libdbi-devel — Discussion about development of the core library and DBD interface

Re: [libdbi-devel] libdbi-drivers-0.8.1 to be released soon

[Christian M. S. <cm...@ce...>]

Hi Markus,=20

On Sunday 18 September 2005 01.01, Markus Hoenicka wrote:
> Hi Christian,
>
> another bit of information. In another desperate attempt to see the
> firebird driver run ok I've built the firebird server on FreeBSD from
> the sources. The binary package I tried a while ago was broke as you
> may recall. The more recent version from the ports tree works ok from
> the command line, so I've got something new to play with.
>
> However, there's two bad news:
>
> 1) The firebird driver as built from the 0.8.0 sources would crash
>    applications that don't even use the firebird driver. The fact that
>    the driver is loaded by libdbi is apparently enough to wreak havoc
>    on an application even if it uses only e.g. the MySQL driver. This
>    appears to be a problem with the FreeBSD pthreads implementation
>    (which libfbclient.so depends on). I've changed acinclude.m4 in CVS
>    to first try libc_r and then libpthread, this apparently fixes the
>    problem. I have not seen this problem on Debian.
>
> 2) Using this version of Firebird on FreeBSD, I don't even get as far
>    as to the create table failure that I saw on Debian (albeit without
>    useful error messages, see my previous mail). test_dbi dumps core
>    when trying to connect to the freshly created database:
>
> #0  0x284fcce3 in THD_restore_specific () from
> /usr/local/lib/libfbclient.so.1 #1  0x28512120 in error () from
> /usr/local/lib/libfbclient.so.1
> #2  0x2850bd2e in REM_attach_database () from
> /usr/local/lib/libfbclient.so.1 #3  0x284fec97 in isc_attach_database ()
> from /usr/local/lib/libfbclient.so.1 #4  0x284deca8 in _dbd_real_connect =
()
> from /usr/local/lib/dbd/libfirebird.so #5  0x284dd901 in dbd_connect ()
> from /usr/local/lib/dbd/libfirebird.so #6  0x28089f49 in dbi_conn_connect
> () from /usr/local/lib/libdbi.so.0 #7  0x0804980f in main (argc=3D1,
> argv=3D0xbfbfe7f4) at test_dbi.c:84
>
> firebird.log contains two types of error messages (several entries for
> each of these per crash):
>
> yeti.mininet    Sun Sep 18 00:58:14 2005
>         SERVER/process_packet: connection rejected for markus
>
> yeti.mininet    Sun Sep 18 00:58:14 2005
>         SERVER/process_packet: connect reject, server exiting

I don't get these problems or error on my development platform (Linux Gento=
o).
That libdbi dumps core when a user have the wrong rights to connect is not =
the=20
right thing obviously. =20


> Once again, I can use the command-line client isql without any
> problems. It is only through libdbi that I keep getting problems.

Are you using isql as root?=20
Then you have =FCber-access rights, and can connect and perfrom task on eve=
ry=20
database on this server.=20


Its not the same thing as accessing the databases through the C API which u=
ses=20
the user you have supplied in your program.=20
Have you setupped a user with the gsec program?=20


I have just tested the "make check" option and on my platform with a user w=
ith=20
correct rights, libdbi can create tables.  Both from "make check"'s test=20
program and other applications written that uses libdbi.

I will spend some time today trying to get the test program to run on Linux=
=20
and then it should run on freebsd as well.=20



regards,=20

Christian

Re: [libdbi-devel] libdbi-drivers-0.8.1 to be released soon

[Markus H. <mar...@mh...>]

Hi Christian,

another bit of information. In another desperate attempt to see the
firebird driver run ok I've built the firebird server on FreeBSD from
the sources. The binary package I tried a while ago was broke as you
may recall. The more recent version from the ports tree works ok from
the command line, so I've got something new to play with.

However, there's two bad news:

1) The firebird driver as built from the 0.8.0 sources would crash
   applications that don't even use the firebird driver. The fact that
   the driver is loaded by libdbi is apparently enough to wreak havoc
   on an application even if it uses only e.g. the MySQL driver. This
   appears to be a problem with the FreeBSD pthreads implementation
   (which libfbclient.so depends on). I've changed acinclude.m4 in CVS
   to first try libc_r and then libpthread, this apparently fixes the
   problem. I have not seen this problem on Debian.

