[railroad-dev] Some help wanted]

[Julian Davison]

JR wrote:
> Hi,
> 
> I will have to build a new website and we're looking for the opportunity 
> to use Plone.
> Railroad seems to be very useful cause we have to put many large files...

Doesn't it? :)

> The main problem is that I didn't manage to install RailRoad.
> I feel like a newbie cause I didn't understand that the server side must 
> be installed.
> Maybe I am the only one to think that but I feel there is not too much 
> documentation.

There doesn't seem to be. I found a couple of key things
missing. Particularly with regard to the way RailRoad will
call back to your CMS (in our case, Plone) to determine
whether someone has access to an item or not.


[snip]
> This is the story of a non easy installation of RR.
> I hope it is the end of my trouble.

:)
Is it all still going well?

> But from now on I don't know how to check if it is running well...

What do you mean by 'well'? That it's running at full speed?



I've had a couple of problems with meta-data not storing properly
(any time you viewed a PloneRailroadProxy object all the metadata
got unset) and an associated flood of DB traffic (postgres is on
another machine accessed via 100Mb TCP LAN); a quirk with viewing
Proxy objects with an empty resource_url in a Plone 'folder contents'
listing and trouble uploading resources from IE (which hands over the
full path of the file you pick).

I've now got a subclass of PloneRailroadProxy with which we can
add images to our repository and the proxy object makes itself
a few (smaller) copies of the image for folder listings/previews.
Subjecting these objects to Plone's workflow also raised some
quirks with the metadata not being unmapped correctly, though
I think this was a direct result of a change I made when trying
to solve the metadata issue from above.

I presume the subversion repository is as up-to-date as it gets?




J,



J,



J,