check this example by Tank:

http://www.customvirtualdesigns.com/modules/AMS/article.php?storyid=8

works great trabis, thank you!

Start from the first error and look in the path for the file which failed to open. If the file isn't there try uploading the missing file to the path mentioned.

Does this file exist on your v-host:
/hermes/web01/b1364/pow.myphilippines/htdocs/bacolodjobs/include/common.php

I recently had this issue with my wamp install... It was Windows Defender which decided after 2 years to look for and change localhost directives in my hosts file.

Its an easy fix open the following file:
C:\Windows\System32\drivers\etc\hosts

and make sure you have at least this directive for localhost:

This is not a recommended setting but you can turn this on if you have control over your php.ini

Try this is a custom php block to test:


edit: reverse the echos:

If you are using as a custom php block the php open/close tags are not needed.

There are arguments for/against either approach. My implementation I suspect is very rare.

I would think it depends on the complexity of the modules involved. The core only approach may be useful for modules that bridge other applications (WordPress, MediaWiki, etc). Also would be useful for mature modules which are no longer in development. If the core hook could encompass the system tables (users, configuration, etc) this would open many possibilities.

On the other hand I am sure there are modules with complex data structures, non-database related resources and cascading read/write functionality which may make a module hook a better choice.

Future developments in the core and legacy modules make it tough to say yes or no either way. I encourage other developers to add their 2 cents. For me this is an exciting topic :)

The hook class in general would save me a ton of work. @zyspec: Catz's suggestions would exceed my current implementation processes and allow a more generic approach to data read/write operations.

Currently I use XOOPS as an extension or library for other applications created for our organization.

As an example lets assume some event in our internal application requires a news article be posted to our XOOPS website via a specific trigger like a form submission. Within our form handler method I include functionality to write data to the news module tables incorporating data captured in our form submission to populate the news tables where applicable. Inversely our end-users may be submitting items to modules within our XOOPS site that we use within our internal app. The required functionality for XOOPS module data is included in our data access methods or into our applications data model.

With the hook class I would likely convert our internal applications to modules and use XOOPS more like a framework then a library :) I like the services concept as well, in a way that probably best describes what I do now.

As kaotik mentions it would be nice and probably necessary to extend hooking to resources other then database tables. There are many possibilities for module plugins and application bridging using hooks.

Nice work Kaotik! I have been doing this retrieving data directly from other modules tables using my own methods. Hooked access to the modules functions is much cleaner. A future core implementation perhaps?