2) Using this version of Firebird on FreeBSD, I don't even get as far
   as to the create table failure that I saw on Debian (albeit without
   useful error messages, see my previous mail). test_dbi dumps core
   when trying to connect to the freshly created database:

#0  0x284fcce3 in THD_restore_specific () from /usr/local/lib/libfbclient.so.1
#1  0x28512120 in error () from /usr/local/lib/libfbclient.so.1
#2  0x2850bd2e in REM_attach_database () from /usr/local/lib/libfbclient.so.1
#3  0x284fec97 in isc_attach_database () from /usr/local/lib/libfbclient.so.1
#4  0x284deca8 in _dbd_real_connect () from /usr/local/lib/dbd/libfirebird.so
#5  0x284dd901 in dbd_connect () from /usr/local/lib/dbd/libfirebird.so
#6  0x28089f49 in dbi_conn_connect () from /usr/local/lib/libdbi.so.0
#7  0x0804980f in main (argc=1, argv=0xbfbfe7f4) at test_dbi.c:84

firebird.log contains two types of error messages (several entries for
each of these per crash):

yeti.mininet    Sun Sep 18 00:58:14 2005
        SERVER/process_packet: connection rejected for markus

yeti.mininet    Sun Sep 18 00:58:14 2005
        SERVER/process_packet: connect reject, server exiting


Once again, I can use the command-line client isql without any
problems. It is only through libdbi that I keep getting problems.

I'm sorry that I can't contribute much to fix these problems. If you
can make sense of the backtrace, or if you want me to test further
things on FreeBSD, just let me know.

regards,
Markus

Markus Hoenicka writes:
 > Hi Christian,
 > 
 > I've spent another two hours trying to sort this out but failed. Now I
 > have a closer understanding as to how firebird handles access control,
 > but at the moment this is only sufficient to create databases and
 > tables using the isql command-line tool.
 > 
 > The real bad thing is that firebird does not run at all on FreeBSD and
 > is broke on Debian. All error messages on Debian read like "cannot
 > format message" which means that the firebird.msg file is
 > corrupt. Therefore I don't get useful diagnostics when running
 > libdbi. Is there any platform where firebird actually runs? BTW I'm
 > running firebird 1.5.1. If you happen to use the same version, could
 > you send me your firebird.msg file? I'd just like to try.
 > 
 > Still I can't see why the test program should fail if other apps using
 > the firebird driver work. The test program is very simple and does not
 > attempt to do weird things. It must be possible to fix the test
 > program.
 > 
 > regards,
 > Markus
 > 
 > Christian M. Stamgren writes:
 >  > I guess, You have to take my word for it. Why the test program cant create 
 >  > tables have something to do with the user not having the right access 
 >  > permissions. I have tried to sort that out but failed. 
 >  > 
 > 
 > 
 > -- 
 > Markus Hoenicka
 > mar...@ca...
 > (Spam-protected email: replace the quadrupeds with "mhoenicka")
 > http://www.mhoenicka.de
 > 
 > 
 > 
 > -------------------------------------------------------
 > SF.Net email is sponsored by:
 > Tame your development challenges with Apache's Geronimo App Server. Download
 > it for free - -and be entered to win a 42" plasma tv or your very own
 > Sony(tm)PSP.  Click here to play: http://sourceforge.net/geronimo.php
 > _______________________________________________
 > libdbi-devel mailing list
 > lib...@li...
 > https://lists.sourceforge.net/lists/listinfo/libdbi-devel
 > 

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: [libdbi-devel] libdbi-drivers-0.8.1 to be released soon

[Markus H. <mar...@mh...>]

Hi Christian,

I've spent another two hours trying to sort this out but failed. Now I
have a closer understanding as to how firebird handles access control,
but at the moment this is only sufficient to create databases and
tables using the isql command-line tool.

The real bad thing is that firebird does not run at all on FreeBSD and
is broke on Debian. All error messages on Debian read like "cannot
format message" which means that the firebird.msg file is
corrupt. Therefore I don't get useful diagnostics when running
libdbi. Is there any platform where firebird actually runs? BTW I'm
running firebird 1.5.1. If you happen to use the same version, could
you send me your firebird.msg file? I'd just like to try.

