squirrel-sql-develop Mailing List for SQuirreL SQL Client (Page 17)

On Mon, Aug 25, 2008 at 10:36 AM, Johnathan James <ha...@gm...> wrote:

> Hello all..
>
> I have been trying to work with the source this weekend to fix a bug that
> has been bugging me. Basically, if you open a new Session from the Aliases
> window, then CTRL-N to create a new window, if you press ALT-ForwardArrow,
> usually the focus will remain in the second window, when it should have
> moved to the second. So if you type, it will affect the window just hidden,
> not the one you just made visible.
>
> Anyway, my question is where is the KeyListener for ALT-ForwardArrow? When
> I
> press that key, what code runs to process it?


All of the key accelerator mappings are found at the bottom of :

sql12/app/src/net/sourceforge/squirrel_sql/client/resources/squirrel.properties

These are translated to KeyStrokes and assigned to the appropriate action in
:

sql12/fw/src/net/sourceforge/squirrel_sql/fw/util/Resources.java

The accelerator you are referring to is defined by the mapping:

menuitem.net.sourceforge.squirrel_sql.client.session.action.NextSessionAction.accelerator=alt
RIGHT

So the code that gets executed is found in :

sql12/app/src/net/sourceforge/squirrel_sql/client/session/action/NextSessionAction.java

Incidentally, I tried the series of events you described and the cursor is
consistently correctly moved
to the window that has focus.  I wonder if this is related to platform /
java version.  I'm using Sun Java
1.6.0_05-ea-b05 on Ubuntu Linux.  How about you ?

Rob

Hello all..

I have been trying to work with the source this weekend to fix a bug that
has been bugging me. Basically, if you open a new Session from the Aliases
window, then CTRL-N to create a new window, if you press ALT-ForwardArrow,
usually the focus will remain in the second window, when it should have
moved to the second. So if you type, it will affect the window just hidden,
not the one you just made visible.

Anyway, my question is where is the KeyListener for ALT-ForwardArrow? When I
press that key, what code runs to process it?

Thanks..

On Wed, Aug 13, 2008 at 1:34 PM, <Ric...@ge...> wrote:

> Has anyone tried or been successful in running SQuirreL via Java Web
> Start? I am looking into possibly using SQuirreL as a client that can be
> executed via Java Web Start, be resticted to connecting to specific
> databases, and require a login to use. The Java Web Start isn't strickly
> necessary, but it would be a benefit and a selling point to my management.
>
> I managed to revive the old ant task that signs all of the jars.  I also
updated the
jnlp file that was vastly out-of-date.  Treat this as experimental for now.
I was
able to launch from the URL:

http://www.squirrelsql.org/jws/squirrel.jnlp

Rob

Has anyone tried or been successful in running SQuirreL via Java Web 
Start? I am looking into possibly using SQuirreL as a client that can be 
executed via Java Web Start, be resticted to connecting to specific 
databases, and require a login to use. The Java Web Start isn't strickly 
necessary, but it would be a benefit and a selling point to my management.

Thanks,

Richard Allen

Meier, Stefan wrote:
>>> Here is a different approach that we use not for security reasons
> but 
>>> merely for logging:
>>>
>>> Create a wrapper JDBC driver that wraps your original driver. Then 
>>> register this driver instead of the original driver within SQuirreL.
> 
>>> Any call that SQuirreL sends to the database will then pass through 
>>> the wrapper. Here you'd be in control of everything. If you don't
> want 
>>> a call to execute simply make the wrapper throw an SQLException. 
>>> SQuirreL should handle such exceptions the way it does any other
> database exceptions.
>>> To create such a driver wrapper can be simplified a lot using JDKs 
>>> Proxy API. See 
>>> http://java.sun.com/j2se/1.5.0/docs/api/java/lang/reflect/Proxy.html
>>>
>> This has the benefit that it is portable to other JDBC front-ends and
> you can use the current version of SQuirreL unmodified.  Nice!
> 
> But then you must also prevent the user from changing the driver
> configuration in SQuirreL to using the original JDBC driver again or
> creating a new alias with this driver. They could also just use a
> different SQL-Client that you can download and use without having to
> install anythin as Administrator (e.g. SQuirrel SQL-Client). It seems to
> me that the "security" features suggested so far all rely on the end
> users being not very tech-savvy so they don't know how to deal with JDBC
> drivers and stuff.
> I think it would be much more secure to use some kind of middleware and
> not rely on restrictions built or configured into the client. Perhaps a
> SQL-proxy would be sufficient to achieve this (if there exists one for
> your RDBMS).


