Full Unicode/Ansi-Support in /testing branch
Moderators: gto, EgonHugeist, olehs
- EgonHugeist
- Zeos Project Manager
- Posts: 1936
- Joined: 31.03.2011, 22:38
miab3,
Yes i think so, Michal. If i've the time then this will be done this week. Imagine i've got a private live too.
But where i see no generic way, is to have the edit option for that type. Here i think you've got to update the rows with a statement by your selves. But accessing these files should work (i hope).
Michael
Yes i think so, Michal. If i've the time then this will be done this week. Imagine i've got a private live too.
But where i see no generic way, is to have the edit option for that type. Here i think you've got to update the rows with a statement by your selves. But accessing these files should work (i hope).
Michael
Best regards, Michael
You want to help? http://zeoslib.sourceforge.net/viewtopic.php?f=4&t=3671
You found a (possible) bug? Use the new bugtracker dude! http://sourceforge.net/p/zeoslib/tickets/
You want to help? http://zeoslib.sourceforge.net/viewtopic.php?f=4&t=3671
You found a (possible) bug? Use the new bugtracker dude! http://sourceforge.net/p/zeoslib/tickets/
@EgonHugeist
All right Michael,
For now you can make a small change in the file ZDbcCache.pas the 74th line:
TZByteArray = array [0 ..131062] of Byte; / / M.A. was 65535
// TZRowBuffer size - LazarusWin 131 072B; - Delphi 131 069B
I checked on:
Zeos7eh1439 (compiles and runs) D2006, DXE2 32/64, LazarusWin1.1.0/fpc 2.7.1 32/64
and does not slow down, and the memory consumption is negligibly higher.
(On earlier versions also)
I do not know how in Linux but it seems to me that it will be well too.
Michal
All right Michael,
For now you can make a small change in the file ZDbcCache.pas the 74th line:
TZByteArray = array [0 ..131062] of Byte; / / M.A. was 65535
// TZRowBuffer size - LazarusWin 131 072B; - Delphi 131 069B
I checked on:
Zeos7eh1439 (compiles and runs) D2006, DXE2 32/64, LazarusWin1.1.0/fpc 2.7.1 32/64
and does not slow down, and the memory consumption is negligibly higher.
(On earlier versions also)
I do not know how in Linux but it seems to me that it will be well too.
Michal
- EgonHugeist
- Zeos Project Manager
- Posts: 1936
- Joined: 31.03.2011, 22:38
miab3,
This guy is imbelievable! Addition Hotfixes on private line:
http://bugs.freepascal.org/view.php?id=22381
http://bugs.freepascal.org/view.php?id=22377
He's cleaning up our failing FPC tests like nobody alse did it before! Not Zeos had problems it was the FPC.
Also i was going two steps further with the BFILE. I didn't test it yes. My girlfriends birthday is today...
Michael
This guy is imbelievable! Addition Hotfixes on private line:
http://bugs.freepascal.org/view.php?id=22381
http://bugs.freepascal.org/view.php?id=22377
He's cleaning up our failing FPC tests like nobody alse did it before! Not Zeos had problems it was the FPC.
Also i was going two steps further with the BFILE. I didn't test it yes. My girlfriends birthday is today...
Michael
Best regards, Michael
You want to help? http://zeoslib.sourceforge.net/viewtopic.php?f=4&t=3671
You found a (possible) bug? Use the new bugtracker dude! http://sourceforge.net/p/zeoslib/tickets/
You want to help? http://zeoslib.sourceforge.net/viewtopic.php?f=4&t=3671
You found a (possible) bug? Use the new bugtracker dude! http://sourceforge.net/p/zeoslib/tickets/
@EgonHugeist
Well, this is even more interesting in Oracle (than BFILE):
http://psoug.org/reference/collections.html
http://www.orafaq.com/wiki/NESTED_TABLE
Here is an example table for the ZEOS reports an error when opening:
CREATE TABLE SYSTEM.AQ$_QUEUES
(
OID RAW(16),
EVENTID NUMBER NOT NULL,
"NAME" VARCHAR2(30) NOT NULL,
TABLE_OBJNO NUMBER NOT NULL,
USAGE NUMBER NOT NULL,
ENABLE_FLAG NUMBER NOT NULL,
MAX_RETRIES NUMBER,
RETRY_DELAY NUMBER,
PROPERTIES NUMBER,
RET_TIME NUMBER,
QUEUE_COMMENT VARCHAR2(2000),
SUBSCRIBERS SYS.AQ$_SUBSCRIBERS, <-------
MEMORY_THRESHOLD NUMBER,
SERVICE_NAME VARCHAR2(64),
NETWORK_NAME VARCHAR2(256),
CONSTRAINT AQ$_QUEUES_CHECK
UNIQUE(
"NAME",
TABLE_OBJNO)
Michal