Still I can't see why the test program should fail if other apps using
the firebird driver work. The test program is very simple and does not
attempt to do weird things. It must be possible to fix the test
program.

regards,
Markus

Christian M. Stamgren writes:
 > I guess, You have to take my word for it. Why the test program cant create 
 > tables have something to do with the user not having the right access 
 > permissions. I have tried to sort that out but failed. 
 > 


-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

[libdbi-devel] $BB(%O%a!*B(%(%C%A!*(B

[<in...@ju...>]

　http://free-deai.1192296.net/

おいらのモットーと言えば、即ハメ・即Ｈ！
だって、出来るだけたくさんの子としたいし、
可愛い娘を見つけたら即行ハメたいと思うのはオイラだけ？
できれば、メールの駆け引きとか面倒なことはしたくないよね？

　http://free-deai.1192296.net/

今回のターゲットはこの子デース！
名前：まき
年齢：22歳
まきちゃんは現在、某ファーストフード店でアルバイト中。

　http://free-deai.1192296.net/

拒否はこちら。
 　 don...@ya...

.

Re: [libdbi-devel] libdbi-drivers-0.8.1 to be released soon

[Markus H. <mar...@mh...>]

Hi Christian,

"Christian M. Stamgren" <cm...@ce...> was heard to say:

> I guess, You have to take my word for it. Why the test program cant create
> tables have something to do with the user not having the right access
> permissions. I have tried to sort that out but failed.
>
> Ofcourse I see the test program as a very good thing to have, but I really
> dont see that it should be used as the things that decides if a driver goes
> into stable or not.
>

Please don't take me wrong. It's not like I'm saying I don't trust you that you
have the driver working ok. The problem is, if even I as a libdbi developer
can't get the driver to do the right thing for me, how should a plain user out
there make good use of the driver? I promise I'll look into the access rights
stuff again, it may very well be that got something wrong. But as things are at
the moment, I can create tables from the command line, but not through the
libdbi driver using the very same username/password combo.


> Ok fair enough. Is the CVS taged for 8.0 so that we can start to commit
> things
> that are going into 0.9?
>

Yes, both libdbi and libdbi-drivers are tagged and ready for new commits.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: [libdbi-devel] libdbi-drivers-0.8.1 to be released soon

[Markus H. <mar...@mh...>]

"Christian M. Stamgren" <cm...@ce...> was heard to say:

> BTW, Gentoo have 0.8.0 included and support for
> Mysql, Oracle, Postgres and Sqlite drivers.
>

Oh, cool. Last time I talked to Ashish he told me the Oracle driver was not done
yet. Apparently it works better than he thought.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: [libdbi-devel] libdbi-drivers-0.8.1 to be released soon

[Christian M. S. <cm...@ce...>]

Hi again,

On Sunday 11 September 2005 22.41, Markus Hoenicka wrote:
[..]
> This would require moving to 0.9.0 two weeks after releasing
> 0.8.0. I'm afraid this is going to confuse packagers. So far I've seen
> 0.8.0 only as a Fink package, thanks to Peter O'Gorman.

BTW, Gentoo have 0.8.0 included and support for 
Mysql, Oracle, Postgres and Sqlite drivers. 

Regards, 

Christian

Re: [libdbi-devel] libdbi-drivers-0.8.1 to be released soon

[Christian M. S. <cm...@ce...>]

Hi Markus, 

On Sunday 11 September 2005 22.41, Markus Hoenicka wrote:
> Hi Christian,
>
> Christian M. Stamgren writes:
>  > The firebird driver is working as it should.
>  > Its just the test program that doesn't work with firebird. I don't see
>  > that as a reason to not include it.
>
> Ok, but the test program is currently my only way to test the
> driver. Last time I tried I was not able to create tables. Do you have
> a simple program on your box that is suitable to do some testing on
> Debian?

No not at the moment. 
I guess, You have to take my word for it. Why the test program cant create 
tables have something to do with the user not having the right access 
permissions. I have tried to sort that out but failed. 

Ofcourse I see the test program as a very good thing to have, but I really 
dont see that it should be used as the things that decides if a driver goes 
into stable or not.   


> I'd suggest to invest some more work into 0.9.0. A couple of things
> come to mind, like support for sequences, a create_db() function which
> handles encoding stuff transparently and so on.


Ok fair enough. Is the CVS taged for 8.0 so that we can start to commit things 
that are going into 0.9? 



