1

Topic: Firebird: passage on x64

Friends, prompt, I use Firebird Embedded
Whether it is necessary at passage with Firebird x86 on Firebird x64 to recreate a database or it it is not anchored to digit capacity?

2

Re: Firebird: passage on x64

scorpion235;
No, it is not necessary, to digit capacity it is not anchored.

3

Re: Firebird: passage on x64

scorpion235;
If are used any  - it is necessary to find or do their 64-digit versions. Or to refuse them.

4

Re: Firebird: passage on x64

scorpion235;
And here a question - and what for do transfer of application-oriented application with embedded on 64 bits? Storages do not suffice? Or what?

5

Re: Firebird: passage on x64

kdv wrote:

what for do transfer of application-oriented application with embedded on 64 bits?

Fashionably, stylishly, . All pass.

6

Re: Firebird: passage on x64

wadman wrote:

it is passed...
Fashionably, stylishly, . All pass.

There are two simple councils about usage FB:
1. Do not use 32-bit assembly of the server and the client.
2. Do not use the built in server.

7

Re: Firebird: passage on x64

Dmitry wrote:

1. Do not use 32-bit assembly of the server and the client.

It is possible more in detail?

8

Re: Firebird: passage on x64

Dmitry wrote:

Is two simple councils about usage FB:
1. Do not use 32-bit assembly of the server and the client.
2. Do not use the built in server.

Well and  xD

9

Re: Firebird: passage on x64

Dmitry wrote:

1. Do not use 32-bit assembly of the server and the client.

Then it is possible council: What to do, if at the client  32-bit? To put requirements?

10

Re: Firebird: passage on x64

YuRock;
It is the general spirit, probably. 32-bit systems it will not be fast. It is clear. But the file of a DB does not depend from , therefore an oddish message. Anyway at present. There would be a speech about an apple yes, there is no place to disappear.

11

Re: Firebird: passage on x64

KreatorXXI wrote:

YuRock;
It is the general spirit, probably. 32-bit systems it will not be fast. It is clear. But the file of a DB does not depend from , therefore an oddish message. Anyway at present. There would be a speech about an apple yes, there is no place to disappear.

Soon - concept extensible. For decades the extensible. And when in park of computers of the client is though one from 32-bit OS (and them now the big percent since was not and is not present sense to put on a computer with 1-2  64-bit OS, and it is not necessary for the user for operation any more), if on it does not work the system - it will not work anywhere.
To refuse that-whether this client? To wait for 20 years?

12

Re: Firebird: passage on x64

wrote:

there is a mass of the libraries, to new versions which at least Vista submit

It is a little not about that speech. There is a mass of computers on Windows 10 32-bit.

13

Re: Firebird: passage on x64

The author still while answered nothing.

14

Re: Firebird: passage on x64

CyberMax wrote:

it is passed...
It is possible more in detail?

What you are is specific particulars want to hear? Type an error in such line?
At "unexpected" problems (for example, at exhaustion of resources) the server and the client come to itself on a tail and fall/hang.
On 64 bits, this moment can be postponed for some time.
---
And it is a problem not only FB. It generally the global. Because, @#!, to write the reliable code very much it is expensive.

15

Re: Firebird: passage on x64

wrote:

And it is a problem not only FB. It generally the global. Because, @#!, to write the reliable code very expensively.

While there will be adherents of old versions Delphi, 32 the code anywhere does not get to.

16

Re: Firebird: passage on x64

kdv wrote:

what for do transfer of application-oriented application with embedded on 64 bits?

Components FibPlus and TJvMemoryData from JVCL work on x64 (a gain of productivity of 20-30 %) is more bright

17

Re: Firebird: passage on x64

scorpion235 wrote:

Components FibPlus and TJvMemoryData from JVCL work on x64 (a gain of productivity of 20-30 %)

is more bright
There are any acknowledgement?