Good point. But perhaps one could equip the wrapper with some log in information 
which one would lack that is using the plain driver. Simplest thing could be URL 
and/or password hard coded in the wrapper. A more sophisticated and surely more 
secure way could be to let the wrapper work with the kind of middleware you suggest.

If an SQL-Proxy exists for the RDBMS in question than sure all security issues 
could be handled there. On the Client/SQuirreL side I'm pretty sure the 
SQL-Proxy would just behave like a wrapper.

Gerd

> > Here is a different approach that we use not for security reasons
but 
> > merely for logging:
> >
> > Create a wrapper JDBC driver that wraps your original driver. Then 
> > register this driver instead of the original driver within SQuirreL.

> > Any call that SQuirreL sends to the database will then pass through 
> > the wrapper. Here you'd be in control of everything. If you don't
want 
> > a call to execute simply make the wrapper throw an SQLException. 
> > SQuirreL should handle such exceptions the way it does any other
database exceptions.
> >
> > To create such a driver wrapper can be simplified a lot using JDKs 
> > Proxy API. See 
> > http://java.sun.com/j2se/1.5.0/docs/api/java/lang/reflect/Proxy.html
> >
> 
> This has the benefit that it is portable to other JDBC front-ends and
you can use the current version of SQuirreL unmodified.  Nice!

But then you must also prevent the user from changing the driver
configuration in SQuirreL to using the original JDBC driver again or
creating a new alias with this driver. They could also just use a
different SQL-Client that you can download and use without having to
install anythin as Administrator (e.g. SQuirrel SQL-Client). It seems to
me that the "security" features suggested so far all rely on the end
users being not very tech-savvy so they don't know how to deal with JDBC
drivers and stuff.
I think it would be much more secure to use some kind of middleware and
not rely on restrictions built or configured into the client. Perhaps a
SQL-proxy would be sufficient to achieve this (if there exists one for
your RDBMS).

Stefan

On Wed, Jul 23, 2008 at 5:37 AM, Gerd Wagner <ger...@t-...>
wrote:

> Here is a different approach that we use not for security reasons but
> merely for logging:
>
> Create a wrapper JDBC driver that wraps your original driver. Then register
> this driver instead of the original driver within SQuirreL. Any call that
> SQuirreL sends to the database will then pass through the wrapper. Here
> you'd be in control of everything. If you don't want a call to execute
> simply make the wrapper throw an SQLException. SQuirreL should handle such
> exceptions the way it does any other database exceptions.
>
> To create such a driver wrapper can be simplified a lot using JDKs Proxy
> API. See
> http://java.sun.com/j2se/1.5.0/docs/api/java/lang/reflect/Proxy.html
>

This has the benefit that it is portable to other JDBC front-ends and you
can use the current version of SQuirreL unmodified.  Nice!

Rob

Here is a different approach that we use not for security reasons but merely for 
logging:

Create a wrapper JDBC driver that wraps your original driver. Then register this 
driver instead of the original driver within SQuirreL. Any call that SQuirreL 
sends to the database will then pass through the wrapper. Here you'd be in 
control of everything. If you don't want a call to execute simply make the 
wrapper throw an SQLException. SQuirreL should handle such exceptions the way it 
does any other database exceptions.

To create such a driver wrapper can be simplified a lot using JDKs Proxy API. 
See http://java.sun.com/j2se/1.5.0/docs/api/java/lang/reflect/Proxy.html

Hope this helps.

Gerd

Robert Manning wrote:
> On Mon, Jul 14, 2008 at 6:32 AM, <Gui...@lo...> wrote:
> 
>> Hi,
>>
>> I am looking for a SQL tool we could give to our developpers in order to
>> access safely our production databases.
>> Because our databases contain confidential data, we would like to trace or
>> restrict the actions the developers could perform. This is a requirement
>> of our Security Dept (I am working in a bank, security is strict). The
>> tool would be installed on dedicated workstations with restricted
>> permissions (no administrative rights).
>>
>> One possibility is to use Squirrel, and develop our plugin to meet our
>> requirements.
>> Basicaly, the plugin should :
>> - prevent the user to export data by disabling some menu actions
>> - trace in a log-file all sql queries that are executed (not only those
>> sent by clicking on the "Run SQL" button, but also when the user edits the
>> results returned in the "Results" tab)
>>
>>
>> Can you confirm me that it is possible to develop such a plugin
>> (technically speaking) ?
>> I have no doubt that the first point (disabling a menu action) could be
>> easily developed. But about the second point (trace in a logfile), does
>> IPlugin have a callback method that I could implement and that would be
>> called before any SQL command is sent through jdbc ?
> 
> 
> Our current table editing component (DataSetUpdateableTableModelImpl)
> doesn't allow
> you to register a listener for it's SQL executing actions.  We would need to
> add support for
> that.  However, for the SQL tab, your plugin can register a
> ISQLExecutionListener  and
> get a callback prior to each statement to support auditing.  Of course,
> plugins can be
> disabled using the plugin summary dialog - if that's a show stopper, that
> menu item could
> be disabled as well.
> 
> Rob
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> Squirrel-sql-develop mailing list
> Squ...@li...
> https://lists.sourceforge.net/lists/listinfo/squirrel-sql-develop
> 