Regards, 

Christian

Re: [libdbi-devel] libdbi-drivers-0.8.1 to be released soon

[Markus H. <mar...@mh...>]

Hi Christian,

Christian M. Stamgren writes:
 > The firebird driver is working as it should. 
 > Its just the test program that doesn't work with firebird. I don't see that as 
 > a reason to not include it. 
 > 

Ok, but the test program is currently my only way to test the
driver. Last time I tried I was not able to create tables. Do you have
a simple program on your box that is suitable to do some testing on
Debian?

 > I would really like to include the patch from Henrique that brings uniform 
 > transaction support to libdbi included before the next release. I was just on 
 > my way to submit that patch to cvs. 
 > 
 > What do you think about that? 
 > 

This would require moving to 0.9.0 two weeks after releasing
0.8.0. I'm afraid this is going to confuse packagers. So far I've seen
0.8.0 only as a Fink package, thanks to Peter O'Gorman.

I'd suggest to invest some more work into 0.9.0. A couple of things
come to mind, like support for sequences, a create_db() function which
handles encoding stuff transparently and so on.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: [libdbi-devel] libdbi-drivers-0.8.1 to be released soon

[Christian M. S. <cm...@ce...>]

On Sunday 11 September 2005 21.44, Markus Hoenicka wrote:
> Hi fellow-developers,
>
> I've spotted a bug that hits both the sqlite and the sqlite3
> driver. In some cases the parser that isolates the field and table
> names in order to determine the data type of a field fails to get the
> table name right. This requires an update of the drivers as soon as
> possible.
>
> I intend to release 0.8.1 somewhere during the next week. I'll include
> Vadym's FreeTDS driver, although there is no feedback yet from someone
> else. My question is, is there any chance to include the Firebird and
> Oracle drivers too? I could spend a few hours during the next week
> trying to fix the Firebird driver, but there's nothing I can do about
> Oracle.

The firebird driver is working as it should. 
Its just the test program that doesn't work with firebird. I don't see that as 
a reason to not include it. 

About the Oracle driver. This driver worked good before the API changes for 
8.0. So to get the Oracle driver up speed should be straight forward.
Finally I'm starting to have some more time for projects like this again. So 
if somebody will just give me access to an Oracle server I will do the 
changes that is needed quickly.


I would really like to include the patch from Henrique that brings uniform 
transaction support to libdbi included before the next release. I was just on 
my way to submit that patch to cvs. 

What do you think about that? 

Christian

[libdbi-devel] libdbi-drivers-0.8.1 to be released soon

[Markus H. <mar...@mh...>]

Hi fellow-developers,

I've spotted a bug that hits both the sqlite and the sqlite3
driver. In some cases the parser that isolates the field and table
names in order to determine the data type of a field fails to get the
table name right. This requires an update of the drivers as soon as
possible.

I intend to release 0.8.1 somewhere during the next week. I'll include
Vadym's FreeTDS driver, although there is no feedback yet from someone
else. My question is, is there any chance to include the Firebird and
Oracle drivers too? I could spend a few hours during the next week
trying to fix the Firebird driver, but there's nothing I can do about
Oracle.

Christian, Ashsish, please let me know about your plans.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: Re: [libdbi-devel] libdbi

[Henrique <on...@ya...>]

> I don't see a reson for this. The engines will
handle
> transactions internaly the only thing needed to be
> provided by the libdbi user is the name of the
> savepoint so that she/he can rollback with that
> savepoint as an argument.

Ok. Though, like I as said, isn't all databases that
will offers support to savepoints, but all of
them offers (a minimum) support to transactions.
So, maybe makes sense create news functions to
API that *all* databases implement, and not what
only one or two do.

By other side, I think that savepoints is something
that is really in the roadmap for the "toy"
databases and seems that is more interesting works
with it instead pure transactions

> A transcation will only be tied to a connection
> anyway so why extract it in to some other level
> where it doesn't belong?

My initial (bad) idea was offers savepoints support
based in sample transactions for databases that don't
support this.  So, removing the functions for
savepoints and implementing only a "native support"
for transactions seems the right way.  But, as you
said, we don't will implement an interface for a
support that don't exists in the real database, or
in another words, we don't will implement savepoints
if the database doesn't support this.

> I think the interface provided in your email is
> a sane way of implementing transaction support.

Yeah. I think the same now.

