MS RFC 8: Pluggable External Feature Layer Providers

[Jani Averbach]

> -----Original Message-----
> From: Daniel Morissette [mailto:dmorissette at dmsolutions.ca]
> Subject: Re: MS RFC 8: Pluggable External Feature
> Layer Providers
> 
> Jani Averbach wrote:
> > 
> > I would like to ask MS TSC to vote for this proposal. I am sorry
> > that this came so late, but if this passes, it will be ready and
> > implemented for coming 4.8 before feature freeze.
> > 
> 
> You realize that Oct 31st is only 4 days away?

Yes.  We have used the external plugins internally now about two
months, so the code is there and it's working.  I am also prepared to
work with this during weekend so it will be ready for Monday.

The general status of MS RFC 3 (bug 1477) is following:

- I have done all refactoring and implementing, and synced with CVS
  head at this morning.

- I have run various hand tests on Win32, and msautotest +
  mapscript/python tests on Linux(amd64)

- I have verified that php-mapscript compiles, but I would greatly
  appreciate if someone could help/test this after I have done my main
  commit. Same thing applies for perl,ruby,java, csharp and tcl.

- At the moment I am doing final polishing and testing and I will
  commit ms-rfc-3 this afternoon/evening. 

> How about a more automated mechanism where we set a MS_PLUGIN_DIR
> variable in the environment prior to calling MapServer or via a
> CONFIG directive at the map level, then the PLUGINLIB becomes a
> PLUGIN keyword whose value is the basename of the plugin file?

Otherwise than using environment, I like your idea. I will amend the
MS RFC 8 based on this and other ideas which came up on the list.

Best Regards, 
Jani