching records. > Wow. It differs from IStatement behavior. > > Certainly. May be b

Re: [Firebird-devel] IStatement::free() and result sets

On 07/30/2016 02:51 PM, Dimitry Sibiryakov wrote: > Hello, All. > > Should IResultSet returned by IStatement::openCursor remain in working > state after > free() call? > I.e. is following code allowed? > > IResultSet* rs = stmt->openCursor(...); > stmt->free(); <<

Re: [Firebird-devel] IStatement and externally provided message metadata

On 07/29/2016 09:02 PM, Dimitry Sibiryakov wrote: > 29.07.2016 19:33, Alex Peshkoff wrote: >> It should return default generated automatically metadata - the fact >> that statement was executed with some non-default metadata does not >> affect default parameters format. >

Re: [Firebird-devel] IStatement and externally provided message metadata

On 07/29/2016 06:21 PM, Dimitry Sibiryakov wrote: > Hello, All. > > Consider following code: > > IMessageMetadata* inm = XXX; > IStatement* stmt = att->prepare(); > stmt->execute(, inm, ); > > Should following stmt->getInputMetadata(...) return the same inm

Re: [Firebird-devel] StatementMetadata::getType()

On 07/29/2016 04:27 PM, Dimitry Sibiryakov wrote: > 29.07.2016 15:20, Alex Peshkoff wrote: >> StatementMetadata::getType() returns COLUMN type, but >> Statement::getType() - STATEMENT type. They are not related. > JStatement::getType() seems to have different opinion

Re: [Firebird-devel] StatementMetadata::getType()

On 07/29/2016 03:34 PM, Dimitry Sibiryakov wrote: > 29.07.2016 14:26, Dimitry Sibiryakov wrote: >> StatementMetadata::getType() doesn't perform backward mapping for constants >> received from it. > Perhaps, documentation is wrong and IStatement::getType() is supposed to > return >

Re: [Firebird-devel] IMetadataBuilder and field names

On 07/28/2016 06:30 PM, Dimitry Sibiryakov wrote: > 28.07.2016 17:24, Alex Peshkoff wrote: >> You must have your own implementation, not one created by >> IMetadataBuilder. It's trivial. > Implementation of IMessageMetadata is trivial. Assuring compatibility of

Re: [Firebird-devel] IStatement::getOutput/InputMetadata()

On 07/28/2016 05:55 PM, Dimitry Sibiryakov wrote: > Hello, All. > > If statement has no input or output parameters, what must return > corresponding methods? > NULL or valid IMessageMetadata with zero variables inside? > Valid IMessageMetadata with zero variables inside seems to be

Re: [Firebird-devel] IMetadataBuilder and field names

On 07/28/2016 06:17 PM, Dimitry Sibiryakov wrote: > 28.07.2016 17:13, Alex Peshkoff wrote: >>> How field and relation name is supposed to be set in IMessageMetadata >>> built with >>>> IMetadataBuilder? >>>> >> Obviously empty. > H

Re: [Firebird-devel] IMetadataBuilder and field names

On 07/28/2016 06:08 PM, Dimitry Sibiryakov wrote: > Hello, All. > > How field and relation name is supposed to be set in IMessageMetadata > built with > IMetadataBuilder? > Obviously empty. -- Firebird-Devel

Re: [Firebird-devel] IAttachment::openCursor with single result

On 07/27/2016 01:40 PM, Dimitry Sibiryakov wrote: > 27.07.2016 12:26, Alex Peshkoff wrote: >> Raise error. > What gds code? > > Currently it's isc_random, in fb4 to be fixed. -- What NetFlow Analy

Re: [Firebird-devel] IAttachment::openCursor with single result

On 07/26/2016 03:56 PM, Dimitry Sibiryakov wrote: > Hello, All. > > What must to do openCursor() if it is called with statement that do not > return cursor, > such as execute procedure or update returning? > Raise error.

