Re: [DotGNU]UI and SEE (was Re: What Web Services Are NOT)

[Norbert Bollow]

> Norbert Bollow wrote:
> > Regardless of where webservices are used, they will be used to
> > put together a system that does something useful.  The users of
> > this system deserve to get a decent user interface, hence the
> > plan to create CD-ROMs with a very portable Secure Execution
> > Environment platform that allows a program which implements this
> > user interface to be executed locally, on ther user's PC.
> 
> Speaking of user interface, I believe you are saying that
> there must be both a client ("implements UI") and a server,
> and that the RPC needed for webservices will be handled by the
> programs individually.

yes... I think the RPC should be part of the programs which
implement webservice components, not the deamon... of course
these "untrusted space" programs will make use of libraries
which are provided by the SEE, and which give access to the
machine's network interfaces in a controllable manner.

> Sorry, I originally thought the UI transference would be
> handled by SEE/plugin, much as exported X apps are handled by
> the X server.

no.. I think the SEE daemon will be involved only during the
transfer of the UI and until the UI is running in a process of
its own.

> So this means that:
> 
> * A user (who also has SEE installed locally on the PC) who
>    happens to be the "owner of the data" for a program that runs
>    on a remote webservice server, tells the SEE on his local
>    machine to connect to a SEE on a remote webservice server, and
>    download the data and the application program, so that it can
>    then be executed locally.
> 
> completely contains the point:
> 
> * A user (who has SEE installed locally on the PC) should
>    have the option of accessing a webservice application (which
>    runs on a remote webservice server) in such a way that if that
>    webservice application can transfer to the user's PC code
>    which implements a decent user interface.
> 
> because the SEE shouldn't be able to tell the difference,
> because it doesn't control the RPC of webservices.

no... the SEE will know the difference... in the first case, the
SEE will tell the server:

  "I have a user here who is the 'owner of the data' for your
  service XYZ.  Here are the credentials.  Don't execute service
  XYZ, but rather send me everything that is necessary to
  provide service XYZ, data and portable executables."

The response from the server could contain a GnuPG-encrypted
.tgz with everything that was asked for, and it's the local
SEE's responsibility to either give this to the user as-is,
or unpack it and start providing service XYZ, or both.


In the second case, the SEE will tell the server:

  "I have a user here who has the right to use service XYZ.
  Here are the credentials.  Please execute service XYZ on
  his behalf.  Please tell the service XYZ process that we
  can handle the following types of user interface, with
  this order of preference: pnet3.2, HTML4.01, HTML2.0 and
  that the user's language preferences are German, English,
  French."

The response from the server could be "I've checked those
credentials, they're fine.  Here's a German-language user
interface of type pnet3.2 - execute that, and your user should
be happy."


> Right? Because if so, I can just wipe out the whole
> SEE-development chapter on passing events (a little more
> Texinfo I'm writing, more details to come later).

Even if this isn't necessary for the SEE daemon, don't quite
wipe it out please.   An environment that allows you to
pass events around is a very useful thing.  I think it will
be very good to have such an environment available, as one
of the SEE plugins (which would in turn use other plugins).

Greetings, Norbert.

-- 
A founder of the http://DotGNU.org project and Steering Committee member
Norbert Bollow, Weidlistr.18, CH-8624 Gruet   (near Zurich, Switzerland)
Tel +41 1 972 20 59       Fax +41 1 972 20 69      http://thinkcoach.com
List hosting with GNU Mailman on your own domain name   http://cisto.com