Latest topics for the forum "SQL Power Architect Suggestions"

Data model diff as SQL to recreate changes should by built in order of dependencies added

bronius — GMT

I ran into this issue today:
- SQL diff, see ALTER fields to add a fistful of new fields
- Execute SQL, watch some statements fail due to dependent fields not present
- EITHER Execute SQL again, ignoring dup fields, repeating for each field dependency now allowed
- OR manually sort the ALTER statements to make sure that missing fields are inserted first

Ex:
[code]ALTER TABLE sfstaging_account ADD COLUMN field_payopt_creditcode VARCHAR(255) AFTER field_home_addresszip;
ALTER TABLE sfstaging_build ADD COLUMN company_rep_email VARCHAR(255) AFTER field_company_subpartnerno;
ALTER TABLE sfstaging_build ADD COLUMN field_installation_county VARCHAR(255) AFTER field_installation_addresszip;
ALTER TABLE sfstaging_build ADD COLUMN paymentoption_id INT AFTER account_id;
ALTER TABLE sfstaging_lease ADD COLUMN field_payopt_prepaytime VARCHAR(255) AFTER field_payopt_srecowner;
ALTER TABLE sfstaging_lease ADD COLUMN field_payopt_srecowner VARCHAR(255) AFTER field_quoted_leaserate;
ALTER TABLE sfstaging_lease ADD COLUMN qe_api_taxamount DECIMAL(10,2) AFTER qe_api_taxrate;
ALTER TABLE sfstaging_lease ADD COLUMN qe_api_taxrate DECIMAL(10,2) AFTER field_payopt_prepaytime;[/code]

should be:
[code]ALTER TABLE sfstaging_account ADD COLUMN field_payopt_creditcode VARCHAR(255) AFTER field_home_addresszip;
ALTER TABLE sfstaging_build ADD COLUMN company_rep_email VARCHAR(255) AFTER field_company_subpartnerno;
ALTER TABLE sfstaging_build ADD COLUMN field_installation_county VARCHAR(255) AFTER field_installation_addresszip;
ALTER TABLE sfstaging_build ADD COLUMN paymentoption_id INT AFTER account_id;
ALTER TABLE sfstaging_lease ADD COLUMN field_payopt_srecowner VARCHAR(255) AFTER field_quoted_leaserate;
ALTER TABLE sfstaging_lease ADD COLUMN field_payopt_prepaytime VARCHAR(255) AFTER field_payopt_srecowner;
ALTER TABLE sfstaging_lease ADD COLUMN qe_api_taxrate DECIMAL(10,2) AFTER field_payopt_prepaytime;ALTER TABLE sfstaging_lease ADD COLUMN qe_api_taxamount DECIMAL(10,2) AFTER qe_api_taxrate;[/code]

Notice in the above example that the order of the first four or so ALTERs didn't matter, because the "AFTER fields" already exist, but the last four items had to be rearranged because the AFTER fields were to be inserted by this same SQL diff script.

Cheers
-Bronius

Video conferencing

Herieaty Khalid — GMT

There are lots of video conferencing services available today. If you have never heard about it before then read a review about these services. The most well known are the following

Gomeetnow(http://www.gomeetnow.com/)
GoToMeeting
WebEx
Skype (for 2 persons)

Attaching Objects to Relationships

Harlan — GMT

Hello,
One feature that almost no other Entity Relationship program provides, at least none that I've found, especially in the open source arena, is attaching objects to relationships.

When attaching an object to a relationship, such as amount (of type numeric(10,2)) would then create another table. As it does now, if there are no attached objects, then the relationships just become foreign keys.

This would help immensely to accurately describe databases. Right now, we have to hand-code this, which makes things less clear.

SQL Power Architect is definitely one of the best design products that I've seen in the open source arena.

Thank You for your time and attention!!!

Harlan...

Add ts_vector data type for Postgres

Jeeba — GMT

Im sure im not the only one that want to see the ts_vector datatype from postgres 8.4 in your excellent software. Anyway is there another way to put that datatype available?

Disable edge stickiness in PlayPen

OceanWind — GMT

I've been running into a difficulty with tables wanting to stick to the edges of the playpen. It's a real problem with large tables, since I can't unstick them — particularly from the bottom edge.

Is there a way to turn this feature off?

Salesforce and Force.com JDBC Database Connection

slivey — GMT

I would like to use Power Architect to connect to my Force.com (salesforce platform) database.
Currently, a JDBC driver exists for SchemaSpy. But getting it working is somewhat "clunky" and then you can only view the schema in SchemSpy.

I much prefer SQL Power Architect. And I think a database connection could be easily created with some minor adjustments to the JDBC SchemaSpy connector.

I was able to adjust an existing Database Connection (SQLStream generic connector) and tested the connection successfully... Here are the adjusted parameters:

Driver Class: com.claimvantage.force.jdbc.ForceMetaDataDriver
Connection String Template: jdbc:claimvantage:force::
(where sfuser is the salesforce username, sfpw is the salesforce password + security token)
Jar files from first three downloads here: http://code.google.com/p/force-metadata-jdbc-driver/downloads/list (depends.zip has jar files)
Also - all source code, wiki, etc for the salesforce jdbc may be viewed here.

So, Test Connection is successful with these parameters.

However, when I Add Source Connection, I get a Java null pointer exception (see below). But I think it would only take a few "tweaks" to adapt and get working. And this type of connection would draw a huge following to SQL Power Architect - because there is currently no decent way to create/view data models.