18

Re: Firebird: passage on x64

scorpion235 wrote:

Components FibPlus is more bright and TJvMemoryData from JVCL work on x64 (a gain of productivity of 20-30 %)

Speed of operation of components of access to a DB is only an overhead charge for convenience. These expenditures should not be comparable with an operating time of the DBMS. They should be close to zero. And 0.2 * ~0 = ~0.
If it is truth, of course, in what to me is trusted with great difficulty. As though not on the contrary.

19

Re: Firebird: passage on x64

scorpion235;
With the loading test any hogwash full. At first the data is loaded in fibdataset, in storage, and then overloaded in memory dataset.  these double dancings? What if to change  on fibquery?

20

Re: Firebird: passage on x64

Dmitry wrote:

Is two simple councils about usage FB:
1. Do not use 32-bit assembly of the server and the client.
2. Do not use the built in server.

And nevertheless, why?

21

Re: Firebird: passage on x64

qwer098 wrote:

And nevertheless, why?

Yes . Because it is easy possible to use 32 the server and as it is quiet - embedded.
At it there any specific cases which in remaining 99 % have no sense.

22

Re: Firebird: passage on x64

qwer098 wrote:

it is passed...
And nevertheless, why?

1. I described above. And repeatedly swore here . In a month there will be a year as I final hammered on 32-bit components of the server.
At me recently it was screwed also 64-bit FB3 (in system resources ended - one driver did not release descriptors). Truth then all interesting fell a lot of...
So, of course, 64-bit not panacea.
2. Such things (multi-threaded DLL) should have a pair of functions - initialization and . As, for example, at WinSock. Initialization at the built in server are not present (?). And  is. If at you application with one output without explicit initialization it is possible to live. If something more difficult (issued in the form of DLL) - that most likely are problems with determination of the moment of a call .
1+2. If at you, at shortage of resources, application with the built in server, most likely, the basis will be damaged is screwed.
----
On "" tasks it is possible to use everything smile
----
And yes, here that I while can (carefully tell good about FB (3) - at shortage of a place on a disk it does not fall also basis does not spoil.

23

Re: Firebird: passage on x64

Dmitry wrote:

2. Such things (multi-threaded DLL) should have a pair of functions - initialization and . As, for example, at WinSock. Initialization at the built in server are not present (?). And  is.

It can be considered.

Dmitry wrote:

And yes, here that I while can (carefully tell good about FB (3) - at shortage of a place on a disk it does not fall also basis does not spoil.

It is inherited from 2.5 smile

24

Re: Firebird: passage on x64

Dmitry wrote:

2. Such things (multi-threaded DLL) should have a pair of functions - initialization and
. As, for example, at WinSock. Initialization at the built in server are not present (?). And
is. If at you application with one output, without explicit initialization
It is possible to live. If something more difficult (issued in the form of DLL) - that most likely arise
Problems with determination of the moment of a call .

Oh yes what there problems? MSDN to read yes a hand from the face to manage to unstick?.

25

Re: Firebird: passage on x64

Dmitry wrote:

1. I described above. And repeatedly swore here . In a month there will be a year as I final hammered on 32-bit components of the server.
At me recently it was screwed also 64-bit FB3 (in system resources ended - one driver did not release descriptors). Truth then all interesting fell a lot of...
So, of course, 64-bit not panacea.
2. Such things (multi-threaded DLL) should have a pair of functions - initialization and . As, for example, at WinSock. Initialization at the built in server are not present (?). And  is. If at you application with one output without explicit initialization it is possible to live. If something more difficult (issued in the form of DLL) - that most likely are problems with determination of the moment of a call .
1+2. If at you, at shortage of resources, application with the built in server, most likely, the basis will be damaged is screwed.
----
On "" tasks it is possible to use everything smile

I will sum up Dmitry's expressions:
1. There are cucumbers it is impossible in no event for the Universe collapse comes
...
(In small print)
10. Only provided that before it you stroked a baobab
11. And at you the tail of a jerboa oriented to North star has been clamped in the left hand