The open proxy has been implemented a lot here in the Netherlands back in
the days when "the internet was a safe place".

I recently notified the security officer of a large governemental body of
this issue as I think handing out wildcards to use your servers as a proxy
is a no-go. He is looking into the problem.

In my opinion, a better solution would be CORS (
http://en.wikipedia.org/wiki/Cross-origin_resource_sharing) but this would
mean that the (open) dataproviders will need some sort of policy to allow
applications access. I do however think thats the least-worse way to go, or
the data providers would have to set up a key validation mechanism.

Anyway, knowing who are consuming your services is not a bad thing. Neither
is telling a service provider that you are going to consume their service.


2013/12/17 Phil Scadden <[email protected]>

>
>  Anyway the decision is not mine, I have to provide a working web client
>> bringing the functionality of modify features, so I need a WFS-t in the
>> backend. Moreover I need to provide a wfs-t accessible without a proxy for
>> architecture reasons since the web browser and the web server will be
>> embebbed in a C++ standalone application and we cannot modify the web
>> server to host  a proxy.
>>
>
> The need for a proxy is dictated by security features of the browser. Most
> browsers will not do cross-domain XHR for very very good reasons. Its a
> pretty strange server that cant host a proxy. What is the servlet container
> that hosts your the WFS server?(and why cant the proxy live in that?). If
> the only WFS features you are dealing are hosted on embedded server, then
> you may be able to configure so request is not cross-domain. Might also
> depend on what exactly this embedded web browser actually is. The
> alternative would be protocol.script but you will be writing a lot of code
> to work around its limitations. Effective web apps work to principle of
> having the server do most of the computing so I am curious as to how
> serverside works for your application.  I would have to say that using
> embedded web-server and browser + OL for a standalone C# mapping
> application is a pretty strange choice. OL has to work around web
> limitations that simply dont exist to a standalone C# application.
>
> Notice: This email and any attachments are confidential.
> If received in error please destroy and immediately notify us.
> Do not copy or disclose the contents.
>
> _______________________________________________
> Users mailing list
> [email protected]
> http://lists.osgeo.org/mailman/listinfo/openlayers-users
>
_______________________________________________
Users mailing list
[email protected]
http://lists.osgeo.org/mailman/listinfo/openlayers-users

Reply via email to