On Mon, Jul 14, 2008 at 6:32 AM, <Gui...@lo...> wrote:

> Hi,
>
> I am looking for a SQL tool we could give to our developpers in order to
> access safely our production databases.
> Because our databases contain confidential data, we would like to trace or
> restrict the actions the developers could perform. This is a requirement
> of our Security Dept (I am working in a bank, security is strict). The
> tool would be installed on dedicated workstations with restricted
> permissions (no administrative rights).
>
> One possibility is to use Squirrel, and develop our plugin to meet our
> requirements.
> Basicaly, the plugin should :
> - prevent the user to export data by disabling some menu actions
> - trace in a log-file all sql queries that are executed (not only those
> sent by clicking on the "Run SQL" button, but also when the user edits the
> results returned in the "Results" tab)
>
>
> Can you confirm me that it is possible to develop such a plugin
> (technically speaking) ?
> I have no doubt that the first point (disabling a menu action) could be
> easily developed. But about the second point (trace in a logfile), does
> IPlugin have a callback method that I could implement and that would be
> called before any SQL command is sent through jdbc ?


Our current table editing component (DataSetUpdateableTableModelImpl)
doesn't allow
you to register a listener for it's SQL executing actions.  We would need to
add support for
that.  However, for the SQL tab, your plugin can register a
ISQLExecutionListener  and
get a callback prior to each statement to support auditing.  Of course,
plugins can be
disabled using the plugin summary dialog - if that's a show stopper, that
menu item could
be disabled as well.

Rob

Hi, 

I am looking for a SQL tool we could give to our developpers in order to 
access safely our production databases.
Because our databases contain confidential data, we would like to trace or 
restrict the actions the developers could perform. This is a requirement 
of our Security Dept (I am working in a bank, security is strict). The 
tool would be installed on dedicated workstations with restricted 
permissions (no administrative rights).

One possibility is to use Squirrel, and develop our plugin to meet our 
requirements.
Basicaly, the plugin should :
- prevent the user to export data by disabling some menu actions
- trace in a log-file all sql queries that are executed (not only those 
sent by clicking on the "Run SQL" button, but also when the user edits the 
results returned in the "Results" tab)


Can you confirm me that it is possible to develop such a plugin 
(technically speaking) ?
I have no doubt that the first point (disabling a menu action) could be 
easily developed. But about the second point (trace in a logfile), does 
IPlugin have a callback method that I could implement and that would be 
called before any SQL command is sent through jdbc ?



Thanks for your answer
Guillaume



************************ DISCLAIMER ************************
This message is intended only for use by the person
to whom it is addressed. It may contain information
that is privileged and confidential. Its content does
not constitute a formal commitment by Lombard
Odier Darier Hentsch Group and any of its affiliates.
If you are not the intended recipient of this message,
kindly notify the sender immediately and destroy this
message. Thank You.
*****************************************************************

Hi!

Before I send you the code, I would like to add the user information during saving the temp pictures:

> >   I tried to set the status line: setStatusMessage(...), but the screen
> is not changing - the clock does not work, too.
> >   The screen is changed after exporting the images.
> >   How can I do a user information, how can i show an hourglass?


Last night I tried to remove the compatibilty problems to old files. I think I succeded...

Is it possible to check the compatitbility with some 'old' xml files of other databases - or isn't it possible to read them without database connection (of course update table, and other menu items  cant work...)
Some pictures to check the result would be neccessary, too.