--
Henrique



__________________________________________________
Converse com seus amigos em tempo real com o Yahoo! Messenger 
http://br.download.yahoo.com/messenger/ 

Re: [libdbi-devel] libdbi

[Christian M. S. <chr...@st...>]

Hi, 

On Thursday 01 September 2005 23.32, Markus Hoenicka wrote:
> [I've cc'ed the list as this is of general interest. Please keep these
> discussions on the list. The discussion will be archived and it is
> open for everyone to join in]
>
> Henrique writes:
>  > The functions proposed was:
>  >
>  > dbi_conn_commit(dbi_connt conn);
>  > dbi_conn_rollback(dbi_conn conn, const char
>  > *savepoint);
>  > dbi_conn_rollback_all(dbi_conn conn);
>  > dbi_conn_set_savepoint(dbi_conn conn, const char
>  > *savepoint);
>  >
>  > Though, I think that we "must" offers  a "native"
>  > transactions support instead "const strings"
>  > savepoints, something like creating a new
>  > dbi_transaction data type, what do you think?

I don't see a reson for this. The engines will handle transactions internaly 
the only thing needed to be provided by the libdbi user is the name of the 
savepoint so that she/he can rollback with that savepoint as an argument.

A transcation will only be tied to a connection anyway so why extract it in to 
some other level where it doesn't belong?

I think the interface provided in your email is a sane way of implementing 
transaction support.  

Regards, 

Christian

[libdbi-devel] libdbi

[Markus H. <mar...@mh...>]

[I've cc'ed the list as this is of general interest. Please keep these
discussions on the list. The discussion will be archived and it is
open for everyone to join in]

Henrique writes:
 > The functions proposed was:
 > 
 > dbi_conn_commit(dbi_connt conn);
 > dbi_conn_rollback(dbi_conn conn, const char
 > *savepoint);
 > dbi_conn_rollback_all(dbi_conn conn);
 > dbi_conn_set_savepoint(dbi_conn conn, const char
 > *savepoint);
 > 
 > Though, I think that we "must" offers  a "native"
 > transactions support instead "const strings"
 > savepoints, something like creating a new
 > dbi_transaction data type, what do you think?
 > 

Sure, if you can handle this transparently. We need a similar
abstraction for auto-incrementing columns which in some engines
require that you know the name of the sequence.

 > Note that sqlite driver will be a problem, since it
 > don't offers a complete transactions support, so we'll
 > can works with one, and only one, opened transaction
 > at time :(
 > 

Not only SQLite. mSQL does not support transactions at all. MySQL
supports transactions only since approx. 4.0, and only if you use a
transaction-safe table type. If you run the fastest tables (MyISAM),
there is no transaction support (although sending a BEGIN or COMMIT
does not seem to cause an error).

 > An another thing, I was thinking about offers tables
 > and columns management, i.e, the library's user don't
 > will need create and use tables and its columns by
 > simple query strings, a dbi datatype will be provide
 > for this.  I don't have a plan to this, but if you
 > like the idea I can create some document or works more
 > on this.
 > 

Could you please elaborate on this a little further? I'm not sure
whether I understand you here.

 > PS: Sorry by my bad, very bad, english
 > 

Don't worry. The Portugese of most list subscribers is far worse. All
simply do their best.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

[libdbi-devel] 0.8.0 release date

[Markus H. <mar...@mh...>]

Hi there,

we're already past the previously announced release date, and I'd like to avoid
that this process keeps dragging on. Unless any of you has objections, I'd like
to fix Sunday, Aug. 28, as the release date. I think we should release whatever
passes make check, and leave everything else to a libdbi-drivers-0.8.1 release.
I'll try to fix the debian build code until Sunday, so that the .debs can be
packaged by the Debian folks without any downstream patches.

As of today the picture is as follows:

libdbi framework: builds and runs on FreeBSD, Debian, FC4 (using gcc4), Solaris,
Windows/Cygwin
mysql, sqlite3: passes make check on FreeBSD, Cygwin and on Debian
pgsql, sqlite: passes make check on FreeBSD
firebird: fails make check on Debian
msql: no test report available
oracle: test suite not yet adapted to this driver
freetds: kinda works, but not yet added to the codebase (in fact, I don't know
whether the guys developing the driver want to contribute it or have some other
plans with it)

If you can add any success stories, please let me know. I plan to leave all
drivers in the distribution, but change the configure switch to something like
--with-oracle-experimental for those who do not pass make check yet.

