Subject | RE: [Firebird-Java] JayBird - is a version update scehdule? |
---|---|
Author | Rick Debay |
Post date | 2005-03-31T16:27Z |
What is not being cached in auto-commit mode?
-----Original Message-----
From: Roman Rokytskyy [mailto:rrokytskyy@...]
Sent: Wednesday, March 30, 2005 1:58 PM
To: Firebird-Java@yahoogroups.com
Subject: Re: [Firebird-Java] JayBird - is a version update scehdule?
Ron,
- full support for Services API (backup/restore, user management, etc);
- support for updatable result sets
- execution plan in statements
- no caching in auto-commit mode
- reworked JCA code
First three issues are solved, last is almost solved (reconnecting an
in-limbo transaction for recovery is the only problem there), I am
currently doing big refactoring of the transaction management to fix
caching in auto-commit mode. Current state of the driver is not
acceptable even for alpha release, it's in the middle of refactoring. I
think I need one month to finish this task, and if there will be no
pending changes from Gabriel Reid and Steven Jardine, then we can start
with the release cycle.
I do not plan to release 1.5.6 version, but if there is demand for it -
I will check how hard would it be to backport fixes from HEAD to
Branch_1_5.
If that is not too complex, I will do this.
Can you please provide a list of bugfixes to port (the best would a list
of ChangeLog entries, file is available from CVS)?
Roman
Yahoo! Groups Links
-----Original Message-----
From: Roman Rokytskyy [mailto:rrokytskyy@...]
Sent: Wednesday, March 30, 2005 1:58 PM
To: Firebird-Java@yahoogroups.com
Subject: Re: [Firebird-Java] JayBird - is a version update scehdule?
Ron,
> What is the schedule for the next release of the JayBird driver?Next release will be 2.0 and will include many new features:
- full support for Services API (backup/restore, user management, etc);
- support for updatable result sets
- execution plan in statements
- no caching in auto-commit mode
- reworked JCA code
First three issues are solved, last is almost solved (reconnecting an
in-limbo transaction for recovery is the only problem there), I am
currently doing big refactoring of the transaction management to fix
caching in auto-commit mode. Current state of the driver is not
acceptable even for alpha release, it's in the middle of refactoring. I
think I need one month to finish this task, and if there will be no
pending changes from Gabriel Reid and Steven Jardine, then we can start
with the release cycle.
> The latest version is still 1.5.5, correct?Correct.
> Since December there have been multiple enchancments and bugs fixedetc.
> (?) - I am particularly interested in bugs related to memory leaks,
I do not plan to release 1.5.6 version, but if there is demand for it -
I will check how hard would it be to backport fixes from HEAD to
Branch_1_5.
If that is not too complex, I will do this.
Can you please provide a list of bugfixes to port (the best would a list
of ChangeLog entries, file is available from CVS)?
Roman
Yahoo! Groups Links