Ciao, Karl.
-------- Original-Nachricht --------
> Datum: Mon, 09 Jun 2008 21:39:20 +0200
> Von: Gerd Wagner <ger...@t-...>
> An: Karl Gorton <Kar...@gm...>
> CC: squ...@li...
> Betreff: Re: [Squirrel-sql-develop] graph-plugin: user information during long	processings

> Hi Karl,
> 
> I hope you don't mind me being lazy just answering in English.
> 
> What you describe sounds grate. Thanks a lot for your efforts to improve
> SQuirreL.
> 
> If you would send me your code I'll have a look  at it. If the code is
> alright 
> and if you agree I'd check it in to our CVS repository.
> 
> Gerd
> 
> Karl Gorton wrote:
> > Grüß Gott aus Kärnten.
> > 
> > (english following)
> > 
> > Ich habe ein bisschen am graph plugin geändert.
> > 
> > Layout der Tabellen
> > Views können dargestellt werden
> > Notizen können ergänzt werden
> > Die Bilder bei der Ausgabe werden einzeln in einem temp-Verzeichnis
> zwischengespeichert.
> > Ein Spalte kann mehrere 'imported from'-Einträge haben
> >    ist zwar nicht schön, aber ich habe so eine Db zu warten...
> >    Deshalb bin ich derzeit auch zu alten xml-Dateien nicht kompatibel -
> aber den Import der alten mache ich bald.
> > 
> > Jetzt meine Frage:
> >   Die Zwischenpeicherung der Bilder kann dauern (beim Test für kapp
> über 2700 rund 5min)
> >   Ich habe versucht in die Statuszeile mit setStatusBarMessage(oder so
> ähnlich) etwas auszugeben,
> >      aber der Bildschirm ändert sich erst am Ende (es steht auch die
> Uhr).
> >   Wie mache ich das richtig?
> > 
> > Wenn alles klappt, wem soll ich dann meinen Code schicken - Ich habe
> doch gößere Änderungen vorgenommen - dafür habe ich einiges an Kommentar
> eingebaut - auch in 'fremden' Funktionen.
> > 
> > Mit freundlichen Grüßen, Karl Gorton.
> > -- english -------------------------------------------------------------
> > I tried to improve the graph plugin:
> >   Layout of tables
> >   Views can be displayed
> >   Notices
> >   Export of images - save images in temp path to save memeory
> >   More than one 'imported from' entries are possible
> >     bad design, but possible - and neccessary for me
> > 	currently i have problems with the compatibiltiy to old xml-files, but
> i hope i can correct this
> > 
> > My problem is:
> >   Saving of images needs much time: about 5 minutes for more then 2700
> images
> >   I tried to set the status line: setStatusMessage(...), but the screen
> is not changing - the clock does not work, too.
> >   The screen is changed after exporting the images.
> >   How can I do a user information, how can i show an hourglass?
> >   
> > If everthing works fine, to whom should i send my code?
> > I did large changes, but I have added some comments (to some unchanged
> funtions, too).
> > 
> > Bye, Karl Gorton.

-- 
-----------------------------
Karl Gorton
Kar...@gm...

GMX startet ShortView.de. Hier findest Du Leute mit Deinen Interessen!
Jetzt dabei sein: http://www.shortview.de/?mc=sv_ext_mf@gmx

Hi Karl,

I hope you don't mind me being lazy just answering in English.

What you describe sounds grate. Thanks a lot for your efforts to improve SQuirreL.

If you would send me your code I'll have a look  at it. If the code is alright 
and if you agree I'd check it in to our CVS repository.

Gerd

Karl Gorton wrote:
> Grüß Gott aus Kärnten.
> 
> (english following)
> 
> Ich habe ein bisschen am graph plugin geändert.
> 
> Layout der Tabellen
> Views können dargestellt werden
> Notizen können ergänzt werden
> Die Bilder bei der Ausgabe werden einzeln in einem temp-Verzeichnis zwischengespeichert.
> Ein Spalte kann mehrere 'imported from'-Einträge haben
>    ist zwar nicht schön, aber ich habe so eine Db zu warten...
>    Deshalb bin ich derzeit auch zu alten xml-Dateien nicht kompatibel - aber den Import der alten mache ich bald.
> 
> Jetzt meine Frage:
>   Die Zwischenpeicherung der Bilder kann dauern (beim Test für kapp über 2700 rund 5min)
>   Ich habe versucht in die Statuszeile mit setStatusBarMessage(oder so ähnlich) etwas auszugeben,
>      aber der Bildschirm ändert sich erst am Ende (es steht auch die Uhr).
>   Wie mache ich das richtig?
> 
> Wenn alles klappt, wem soll ich dann meinen Code schicken - Ich habe doch gößere Änderungen vorgenommen - dafür habe ich einiges an Kommentar eingebaut - auch in 'fremden' Funktionen.
> 
> Mit freundlichen Grüßen, Karl Gorton.
> -- english -------------------------------------------------------------
> I tried to improve the graph plugin:
>   Layout of tables
>   Views can be displayed
>   Notices
>   Export of images - save images in temp path to save memeory
>   More than one 'imported from' entries are possible
>     bad design, but possible - and neccessary for me
> 	currently i have problems with the compatibiltiy to old xml-files, but i hope i can correct this
> 
> My problem is:
>   Saving of images needs much time: about 5 minutes for more then 2700 images
>   I tried to set the status line: setStatusMessage(...), but the screen is not changing - the clock does not work, too.
>   The screen is changed after exporting the images.
>   How can I do a user information, how can i show an hourglass?
>   
> If everthing works fine, to whom should i send my code?
> I did large changes, but I have added some comments (to some unchanged funtions, too).
> 
> Bye, Karl Gorton.