Re: [Firebird-devel] IAttachment::openCursor without outMetadata

On 07/25/2016 05:29 PM, Dimitry Sibiryakov wrote: > Hello, All. > > In new API guide it is written that "Format of output data is defined by > outMetadata > parameter, leaving it NULL default format may be used". But in code I > everywhere see > comparison with DELAYED_OUT_FORMAT, which

Re: [Firebird-devel] Providers and interfaces in parameters

On 07/22/2016 04:18 PM, Dimitry Sibiryakov wrote: > Hello, All. > > When a provider receive, for example, ITransaction* as a parameter for a > method, can it > be sure that this is the object it returned from startTransaction() before? > For example, can Y-valve get JTransaction from

Re: [Firebird-devel] Intl library compatibility

On 07/20/2016 05:43 PM, Dimitry Sibiryakov wrote: > 20.07.2016 16:35, Alex Peshkoff wrote: >> Does it have any practical meaning? > If I load Firebird 3 library in embedded mode and Firebird 2.5 Embedded > engine into the > same process space, I'd prefer them not to fig

Re: [Firebird-devel] Intl library compatibility

On 07/20/2016 04:56 PM, Dimitry Sibiryakov wrote: > Hello, All. > > Can Firebird 2.5 work with fbintl.dll from Firebird 3? > If built with same compiler - may be yes. Does it have any practical meaning? -- What

Re: [Firebird-devel] Unloading plugin

On 07/18/2016 10:14 PM, Jiří Činčura wrote: > By "engine" you mean server? yes, till the end of 'firebird' process > > -- > Mgr. Jiří Činčura > Independent IT Specialist > > > On Mon, Jul 18, 2016, at 14:15, Alex Peshkoff wrote: >> On 07/13/201

Re: [Firebird-devel] Unloading plugin

On 07/13/2016 11:38 AM, Jiří Činčura wrote: > I've let it sit idle, with no connections whatsoever, for 5 minutes and > the release method of plugin was never called. I can see the DLL loaded > in list of handles kept by firebird.exe, if I'm not attaching debugger. > > It's not even called when I

Re: [Firebird-devel] Unloading plugin

On 07/12/2016 05:16 PM, Jiří Činčura wrote: >>> 1 min after release of last reference >> I suppose "last reference" means after all (1 in my case) connections to >> the database are closed the 1 minute starts counting. Then the `release` >> method of IExternalEngineImpl will be called, right?

Re: [Firebird-devel] Minimal required set of interfaces for a provider

On 07/09/2016 03:55 PM, Dimitry Sibiryakov wrote: > Hello, All. > > Which interfaces a provider has to implement? > # grep -B 1 Impl EngineInterface.h -- class JBlob FB_FINAL : public Firebird::RefCntIface > -- class

Re: [Firebird-devel] Crash on developer build

On 07/07/2016 01:53 PM, Roman Simakov wrote: > The reason is an attempt to get config manager of DummyMasterImpl. The > root of call is here > Fixed it - at least for linux. Also had double fclose() issue. -- Attend

Re: [Firebird-devel] Centos7 and FB_CPU

On 06/29/2016 04:28 PM, Dimitry Sibiryakov wrote: > 29.06.2016 15:22, Alex Peshkoff wrote: >> We talk about correctness of our internal commands. > They are known to be incorrect, but they cannot be fixed because makefile > cannot commit > suicide

Re: [Firebird-devel] Centos7 and FB_CPU

On 06/29/2016 04:06 PM, Dimitry Sibiryakov wrote: > 29.06.2016 15:00, Alex Peshkoff wrote: >> That's possible. Did you try 'make clean_all'? > "git clean -d -x -f" is more effective. > > May be. Fresh checkout too :) We talk about correctn

Re: [Firebird-devel] Centos7 and FB_CPU

On 06/29/2016 02:56 PM, Paul Reeves wrote: > On Wed, 29 Jun 2016 12:41:28 +0300 > Alex Peshkoff <peshk...@mail.ru> wrote: > >> In common.h there is >> >> #ifdef AMD64 >> #define FB_CPU CpuAmd >> >> 2 master defines (-DLINUX -DAMD64) are pla