Well, this is even more interesting in Oracle (than BFILE):
http://psoug.org/reference/collections.html
http://www.orafaq.com/wiki/NESTED_TABLE
Here is an example table for the ZEOS reports an error when opening:
CREATE TABLE SYSTEM.AQ$_QUEUES
(
OID RAW(16),
EVENTID NUMBER NOT NULL,
"NAME" VARCHAR2(30) NOT NULL,
TABLE_OBJNO NUMBER NOT NULL,
USAGE NUMBER NOT NULL,
ENABLE_FLAG NUMBER NOT NULL,
MAX_RETRIES NUMBER,
RETRY_DELAY NUMBER,
PROPERTIES NUMBER,
RET_TIME NUMBER,
QUEUE_COMMENT VARCHAR2(2000),
SUBSCRIBERS SYS.AQ$_SUBSCRIBERS, <-------
MEMORY_THRESHOLD NUMBER,
SERVICE_NAME VARCHAR2(64),
NETWORK_NAME VARCHAR2(256),
CONSTRAINT AQ$_QUEUES_CHECK
UNIQUE(
"NAME",
TABLE_OBJNO)
Michal
- EgonHugeist
- Zeos Project Manager
- Posts: 1936
- Joined: 31.03.2011, 22:38
@miab3,
Michal like you've recognized (i hope) is the Project-Manger MDeams and me the only people who maintain the Zeos-Code. Your bugreports a a great help for everybody and they are welcome. For me personally fatherhood is coming in at least three weeks. Which means i won't have the time to help everywhere and verytime. So i propose to grant you svn commit right's and you do some jobs by your selves. Start to run our testsuites. Learn to compare the results. If they are save for all compilers then commit them. I already wrote a little Lazarus app which makes it very easy to setup the testsuites ->\test\ZTestConfig. I'm planning to upgrade these app to have testgroups and to allow to run the same Driver in different environment settings like Ansi vs. Unicode, PreprepareSQL or not.
One member Andre van Zuydem is comming back and plans to implement the FB3 support and eventually (realy not promissed) the CUBRID support. Ludob did a great job for us to clear the FPC test results. An additional hotfix: http://bugs.freepascal.org/view.php?id=22387
Also i slightly do prefer to switch to the private contact if you doubt.
So what do you think about this?
Michael
Btw. BFILE is now supported Rev.1464
Michal like you've recognized (i hope) is the Project-Manger MDeams and me the only people who maintain the Zeos-Code. Your bugreports a a great help for everybody and they are welcome. For me personally fatherhood is coming in at least three weeks. Which means i won't have the time to help everywhere and verytime. So i propose to grant you svn commit right's and you do some jobs by your selves. Start to run our testsuites. Learn to compare the results. If they are save for all compilers then commit them. I already wrote a little Lazarus app which makes it very easy to setup the testsuites ->\test\ZTestConfig. I'm planning to upgrade these app to have testgroups and to allow to run the same Driver in different environment settings like Ansi vs. Unicode, PreprepareSQL or not.
One member Andre van Zuydem is comming back and plans to implement the FB3 support and eventually (realy not promissed) the CUBRID support. Ludob did a great job for us to clear the FPC test results. An additional hotfix: http://bugs.freepascal.org/view.php?id=22387
Also i slightly do prefer to switch to the private contact if you doubt.
So what do you think about this?
Michael
Btw. BFILE is now supported Rev.1464
Best regards, Michael
You want to help? http://zeoslib.sourceforge.net/viewtopic.php?f=4&t=3671
You found a (possible) bug? Use the new bugtracker dude! http://sourceforge.net/p/zeoslib/tickets/
You want to help? http://zeoslib.sourceforge.net/viewtopic.php?f=4&t=3671
You found a (possible) bug? Use the new bugtracker dude! http://sourceforge.net/p/zeoslib/tickets/
@EgonHugeist
Michael,
Zeos7eh1466 (compiles and runs) on D2006, DXE2 32/64, LazarusWin1.1.0/fpc 2.7.1 32/64.
It was necessary to make the change to PostgreSQL ZDbcPostgreSqlResultSet.pas
line 203: 24: ColumnInfo.Precision := 64; { regproc } // M.A. was 10
As for the problem with the type of pg_catalog.aclitem [] I suggest to disable array types at the moment, in PostgreSQL.
As for commiting to svn I have to well prepare the test environment before.
Michal
Michael,
Zeos7eh1466 (compiles and runs) on D2006, DXE2 32/64, LazarusWin1.1.0/fpc 2.7.1 32/64.