Grüß Gott aus Kärnten.

(english following)

Ich habe ein bisschen am graph plugin geändert.

Layout der Tabellen
Views können dargestellt werden
Notizen können ergänzt werden
Die Bilder bei der Ausgabe werden einzeln in einem temp-Verzeichnis zwischengespeichert.
Ein Spalte kann mehrere 'imported from'-Einträge haben
   ist zwar nicht schön, aber ich habe so eine Db zu warten...
   Deshalb bin ich derzeit auch zu alten xml-Dateien nicht kompatibel - aber den Import der alten mache ich bald.

Jetzt meine Frage:
  Die Zwischenpeicherung der Bilder kann dauern (beim Test für kapp über 2700 rund 5min)
  Ich habe versucht in die Statuszeile mit setStatusBarMessage(oder so ähnlich) etwas auszugeben,
     aber der Bildschirm ändert sich erst am Ende (es steht auch die Uhr).
  Wie mache ich das richtig?

Wenn alles klappt, wem soll ich dann meinen Code schicken - Ich habe doch gößere Änderungen vorgenommen - dafür habe ich einiges an Kommentar eingebaut - auch in 'fremden' Funktionen.

Mit freundlichen Grüßen, Karl Gorton.
-- english -------------------------------------------------------------
I tried to improve the graph plugin:
  Layout of tables
  Views can be displayed
  Notices
  Export of images - save images in temp path to save memeory
  More than one 'imported from' entries are possible
    bad design, but possible - and neccessary for me
	currently i have problems with the compatibiltiy to old xml-files, but i hope i can correct this

My problem is:
  Saving of images needs much time: about 5 minutes for more then 2700 images
  I tried to set the status line: setStatusMessage(...), but the screen is not changing - the clock does not work, too.
  The screen is changed after exporting the images.
  How can I do a user information, how can i show an hourglass?
  
If everthing works fine, to whom should i send my code?
I did large changes, but I have added some comments (to some unchanged funtions, too).

Bye, Karl Gorton.
-- 
-----------------------------
Karl Gorton
Kar...@gm...

Der GMX SmartSurfer hilft bis zu 70% Ihrer Onlinekosten zu sparen! 
Ideal für Modem und ISDN: http://www.gmx.net/de/go/smartsurfer

All,

Up until today, I have always configured my build system (for release builds
that go on SF) to use Java 1.5 via JAVA_HOME.  I do this because we
guarantee that every new version of SQuirreL will run fine with java 1.5 and
I've been too paranoid to rely on the Ant source/target attributes which are
set to use 1.5.  However, while trying to use the FTP Ant task to automate
uploading update files to the new update site, I noticed that Apache is
shipping 1.7 jars that have been compiled with Java 6 (I get class version
errors if I build with 1.5 and try to use the FTP Ant task, for example -
building with 1.6 works great, though).  I tried downloading the source and
spent several hours attempting to build it from scratch with 1.5.  I gave up
on this for two reasons:

1. It wasn't building all of the optional tasks that we use now in the
build.
2. A custom Ant system built with 1.5 is not a very portable / reproducible
way to build our software.

So, I am switching over to useing java 1.6 for release builds.  This should
make absolutely no difference if our Ant build files are correctly
specifying source / target attributes for the javac Ant task.  That is, a
compile error will still result from using 1.6-only APIs, and users should
still be able to run with Java 1.5.  I noticed that at least one of the
plugin build files doesn't use the source/target attributes for the javac
Ant task.  I'll fix any build files that I spot with this defect.  If anyone
spots any issues that might be related to the switch to building releases
with java 1.6, just send me an email (manningr at users dot sourceforge dot
net).