Re: [Firebird-devel] Centos7 and FB_CPU

On 06/29/2016 12:42 PM, Paul Reeves wrote: > On Wed, 29 Jun 2016 11:07:06 +0200 > Paul Reeves wrote: > >> I'll try building from the epel-7 source rpm and see if that works. >> > That seems to fix the FB_CPU problem > > And now I'm getting a new error : > > symbol lookup >

Re: [Firebird-devel] Centos7 and FB_CPU

On 06/29/2016 12:07 PM, Paul Reeves wrote: > On Wed, 29 Jun 2016 10:36:54 +0200 > Philippe Makowski wrote: > >>> I get this error: >>> >>>In file included from src/include/firebird.h:70:0, >>>from src/gpre/movg.cpp:28: src/include/../common/common.h:555:2: >>>

Re: [Firebird-devel] Settings values in ExternalEngine

On 06/22/2016 01:09 PM, Jiří Činčura wrote: >> input/output metadata (that's IMessageMetadata, yes?) provide you with a > Yes. > >> set of methods needed to access data in buffer correctly >> unavoidable one is getOffset() - it returns offset in a buffer where >> data is placed #include >>

Re: [Firebird-devel] Settings values in ExternalEngine

On 06/22/2016 09:21 AM, Jiří Činčura wrote: > Hello, > > when I finally grasped the FB_MESSAGE I realized that's not going to > help me. It generates struct and I have to know in advance the shape of > it. Which I don't know for whatever external procedure user writes. Yes, that works only for

Re: [Firebird-devel] Wire compression

On 06/20/2016 06:53 PM, Jiří Činčura wrote: > Can this be somewhat requested also by isc_info_firebird_version or > similar? > This is requested by isc_version() in legacy API and IUtil::getFbVersion() in new one. --

Re: [Firebird-devel] Settings values in ExternalEngine

On 06/20/2016 04:17 PM, Jiří Činčura wrote: > Hi *, > > forgot my lack of knowledge of possible helper functions, but are the > some functions to help manipulate memory with values in ExternalEngine? > > I.e. the IExternalProcedure::open has void* inMsg and void* outMsg. > Although I can dig out

Re: [Firebird-devel] Problem with Firebird 2.1 embedded and libncurses 6 (Linux)

On 06/19/2016 06:58 PM, Fulvio Senore wrote: > Thank you for your answer. I have spent some time trying to do it but I > have not been able to force the makefile to use the static ncurses > library instead of the so file. Unfortunately my knowledge of the build > chain is very limited. > > May

Re: [Firebird-devel] Wire compression

On 06/18/2016 06:10 PM, Jiří Činčura wrote: >> Do you receive pflag_compress set in accept->p_acpt_type? You should >> check this value to decide use compression later or not. > Yes, I do. I receive 261, which is ptype_lazy_send | pflag_compress, if > I'm not mistaken. > > Not sure where to look

Re: [Firebird-devel] Transaction id size in Firebird 3

On 06/17/2016 09:23 PM, Mark Rotteveel wrote: > On 2-5-2016 08:40, Dmitry Yemanov wrote: >> 01.05.2016 13:01, Mark Rotteveel wrote: But it seems that these inputs tags were forgotten re. their longer counterparts: isc_spb_rpr_commit_trans, isc_spb_rpr_rollback_trans,

Re: [Firebird-devel] Test bugs.core_3554 randomly fails

On 06/17/2016 07:11 PM, Dimitry Sibiryakov wrote: > 17.06.2016 17:58, Dmitry Yemanov wrote: >> Or perhaps do such a check in the remote client, or even better in the >> Y-valve? > In this case the test will fail always. > > Tests are also fixed sometimes ;)

Re: [Firebird-devel] Wire compression

On 06/17/2016 11:54 AM, Jiří Činčura wrote: > Hi, > >> Yes. May be the only detail is that zlib may compress data in slightly >> different ways. In our case data compressed not as a set of packets - >> the stream is compressed as a whole with flashes in the end of each >> packet. > Can i turn on