Error Report (null pointer exception):

java.lang.RuntimeException: java.lang.NullPointerException
at ca.sqlpower.architect.swingui.DBTree$PokeDBWorker.cleanup(DBTree.java:808)
at ca.sqlpower.swingui.SPSwingWorker$1.run(SPSwingWorker.java:114)
at java.awt.event.InvocationEvent.dispatch(InvocationEvent.java:209)
at java.awt.EventQueue.dispatchEvent(EventQueue.java:633)
at java.awt.EventDispatchThread.pumpOneEventForFilters(EventDispatchThread.java:296)
at java.awt.EventDispatchThread.pumpEventsForFilter(EventDispatchThread.java:211)
at java.awt.EventDispatchThread.pumpEventsForHierarchy(EventDispatchThread.java:201)
at java.awt.EventDispatchThread.pumpEvents(EventDispatchThread.java:196)
at java.awt.EventDispatchThread.pumpEvents(EventDispatchThread.java:188)
at java.awt.EventDispatchThread.run(EventDispatchThread.java:122)
Caused by: java.lang.NullPointerException
at ca.sqlpower.sqlobject.SQLCatalog.fetchCatalogs(SQLCatalog.java:65)
at ca.sqlpower.sqlobject.SQLDatabase.populateImpl(SQLDatabase.java:162)
at ca.sqlpower.sqlobject.SQLObject.populate(SQLObject.java:179)
at ca.sqlpower.sqlobject.SQLObject.getChildCount(SQLObject.java:297)
at ca.sqlpower.architect.swingui.DBTree$PokeDBWorker.pokeDatabase(DBTree.java:789)
at ca.sqlpower.architect.swingui.DBTree$PokeDBWorker.doStuff(DBTree.java:774)
at ca.sqlpower.swingui.SPSwingWorker.run(SPSwingWorker.java:104)
at java.lang.Thread.run(Thread.java:637)

Disable zooming on mouse scroll (x-axis)

dubuscyr — GMT

Hi everyone,

I know the topic had already been discussed but nothing was done by the dev.

I'm asking it again : you [b]really[/b] should allow the user to disable the zooming function on mouse scroll (x-axis). On a mac this is just unusable with the Apple's mice (mighty mouse, magic mouse, multi-touchpad). I'm looking for another software just because I can't scroll properly (without zooming or dezooming like crazy) with SQL Architect. This is just sad because the real job is done by the soft, and this single detail is [b]so[/b] annoying.

References :

Post : http://www.sqlpower.ca/forum/posts/list/1806.page#2343
Bug : http://trillian.sqlpower.ca/bugzilla/show_bug.cgi?id=1515

Thank you,

Cyril

Find/Replace Doesn't Search Physical Names

tonyc — GMT

Hi,

I recently re-factored a table in one of my designs, changing a column name that was referenced in a number of tables. I found that the SQL was being regenerated with the old column name. First I thought it was a caching problem, then a bug and eventually discovered that the search and replace only searches in logical names even when you select the 'anything' option.

Regards, Tony.

Move the application to Netbeans or Eclipse RCP or to SpringDesktop+Java Plugin Framework

EdwinF — GMT

Hi all and thanks for the tool.

The current user interface for the application is kind of hard to follow.
I'd sugest it's moved to a proven and efficient desktop platform such as Netbeans RCP or Eclipse RCP, if any of these two is too heavy, I'd recommend the SpringDesktop RCP along with the Java Plugin Framework to handle packages dependencies (although the Java Services API (SPI) could suffice).

Moving the application to such a platform will give a lot of good things to the application, such as a docking framework, properties management, data editors .... etc ...

Thank you.

Retain Line Breaks and Returns in Remark Fields

mtangorre — GMT

I would like to request that line breaks and hard returns be maintained in the remark boxes for tables and columns. I use these extensively, and they always lose the line breaks!

Preference to auto-copy logical name to physical name

irisheyes — GMT

In the User Preferences, it would be nice to have a checkmark to allow the physical and logical names to be the same as a default preference. When the logical name is entered and the user tabs to the next field (which is the physical name), the program would automatically suggest the logical name as the physical name.

Relationship highlighting

castorp — GMT

Hi,

I love the fact that when I click a relationship, the corresponding columns are highlighted in both tables.

I would like to see this the "other way round" as well.

When selecting a column inside a table, and that column is part of a FK, the relationship line should be highlighted.

Regards
Thomas

YAML support for symfony (Propel/Doctrine) Developers

DBA_Alex — GMT

I use Architect all the time, however there's one feature that I'm sure symfony([url]http://symfony-project.com[/url]) web developers would find useful: being able to export their database code into YAML format for either Propel or Doctrine.

Compare DM - DDL Generation could be optimized

castorp — GMT

Hi,

I noticed that the generated DDL when comparing data models (at least for Oracle and Postgres) is a bit "complicated".

When a table (or column) is dropped, the DDL first contains drop statements for any related constraint. While that is technically correct and is probably needed for e.g. SQL Server or MySQL, Oracle and Postgres offer a CONSTRAINT option when dropping objects.

The resulting DDL script would be a lot smaller if the tables are dropped right at the beginning using "DROP TABLE the_table CASCADE" (Postgres) or "DROP TABLE the_table CASCADE CONSTRAINTS" (for Oracle).

Then all ALTER TABLE ... DROP CONSTRAINT for that table could be removed from the script, making it a bit more concise.

The same is true for dropping columns (simply add the CASCADE keyword and "forget" about any constraint to that column)

Regards
Thomas

PostgreSQL Serial type

forumsareforsissyboysandgirliemen — GMT

Postgresql support a type called serial which is convenient for use as pk, it is automatically connected to a sequence, which is also automatically created.

When programming JPA/EJB3 the serial type comes in handy and is easy to set up and use with JPA.

My suggestion is to add support for Serial in Power Architect.