Thank you for your attention, and have a nice day!

Rob

On Tue, Apr 29, 2008 at 11:06 PM, OPLink <as...@op...> wrote:
> Robert,
>
>  > The DB2 Plugin was developed against DB2 LUW (andsupport for OS/400
>
> > added with help from others).  The "IBM DB2 JDBC Universal Driver"
>  > that I use to connect to LUW doesn't allow you to specify a "System"
>  > property, so I am not able to see this behavior.  Would it be solved
>  > by simply detecting the system property you mentioned and then
>  > re-phrasing the SQL with "/" instead of "." in the table identifiers?
>  > Do table aliases work the same way in this case?
>  Yes it seems the DB2 Plugin will work if it could detect the naming
>  convention selected on the driver properties
>  and then switch to that notation, in this case from SQL to System or '.' to
>  '/'. I tried changing the plugin myself
>  I was able to get rid of the message but then nothing would display if
>  double click on the UDF folder where I am
>  able to display the functions in that folder if I use the SQL notation.
>  However it may be my end as I did not
>  take my time on it. I will attempt again this week. Aliases files should
>  work the exact same way with members or no members.

Ok, let me know if there is anything I can do to help.  If you have a patch
that works, I'd happy to apply it for you in CVS.

>  ----------------
>
>
>  > Again, my driver doesn't seem to support this property.  When you say
>  > doesn't work, I assume you mean that you execute some SQL in the
>  > editor and it fails when using column remarks instead of the column
>  > name - even though the property is set?  It appears as though that
>  > setting might be ignored by the driver?
>  The SQL does not fail it excutes fine, however, it will always display the
>  Column-Name(field name) and not the Remarks(full length text) regardless of
>  the properties setting for Remarks. The Remarks setting gives me a choice of
>  'sql' or 'system' I would think the Remarks data would be displayed for
>  heading if selected 'system'.
>  --------------
>  To re-phrase:
>  --------------
>  When I double click a file from the object tree I can display the right
>  pane. From there - in the 'column' tab you will find both the COLUMN_NAME
>  and  REMARKS. It would be great if SQL could display the REMARKS in place of
>  the COLUMN_NAME based on the driver properties setting for Remarks. But
>  maybe thats not the DB2 Plugin's responsibility ?

I would think that the driver would be responsible for making the
switch, given that
it's a driver property.  We use the metadata that we get from JDBC, so
I would think
it would be easy for someone to write a small test that shows that the
driver doesn't
do this and then perhaps submit it as a bug to the DB2 support people.
 We don't
have APIs for the plugin to make this switch in every place that
column names would
appear.

Rob

Hi Joe,

See my responses below.

On Mon, Apr 28, 2008 at 8:57 PM, oplink <as...@op...> wrote:
> Im not sure if im missing something or if the following is an issue. It seems the DB2 Plugin does not reconize a few
>
>  driver properties :
>
>
>
>  DB2 Plugin does not consider the option selection 'System" for the Driver Properties "naming". DB2 Plugin will always use the SQL dot notation for naming files regardless of the 'naming' selection. (this is not on statements entered in the SQL editor but rather the plugin itself)
>
>
>
>  Example:   in the UDFParentExpander.java the FROM file in the SELECT statement will always us the dot notation  ".SYSIBM.SYSFUNCTIONS..."  regardless of the naming convention selected in the Driver Properties. If I were to select the "System" as the naming convention I can use the standard i5/OS forward slash, "/" instead of the dot notation for SQL statements. In addition, using the 'System" properties will allow the searching of a library list to go into effect.
>
>
>
>  To reproduce this error:
>
>  Select the "IBM DB2 App Driver" then select "Properties", then "Driver properties" tab.
>
>  >From the "driver properties"  check off "Use driver properties'
>
>  Then select "System" for the "naming' properties.
>
>  Then select OK

The DB2 Plugin was developed against DB2 LUW (and support for OS/400
added with help from others).  The "IBM DB2 JDBC Universal Driver"
that I use to connect to LUW doesn't allow you to specify a "System"
property, so I am not able to see this behavior.  Would it be solved
by simply detecting the system property you mentioned and then
re-phrasing the SQL with "/" instead of "." in the table identifiers?
Do table aliases work the same way in this case?

>
>  There is also a nice feature in the Driver Properties called "Remarks". This will allow the use of the remarks as a column name instead of the field name. This remark information is the actual expanded text info of the field name defined in DDS on the i5/OS.
>
>  This feature does not seem to work.