Re: [Firebird-devel] Problem with Firebird 2.1 embedded and libncurses 6 (Linux)

On 06/17/2016 04:12 PM, Fulvio Senore wrote: > I have developed a couple of open source programs that use Firebird as > an embedded database. The programs are some years old so they use > version 2.1. > > I am using an embedded setup under Linux with a copy of Firebird that I > compiled at the

Re: [Firebird-devel] building FB3 for valgrind

On 06/17/2016 10:25 AM, Paul Reeves wrote: > I tried to build FB3 for valgrind yesterday, without any success. > > AFAICT, all that is needed is to declare DEV_FLAGS with -DUSE_VALGRIND > and then run make Debug. > > The first build problem arrives with DELAYED_EXTENT_COUNT being used in >

Re: [Firebird-devel] GPRE, new API and static pointers

On 06/16/2016 07:35 PM, Dimitry Sibiryakov wrote: > Hello, All. > > Currently GPRE used to generate code like this: > >> static Firebird::IRequest* >> fb_396 = 0; /* request handle */ >> >> for (int retries = 0; retries < 2; ++retries) >> { >> { >>

Re: [Firebird-devel] Memory pool: memory block header

On 06/14/2016 05:50 PM, Dimitry Sibiryakov wrote: > 14.06.2016 16:43, Alex Peshkoff wrote: >> If you think it's pool's bug > No, I don't think that missing protection from double deallocation of > blocks is a bug. > > Suggested in your patch has one disadvantage - blo

Re: [Firebird-devel] Memory pool: memory block header

On 06/14/2016 05:46 PM, Dimitry Sibiryakov wrote: > 14.06.2016 16:30, Alex Peshkoff wrote: >> I've asked what does contain a block (i.e. it's hex >> dump) which is deallocated second time... > > Try to apply attached patch and run some tests. Double deallocation > seem

Re: [Firebird-devel] Memory pool: memory block header

On 06/14/2016 05:37 PM, Dimitry Sibiryakov wrote: > 14.06.2016 16:30, Alex Peshkoff wrote: >> That's nice but I've asked what does contain a block (i.e. it's hex >> dump) which is deallocated second time... > I think that you don't need bytes, Then I can't

Re: [Firebird-devel] Memory pool: memory block header

On 06/14/2016 05:22 PM, Dimitry Sibiryakov wrote: > 14.06.2016 16:17, Alex Peshkoff wrote: >> In dev_build - should. Except the head of a block which is reused for >> internal needs. How big is a block? What does it contain? > I got it on Savepoint. Quite small thi

Re: [Firebird-devel] Memory pool: memory block header

On 06/14/2016 05:03 PM, Dimitry Sibiryakov wrote: > 14.06.2016 15:59, Alex Peshkoff wrote: >>>and how to determine which member is >>>> valid? >>>> >>>> union >>>> { >>>>MemPool*pool; >> Use

Re: [Firebird-devel] Memory pool: memory block header

On 06/14/2016 04:14 PM, Dimitry Sibiryakov wrote: > Hello, All. > > Do anybody know why in MemHeader this is an union To save memory - that pointers are never used together. > and how to determine which member is > valid? > > union > { > MemPool*pool; Used when block is

Re: [Firebird-devel] Hostname parameters

On 06/13/2016 08:20 AM, Jiří Činčura wrote: >> AFAIR, server ignores (actually, removes) isc_dpb_host_name if it's >> found in the client-specified DPB. However, isc_dpb_host_name will be >> preserved for the embedded access (if it makes any sense). > What about CNCT_login vs isc_dpb_user_name? >

Re: [Firebird-devel] isc_action_svc_backup with isc_spb_bkp_skip_data format

On 06/13/2016 01:12 PM, Jiří Činčura wrote: > Looking into CrumpletReader I see: > case isc_action_svc_backup: > case isc_action_svc_restore: > switch (tag) > { > case isc_spb_bkp_file: >