Again, if you have any objections, please speak up now. The current releases
with their bugs and incomplete documentation have been out there far too long,
and I'd rather have new releases without additional drivers than any further
delay in the release process. We can add more drivers anytime later.

regards,
Markus


-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: [libdbi-devel] Patch(over 0.8.0pre3) for compiling libdbi core on mingw+msys

[Ashish R. <ash...@ya...>]

Actually it is already a known issue for mingw in
general , but it is going to give a problem only in
some cases of multi-threaded programs. And already in
libdbi, there is no guranteed careful study on
re-entrant code (as much i know), hence for now we can
afford to do with it. Later on i will take up this
issue separately for mingw+msys enhancements. My right
now aim is let the libdbi be up and running on
mingw+msys for the non-reentrant code. As mingw has a
large user-base in open surce community on ms-windows.

bye :-)
Ashish
-- Markus Hoenicka <mar...@mh...>
wrote:

> Hi Ashish,
> 
> yet another question. You've redefined strtok_r() as
> strtok() for
> Mingw, supposedly because the latter does not have
> the reentrant
> form. Does this change do any harm that the Mingw
> users better knew
> about?
> 
> regards,
> Markus
> 
> Ashish Ranjan writes:
>  > I had released one addon patch to compile libdbi
>  > (core) 0.7.x on mingw+msys .Today, i ported that
> for
>  > 0.8.0 pre3 code base, and here i submit my
> changes. If
>  > it is incorporated, then on windows , we can
> compile
>  > libdbi on mingw+msys (apart from the cygwin) out
> of
>  > the box.
>  > 
> 
> -- 
> Markus Hoenicka
> mar...@ca...
> (Spam-protected email: replace the quadrupeds with
> "mhoenicka")
> http://www.mhoenicka.de
> 
> 
> 
>
-------------------------------------------------------
> SF.Net email is Sponsored by the Better Software
> Conference & EXPO
> September 19-22, 2005 * San Francisco, CA *
> Development Lifecycle Practices
> Agile & Plan-Driven Development * Managing Projects
> & Teams * Testing & QA
> Security * Process Improvement & Measurement *
> http://www.sqe.com/bsce5sf
> _______________________________________________
> libdbi-devel mailing list
> lib...@li...
>
https://lists.sourceforge.net/lists/listinfo/libdbi-devel
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

[libdbi-devel] Patch(over 0.8.0pre3) for compiling libdbi core on mingw+msys

[Markus H. <mar...@mh...>]

Hi Ashish,

I've applied the patches in a slightly modified form. If you find some
time please check out the current libdbi cvs version and see whether
it works correctly on Mingw.

Thanks again
Markus

Ashish Ranjan writes:
 > Dear markus,
 > 
 > I hope that i can submit the patches to libdbi core
 > too.
 > 

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

[libdbi-devel] Patch(over 0.8.0pre3) for compiling libdbi core on mingw+msys

[Markus H. <mar...@mh...>]

Hi Ashish,

yet another question. You've redefined strtok_r() as strtok() for
Mingw, supposedly because the latter does not have the reentrant
form. Does this change do any harm that the Mingw users better knew
about?

regards,
Markus

Ashish Ranjan writes:
 > I had released one addon patch to compile libdbi
 > (core) 0.7.x on mingw+msys .Today, i ported that for
 > 0.8.0 pre3 code base, and here i submit my changes. If
 > it is incorporated, then on windows , we can compile
 > libdbi on mingw+msys (apart from the cygwin) out of
 > the box.
 > 

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: [libdbi-devel] Patch(over 0.8.0pre3) for compiling libdbi core on mingw+msys

[Markus H. <mar...@mh...>]

Hi Ashish,

I forgot to ask something: libdbi and libdbi-drivers both have README.w32 files
which currently explain how to compile and install things on Cygwin. These
files would be the appropriate place to explain usage for mingw as well. Why
don't you send a few lines that explain how to get libdbi and the drivers up
and running using mingw? I'd like to include this into the readmes.

regards,
Markus

Ashish Ranjan <ash...@ya...> was heard to say:

> Mingw (www.mingw.org) is port of gnu compiler which
> output native executable on windows platform, unlike
> cygwin. And msys (see, www.mingw.org) is a minimal
> posix environment for windows, it gives us autoconf
> etc tools on windows platform. Biggest advantage over
> cygwin is that there is no runtime emulation done at
> any layer, all the native. You can even cross compile
> using mingw tools, ie sitting on a linux box, you can
> output the native executables for window machines.
> There are other IDE environments also (apart from
> msys), which use mingw in the backend to generate the
> output .exe , like Dev-C++ etc.
>



-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

Re: [libdbi-devel] Patch(over 0.8.0pre3) for compiling libdbi core on mingw+msys

[Markus H. <mar...@mh...>]

Hi Ashish,

Ashish Ranjan <ash...@ya...> was heard to say:

> I had released one addon patch to compile libdbi
> (core) 0.7.x on mingw+msys .Today, i ported that for
> 0.8.0 pre3 code base, and here i submit my changes. If
> it is incorporated, then on windows , we can compile
> libdbi on mingw+msys (apart from the cygwin) out of
> the box.
>

I recall your earlier attempt, but I wasn't promoted to libdbi maintainer back
then, so I didn't follow up on this. Your patches are very welcome, of course.

> Please note that Change 1.1 and 1.2 involve only
> renaming two variables

No big deal.

> And Change 1.3 and 2.1 only are significant change,
> but these are put inside
>     #ifdef ..
> 	  ...
> 	#elif __MINGW32__
> 	 ..
> 	#else
> 	 ..
> 	#endif.

OSX needs special treatment at this point too, so this is nothing to worry
about.

> NOTE: by default , there is no directory /usr/local in msys, though /usr
> exists. So you should mkdir /usr/local, so as use /usr/local/lib etc.

I believe this can be handled when configuring libdbi. Something like
./configure --prefix=/usr will install libdbi in /usr/lib. Even if you use
/usr/local, the mkinstalldirs script should create it. Or does mingw bypass
configure and the configure-created makefiles altogether?

Two minor issues:

- It is easier for me to look at diffs instead of at modified full files. If
possible, send the output of diff -U 3 oldfile newfile > oldfile.diff next
time.

- You've edited the source files on Windows which introduces \r\n linebreaks. If
possible, please use an editor that creates Unix-style linebreaks (jedit,
NTEmacs  or one of many other free programmer's editors for Windows).

I'll use and clean up the files you sent this time though.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

[libdbi-devel] Patch(over 0.8.0pre3) for compiling libdbi core on mingw+msys

[Ashish R. <ash...@ya...>]

Dear markus,

I hope that i can submit the patches to libdbi core
too.

Mingw (www.mingw.org) is port of gnu compiler which
output native executable on windows platform, unlike
cygwin. And msys (see, www.mingw.org) is a minimal
posix environment for windows, it gives us autoconf
etc tools on windows platform. Biggest advantage over
cygwin is that there is no runtime emulation done at
any layer, all the native. You can even cross compile
using mingw tools, ie sitting on a linux box, you can
output the native executables for window machines.
There are other IDE environments also (apart from
msys), which use mingw in the backend to generate the
output .exe , like Dev-C++ etc.

I had released one addon patch to compile libdbi
(core) 0.7.x on mingw+msys .Today, i ported that for
0.8.0 pre3 code base, and here i submit my changes. If
it is incorporated, then on windows , we can compile
libdbi on mingw+msys (apart from the cygwin) out of
the box.

The zip file contains CHANGES_DETAIL_0.8.0pre3.txt, in
which there is all detail as to what changes have been
made in the original file.

Please note that Change 1.1 and 1.2 involve only
renaming two variables
And Change 1.3 and 2.1 only are significant change,
but these are put inside 
    #ifdef .. 
	  ...
	#elif __MINGW32__  
	 .. 
	#else
	 ..
	#endif. 
Hence there will be no re-testing needed for the
modified code on other platforms, like Linux/BSD,AIX
etc due to these changes/additions only.
  The newly added functions have been named like
win_... , because this implementation is not mingw
specific. all those functions will help in compiling
it on VC++ also, if someone tried to port it.
bye :-)
Ashish Ranjan
ash...@ya...
PS: i am mailing the actual patch file separately to you.