Again, my driver doesn't seem to support this property.  When you say
doesn't work, I assume you mean that you execute some SQL in the
editor and it fails when using column remarks instead of the column
name - even though the property is set?  It appears as though that
setting might be ignored by the driver?


Rob

Im not sure if im missing something or if the following is an issue. It seems the DB2 Plugin does not reconize a few 

driver properties :



DB2 Plugin does not consider the option selection 'System" for the Driver Properties "naming". DB2 Plugin will always use the SQL dot notation for naming files regardless of the 'naming' selection. (this is not on statements entered in the SQL editor but rather the plugin itself)

 

Example:   in the UDFParentExpander.java the FROM file in the SELECT statement will always us the dot notation  ".SYSIBM.SYSFUNCTIONS..."  regardless of the naming convention selected in the Driver Properties. If I were to select the "System" as the naming convention I can use the standard i5/OS forward slash, "/" instead of the dot notation for SQL statements. In addition, using the 'System" properties will allow the searching of a library list to go into effect. 

 

To reproduce this error:

Select the "IBM DB2 App Driver" then select "Properties", then "Driver properties" tab.

>From the "driver properties"  check off "Use driver properties'

Then select "System" for the "naming' properties. 

Then select OK 



After setting up the driver properties execute the driver and then  from the objects tree select a library folder then double click UDF. 

The following message should appear:

Error: UDF: java.sql.SQLException: [SQL5016] Qualified object name SYSFUNCS not valid.

-----

There is also a nice feature in the Driver Properties called "Remarks". This will allow the use of the remarks as a column name instead of the field name. This remark information is the actual expanded text info of the field name defined in DDS on the i5/OS. 

This feature does not seem to work.



thanks

joe mcdonald

Just to clarify, it would be good if some kind of progress bar (perhaps 
with a cancel option) could be displayed, to give some feedback. The 
performance isn't too bad actually :)

Neil Gordon wrote:
> Hi there,
>
> If you connect squirrel to a remote database over a low bandwidth 
> connection, it hangs when connecting initially for a long time 
> (obviously loading the schema), but then when it is connected the 
> performance is quite good.
>
> Just an observation , for possible improvement of a great product.
>
> Regards
>
> Neil
>
>

Hi there,

If you connect squirrel to a remote database over a low bandwidth 
connection, it hangs when connecting initially for a long time 
(obviously loading the schema), but then when it is connected the 
performance is quite good.

Just an observation , for possible improvement of a great product.

Regards

Neil

On Mon, Apr 7, 2008 at 6:56 AM, Robert Manning
<rob...@gm...> wrote:
> On Sun, Apr 6, 2008 at 5:12 PM, Dieter Engelhardt <Di...@ew...> wrote:
>  > Hi there,
>  >  i've tried to integrate IzPack in my build environment but when i
>  >  execute thze ant task: make_install_jar i get the following:

Dieter,

It took some work, but I restored the make_install_jar target in the
main build script (i.e. sql12/buid/build.xml)
Make sure you have the latest IzPack installed (3.11.0).  This target
doesn't create all of the release artifacts
that we make available on the downloads site (that build script is in
sql12/build/snapshot/build-weekly.xml)
However, it should provide you with the Windows/Unix installer jar
that you can use to test your plugin integration
with the installer.  Additionally, we usually make new snapshot builds
available on the download site each weekend
if there have been any notable changes during the week. Enjoy!

Rob

On Sun, Apr 6, 2008 at 5:12 PM, Dieter Engelhardt <Di...@ew...> wrote:
> Hi there,
>  i've tried to integrate IzPack in my build environment but when i
>  execute thze ant task: make_install_jar i get the following:
>  LzPack:
>      [java] .::  IzPack - Version 3.11.0 ::.
>      [java] < compiler specifications version: 1.0 >
>      [java] - Copyright (c) 2001-2008 Julien Ponge
>      [java] - Visit http://izpack.org/ for the latest releases
>      [java] - Released under the terms of the Apache Software License
>  version 2.0.
>      [java] -> Fatal error :
>      [java]    bad argument
>      [java] java.lang.Exception: bad argument
>      [java]     at
>  com.izforge.izpack.compiler.CompilerConfig.main(CompilerConfig.java:2208)
>      [java]     at
>  com.izforge.izpack.compiler.Compiler.main(Compiler.java:709)
>      [java] (tip : use -? to get the commmand line parameters)
>
>  The command line help tells me, that i should use -h instead of -HOME
>  but this doesn't work either.