Re: [Firebird-devel] System privileges and database creator

On 06/10/2016 05:59 PM, Dimitry Sibiryakov wrote: > Hello, All. > > Is it expected that database owner cannot run sweep on it? > No And it works for me: # ./isql -user x Use CONNECT or CREATE DATABASE to specify a database SQL> create database 'xxx'; SQL> show database; Database: xxx

Re: [Firebird-devel] [patch] spelling fixes for isc_sync.cpp and messages (perfrom -> perform)

On 03/14/2016 06:48 PM, Damyan Ivanov wrote: > -=| Damyan Ivanov, 14.03.2016 15:46:22 + |=- >> Attached is a patch fixing a couple of spelling mistakes. It is all >> about "perfrom" typo. > Forgot to mention that the patch was made against Firebird 3.0.0.32383 > from the B3_0_Release branch in

Re: [Firebird-devel] POSIX locales

On 06/09/2016 01:30 PM, Dimitry Sibiryakov wrote: > Hello, All. > > For Firebird 4 is there point to care about *nix with locales different > from utf-8? > If possible yes I think. What problems with non-utf8 on *nix? Does iconv() do it's work bad?

Re: [Firebird-devel] Longer metadata names and related things

On 06/07/2016 08:25 PM, Adriano dos Santos Fernandes wrote: > On 07/06/2016 13:42, Alex Peshkoff wrote: >> On 06/07/2016 07:11 PM, Dimitry Sibiryakov wrote: >>> 07.06.2016 18:06, Adriano dos Santos Fernandes wrote: >>>> How should we align this things as show table

Re: [Firebird-devel] Longer metadata names and related things

On 06/07/2016 07:11 PM, Dimitry Sibiryakov wrote: > 07.06.2016 18:06, Adriano dos Santos Fernandes wrote: >> How should we align this things as show table ? >> >> 1) If we do more or less nothing > It would be my choice. > > Agreed

[Firebird-devel] Fwd: [Firebird-checkins] [FirebirdSQL/firebird] 1a4f58: Optimize system privilege check - move permanent j...

Hi all, this commit changes system triggers 1 and 31 in not backward compatible way. Make sure to recreate all ODS13 databases, or you will get segfault or asserts (on dev build) when trying to grant / revoke some rights. Alex. Branch: refs/heads/master Home:

Re: [Firebird-devel] Changing auth plugin in the middle of handshake

On 06/06/2016 12:43 PM, Dimitry Sibiryakov wrote: > 06.06.2016 10:18, Alex Peshkoff wrote: >> Exactly this is not possible. But plugin A can decide that handshake is >> anyway failed and after it server will load plugin B (next in agreed >> between server & client list of

Re: [Firebird-devel] Changing auth plugin in the middle of handshake

On 06/04/2016 12:55 PM, Dimitry Sibiryakov wrote: > 04.06.2016 11:39, Dimitry Sibiryakov wrote: >> Server really can order client to use different plugin for continue >> handshaking?.. > I mean in case of sending op_cont_auth. For example, Srp plugin ask for > more data and > wand to get

Re: [Firebird-devel] Problem in Firebird-3.0.0.32483-0.amd64.tar.gz

> If it's release 3.0.0 it's known issue, fixed here Sorry, I've read everything except subject. :) Yes, this issue is fixed in both master and B3_0. -- What NetFlow Analyzer can do for you? Monitors network bandwidth

Re: [Firebird-devel] Problem in Firebird-3.0.0.32483-0.amd64.tar.gz

On 06/02/2016 04:15 AM, Adriano dos Santos Fernandes wrote: > Alex, > > Look at this: > > asfernandes@cheetah:/opt/firebird$ ll /opt/firebird/include/firebird/ > total 956 > drwxr-xr-x 3 root root 4096 Abr 14 11:12 ./ > drwxr-xr-x 3 root root 4096 Abr 14 11:12 ../ > -r--r--r-- 1 root root

<    5   6   7   8   9   10   11   12   13   14   >