	
		
__________________________________ 
Do you Yahoo!? 
Yahoo! Mail - You care about security. So do we. 
http://promotions.yahoo.com/new_mail

[libdbi-devel] DETAIL OF CHANGES DONE IN Patch(over 0.8.0pre3) for compiling libdbi core on mingw+msys

[Ashish R. <ash...@ya...>]

here are the details of the changes done

changes in file	dbi_main.c
---------------------------
For porting libdbi to mingw+msys, I have used
libdbi-0.8.0-pre3 version
change 1.1) linenumber 1192, 1218,1233 of modified
file(ie, 1175,1201,1216 of original file)=errno
changed to errno1

change 1.2)linenumber 122, 311,319,327,335,
343,351,359(ie, line 105,294,302,310,318,326,334,342
of original file ) =ERROR changed to ERROR1

change 1.3)linenumber 51 to 66,1355 to 1410 of
modified file =newly added
win_dlopen,win_dlsym,win_dlclose,win_dlerror	

changes in file dbi_result.c
-----------------------------
change 2.1)linenumber 40 to 44 of the modified
file=strtok_r defined

==============================================================================
note: For easy viewing of what i have changed
       1)I have commented my all  changed source code
IN A LARGE-BLOCK within 
               /*Patch1:start*/
               ...........
               ...........
               ...........
              /*Patch1:end*/


       2) I have commented my all  changed source code
IN A SINGLELINE AS
              .............  /*Patch1*/

These marks may be removed.
==================================================================
==================================================================
NOTE: by default , there is no directory /usr/local in
msys, though /usr exists. So you should mkdir
/usr/local, so as use /usr/local/lib etc.


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

[libdbi-devel] 0.8.0 status

[Markus H. <mar...@mh...>]

Hi,

I'd just like to summarize my current state of affairs regarding the upcoming
0.8.0 releases of libdbi and libdbi-drivers.

I think the framework is release-worthy after the changes that I've announced in
previous mails. The mysql, pgsql, sqlite, and sqlite3 drivers were adapted to
the new framework and pass all tests too.

I've tested the code on the following platforms without problems:

FreeBSD 5.4
Debian 3.1
Fedora Core 4 (using gcc4.0)
Windows XP/Cygwin

Badly missing from this picture are OSX or PPC-Linux (to check for big-endian
vs. little-endian issues) and Solaris or 64bit Linux (to check for 64bit
cleanliness). Any help in these areas would be greatly appreciated.

You may have guessed that I'd also like to tighten the thumb screws of the
developers another turn. I'd really, really like to release the code asap. I've
managed to install Firebird on a Debian box. I'm not familiar with this engine
at all, but I could try to help if Christian won't get round to fix the driver
in the next couple of days. I can't do much about the msql and oracle drivers
at this time.

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

[libdbi-devel] ANN: libdbi-0.8.0-pre3 available for testing

[Markus H. <mar...@mh...>]

Hi all,

I've packaged the current cvs versions of libdbi and libdbi-drivers:

- a couple of build errors and warnings were fixed

- the dbd_get_encoding() function now also works for MySQL prior to 4.1. I've
tested this on Debian using 4.0.24, which is probably the oldest version in the
default installation of any of the popular end-user OSes. Frankly, I don't know
what happens with 3.23, but if you still use that it's probably about time to
upgrade.

- a new function dbi_conn_get_engine_version() was added. This function
retrieves a numeric representation of the version of the database engine that
serves the current connection. See the manual or the sources for the formula
used to compute the number.

- As mentioned in the previous announcements, I can only test the mysql, pgsql,
sqlite, and sqlite3 drivers. They work on FreeBSD and Debian 3.1 (I'll add more
test platforms shortly). The other drivers may or may not work at this time.

Get the prereleases right here:

 http://libdbi.sourceforge.net/downloads/libdbi-0.8.0-pre3.tar.gz
 http://libdbi.sourceforge.net/downloads/libdbi-drivers-0.8.0-pre3.tar.gz

Please report build and runtime errors. I'd like to release the new versions as
soon as possible. Christian, are you listening?

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de

[libdbi-devel] Freshmeat announcement problem

[Markus H. <mar...@mh...>]

Hi,

if we get a new release out the door, we should announce it on
Freshmeat. The libdbi-drivers announcement will be easy enough as I've
created the account. AFAIK David maintained the libdbi account. Do you
hear me David? If not, does anyone know the procedure to hijack
existing FM accounts?

regards,
Markus

-- 
Markus Hoenicka
mar...@ca...
(Spam-protected email: replace the quadrupeds with "mhoenicka")
http://www.mhoenicka.de