Check again, the "-h" switch should work fine.  In case you are
wondering, we cut releases from the snapshot directory
(build-weekly.xml) under build.  I've had a desire to integrate the
two build files (build.xml and build-weekly.xml) for some time.  We
should probably replace the make_install_jar target from the main
build file when the two are integrated.  For now I'll fix build.xml so
that it uses "-h" instead of "-HOME" (which only works for IzPack
versions prior to 3.7)

>  I'm just want to make shure, that my new plugin is working in a new
>  distribution of squirrel.

You should be able to download the last snapshot build installer from
the downloads page and test it that way for now.

Rob

Hi there,
i've tried to integrate IzPack in my build environment but when i 
execute thze ant task: make_install_jar i get the following:
LzPack:
     [java] .::  IzPack - Version 3.11.0 ::.
     [java] < compiler specifications version: 1.0 >
     [java] - Copyright (c) 2001-2008 Julien Ponge
     [java] - Visit http://izpack.org/ for the latest releases
     [java] - Released under the terms of the Apache Software License 
version 2.0.
     [java] -> Fatal error :
     [java]    bad argument
     [java] java.lang.Exception: bad argument
     [java]     at 
com.izforge.izpack.compiler.CompilerConfig.main(CompilerConfig.java:2208)
     [java]     at 
com.izforge.izpack.compiler.Compiler.main(Compiler.java:709)
     [java] (tip : use -? to get the commmand line parameters)

The command line help tells me, that i should use -h instead of -HOME 
but this doesn't work either.
Here is  my test ant task:
    <target name="LzPack" depends="init">
        <java jar="${izpack.home.dir}/lib/compiler.jar"
              fork="true"
              failonerror="true"
              maxmemory="128m"
        >
            <arg value="-HOME"/>
            <arg value="${izpack.home.dir}"/>
            
            <arg value="${basedir}/build/${izpack.config.file}"/>
            <arg value="-b"/>
            <arg value="${basedir}/build"/>
            <arg value="-k"/>
            <arg value="standard-kunststoff"/>
            <arg value="-o"/>
            <arg 
value="${basedir}/../squirrel-sql-dist/${name}-${version}-install.jar"/>

            <classpath>
                <pathelement 
location="${izpack.home.dir}/lib/compiler.jar"/>
                <pathelement path="${java.class.path}"/>
            </classpath>
        </java>        
    </target>        

Does anybody of you guys have a hint for me whats wrong?

I'm just want to make shure, that my new plugin is working in a new 
distribution of squirrel.

Thanx for your help,
Dieter.

Hi Dieter,

Dieter Engelhardt wrote:
> Hi all,
> i developed a small new plugin which replaces a set of variables in the 
> sql statements before they get executed.
> The background for this is, that i often work in an unix ksh 
> environment, where typicaly  eg. databases are specified as a variable. 
> The plugin just replaces the specified vars before the statements are 
> send to the driver, but leaves the source untouched.
> It contains a pereferences panel, where one can specify a bunch of vars 
> and its replacements and if any of the specifed var are found in the 
> soure, they will be replaced.
> 
> My question now is, should that go into the main project or should i 
> just host it as a external plugin on my own site?
Could you send me the sources of your Plugin. I'll have a look at it and if 
there aren't any questions I'll commit it to CVS and grant you developer rights. 
Would that be alright for you?

There are already two Plugins that address parameterization of SQLs:
1. The Bookmarks Plugin.
2. The SQLParam Plugin.
It would be nice if you could have a look at both (online help should be of 
help) and could give me a short description of the differences in your Plugin.

Thanks

Gerd

Hi all,
i developed a small new plugin which replaces a set of variables in the 
sql statements before they get executed.
The background for this is, that i often work in an unix ksh 
environment, where typicaly  eg. databases are specified as a variable. 
The plugin just replaces the specified vars before the statements are 
send to the driver, but leaves the source untouched.
It contains a pereferences panel, where one can specify a bunch of vars 
and its replacements and if any of the specifed var are found in the 
soure, they will be replaced.

My question now is, should that go into the main project or should i 
just host it as a external plugin on my own site?

Looking forward to hear from you.
Regards
Dieter

This patch that provides support for DB2 XML types is in CVS now.

Cheers,

Rob

2008/2/29  <Lar...@ce...>:
> Hi Rob,
>
>  here is the patch file containing the changes i made.
>
>  Thanks for your help and have a nice day :-)