It was necessary to make the change to PostgreSQL ZDbcPostgreSqlResultSet.pas
line 203: 24: ColumnInfo.Precision := 64; { regproc } // M.A. was 10
As for the problem with the type of pg_catalog.aclitem [] I suggest to disable array types at the moment, in PostgreSQL.
As for commiting to svn I have to well prepare the test environment before.
Michal
- EgonHugeist
- Zeos Project Manager
- Posts: 1936
- Joined: 31.03.2011, 22:38
Slightly we get a litte developer community back!
Great news, Michal.
Michal like you can imagine are there small rules to work together. Things i've learned since joining the Zeos team:
Doing jobs like increasing the RowBuffer again should be discussed in the team or with me if we only talk about my branch. I already wrote you a pm for my contact emails. The second email can be equal used for convinience via google talk. For quick answer.
Remove or changing code from other members can be interpeted as rude. So discuss this changes previously. Fixing a compiler-broken code is out of that rule.
Mark is a great mind with a lot of expierences and he's doing the job since very long times. Contact him too if i'm not available. Write him a pm (MDeams). He's our guiding light and second pair of eyes. Waiting some days for realy is normal for all of us too.
If it is possible then merge the changes on eh and testing (a little deal with Mark).
Ludo want to her us with the FreeTDS support for Lazarus/unix and eventually with the DB2 support.
Michal in the past you did report us a lot of leaks we had. It is total normal to add spezial tests to our testsuite to keep control if everything is right. Here you can add your testtables too.
Hmm what's left? Have fun, Michal. Each help is welcome. And i need your sourceforge.net username to grant you access.
Michael
Great news, Michal.
Michal like you can imagine are there small rules to work together. Things i've learned since joining the Zeos team:
Doing jobs like increasing the RowBuffer again should be discussed in the team or with me if we only talk about my branch. I already wrote you a pm for my contact emails. The second email can be equal used for convinience via google talk. For quick answer.
Remove or changing code from other members can be interpeted as rude. So discuss this changes previously. Fixing a compiler-broken code is out of that rule.
Mark is a great mind with a lot of expierences and he's doing the job since very long times. Contact him too if i'm not available. Write him a pm (MDeams). He's our guiding light and second pair of eyes. Waiting some days for realy is normal for all of us too.
If it is possible then merge the changes on eh and testing (a little deal with Mark).
Ludo want to her us with the FreeTDS support for Lazarus/unix and eventually with the DB2 support.
Michal in the past you did report us a lot of leaks we had. It is total normal to add spezial tests to our testsuite to keep control if everything is right. Here you can add your testtables too.
Hmm what's left? Have fun, Michal. Each help is welcome. And i need your sourceforge.net username to grant you access.
Michael
Best regards, Michael
You want to help? http://zeoslib.sourceforge.net/viewtopic.php?f=4&t=3671
You found a (possible) bug? Use the new bugtracker dude! http://sourceforge.net/p/zeoslib/tickets/
You want to help? http://zeoslib.sourceforge.net/viewtopic.php?f=4&t=3671
You found a (possible) bug? Use the new bugtracker dude! http://sourceforge.net/p/zeoslib/tickets/
- EgonHugeist
- Zeos Project Manager
- Posts: 1936
- Joined: 31.03.2011, 22:38
@miab3,
Michal i do agree. Did you think about my proposal? Did you read my previous post?
Michael
Michal i do agree. Did you think about my proposal? Did you read my previous post?
Michael
Last edited by EgonHugeist on 10.07.2012, 16:42, edited 1 time in total.
Best regards, Michael
You want to help? http://zeoslib.sourceforge.net/viewtopic.php?f=4&t=3671
You found a (possible) bug? Use the new bugtracker dude! http://sourceforge.net/p/zeoslib/tickets/
You want to help? http://zeoslib.sourceforge.net/viewtopic.php?f=4&t=3671
You found a (possible) bug? Use the new bugtracker dude! http://sourceforge.net/p/zeoslib/tickets/
- EgonHugeist
- Zeos Project Manager
- Posts: 1936
- Joined: 31.03.2011, 22:38
miab3,
Ok Michal, no problem did commit your regproc fix. Thought you want to do it by your selves..
Patch done. Rev1474
Michael
Ok Michal, no problem did commit your regproc fix. Thought you want to do it by your selves..
Patch done. Rev1474
Michael
Best regards, Michael
You want to help? http://zeoslib.sourceforge.net/viewtopic.php?f=4&t=3671
You found a (possible) bug? Use the new bugtracker dude! http://sourceforge.net/p/zeoslib/tickets/
You want to help? http://zeoslib.sourceforge.net/viewtopic.php?f=4&t=3671
You found a (possible) bug? Use the new bugtracker dude! http://sourceforge.net/p/zeoslib/tickets/