• Common authentication back-end
• SSL for login
• Improved synchronization from Piccolo

Common authentication back-end

From the time the code for the Ingres DBMS server was published in Subversion, the authentication and authorization mechanism was distinct from that of the rest of the Ingres Community. This meant that those with commit access would need to maintain separate accounts. Alex Trofast and Andrew Ross have converted the authentication system for both http://code.ingres.com and http://bugs.ingres.com to use the same system for the wiki and forums.

SSL for login

Previously only the Community Forum and Wiki had SSL (https) logins, now the bugs and code site do as well. For the time being you can access them both in plain-text mode but at some point secure logins will be required. The updated URLs are – https://code.ingres.com and https://bugs.ingres.com. Note that for the time being the sites are signed using a self-signed certificate. This will change in due course but for the time being you may be required to acknowledge that the certificate has not been validated against any of the root Certificate Authorities.

Improved synchronization from Piccolo

Probably the most important out of the 3 changes is the improved sync up between Ingres’s internal source code system, Piccolo and the external Subversion server. Changes are now replicated out every 15 minutes from Piccolo, assuming there is something to propagate. The process for merging changes from Subversion into Piccolo is still being worked on. So now we have up to date Ingres code in the wild, so to speak.

Some random posts

Related Posts

• Doing More with Even Less – UK IUA 2009 Conference
• UK IUA – 2009 Call for Papers
• UK Ingres User Association Spring Conference

In other news, me being slightly behind the times, Ice Breaker went GA (Generally Available) last week. The all in one database appliance is now available to download from ingres.com. See Google for more info/press/comment.

Some random posts

Some random posts

• Added environment check during initialization
• Fixed SEGV when closing a statement
• Fixed spurious E_WARNING with empty LOB dataset
• Fixed SEGV when ingres_close() is not called
• Fixed auto commit cannot be enabled on shutdown
• Fixed Array index start for php_ii_field_name
• Fixed constant typo

Now that is out of the way I will be looking at the next “major” release of the extension. In particular a change to support multiple result sets. Unfortunately this will require a breakage in backwards compatibility with the way things work at present. For example to run a query now the function ingres_query() is called as followed:
$rc = ingres_query ($sql, $link)or
$rc = ingres_query ($sql)

As you can tell the $link parameter is optional. This will no longer be the case. Also instead the function will not give a return code rather a result resource for ingres_fetch_*. So a simple (no error checking) connect, query, fetch, disconnect will go as follows:

$link = ingres_connect($dbname, $user, $password, $options);
$sql = "select * from iirelation";
$result = ingres_query($link, $sql);
while ( $iirelation = ingres_fetch_object($result) )
{
echo $iirelation->relid . "<br/>\n";
.
.
}
/* release the result set */
ingres_free ($result)
ingres_close ($link)

As you can see it is a bit different and is most likely to cause breakage in any PHP application that makes use of Ingres but it is a necessary change. If you have any comments or suggestions you can email via grantc@php.net

Some random posts

Some random posts

Some random posts