Cool -- I like that the compliance tests themselves are embedded in a gadget
-- this makes it easy to test on lots of containers. In addition, simply
being able to run the complaince test gadget is a test in and of itself.

-Dan

On Thu, Feb 21, 2008 at 10:09 PM, Cassie <[EMAIL PROTECTED]> wrote:

> ha, and by that link (which ends mistakenly in .js) i really meant:
>
> http://opensocial-resources.googlecode.com/svn/tests/trunk/compliancetests.xml
>
> On Thu, Feb 21, 2008 at 9:53 PM, Cassie <[EMAIL PROTECTED]> wrote:
>
> > btw - the initial compliance tests are here:
> >
> >
> http://opensocial-resources.googlecode.com/svn/tests/trunk/compliancetests.js
> >
> > (these are a wip, any patches are very welcome!)
> >
> > - Cassie
> >
> >
> >
> > On Thu, Feb 21, 2008 at 7:32 PM, Kevin Marks <[EMAIL PROTECTED]>
> wrote:
> >
> > > Trying out Cassie's initial spec compliance app which I had to modify
> > > to:
> > > <Require feature="opensocial-samplecontainer"/>
> > > instead of
> > > <Require feature="opensocial-0.7"/>
> > >
> > >  you can see that you're getting there. Presumably not having any
> > > owner info is due to a lack of a real db backend?
> > >
> > > Great progress, looking forward to seeing this soon; we have a lot of
> > > people interested in a fuller PHP version.
> > >
> > > On Thu, Feb 21, 2008 at 9:45 AM, Cassie <[EMAIL PROTECTED]> wrote:
> > > > So Shindig definitely needs its own tests. The spec will only
> include
> > > >  compliance tests because thats all it can/should enforce.
> Compliance
> > > tests
> > > >  will never be equivalent to a regular testing suite because the
> spec
> > > allows
> > > >  a lot of different ways to handle one call.
> > > >
> > > >  It can't test for instance that a person's name field is equal to
> > > what you
> > > >  think it should be. That could only be in a shindig specific test.
> > > The
> > > >  compliance test would only make sure that the call returned an
> > > >  opensocial.Name object and had at least one field filled out.
> > > >
> > > >  Hope that helps explain things.
> > > >
> > > >  - Cassie
> > > >
> > > >
> > > >
> > > >
> > > >  On Thu, Feb 21, 2008 at 9:15 AM, Bruno Bowden <[EMAIL PROTECTED]>
> > > wrote:
> > > >
> > > >  > Fantastic Chris. Thanks for the contributions.
> > > >  >
> > > >  > To help keep the two versions in sync, we should have test cases
> > > that can
> > > >  > be
> > > >  > run against both versions. That's something that belongs with the
> > > spec and
> > > >  > outside of Shindig.
> > > >  >
> > > >  >
> > > >  > On Thu, Feb 21, 2008 at 4:30 AM, Chris Chabot <[EMAIL PROTECTED]>
> > > wrote:
> > > >  >
> > > >  > > Hey guys,
> > > >  > >
> > > >  > > Since no one likes to hear "it's being worked on" without being
> > > able
> > > >  > > to see it, and i now have the first visible workings of a php
> > > version
> > > >  > > i thought i'd post a bit of a progress report.
> > > >  > >
> > > >  > > I've been working for the last 7 or 8 days on a php port of the
> > > java
> > > >  > > shindig code, I've chosen to follow the structure as closely as
> > > >  > > possible where the language differences allows for it to make
> > > >  > > supporting the same features and having the same behavior
> easier.
> > > Now
> > > >  > > it's been a bit of a challenge, partially because the last time
> > > I've
> > > >  > > read any java code is 8 years behind me, and partially since
> it's
> > > a
> > > >  > > huge undertaking to port an entire large feature set and
> service
> > > >  > > without being able to test it until all the components are
> done,
> > > and
> > > >  > > php being a completely different environment then the
> java/jetty
> > > one.
> > > >  > >
> > > >  > > I've now reached the point where the basic structure is
> working,
> > > and
> > > >  > > it successfully renders the todo.xml gadget into a valid and
> > > correct
> > > >  > > gadget.
> > > >  > >
> > > >  > > I'm still working on the js, rpc, proxy and gadget token bits
> so
> > > their
> > > >  > > not included in this, also the remote content gadgets are not
> > > >  > > supported yet (though that will be very quick to follow).
> > > >  > >
> > > >  > > I expect to have those lacking bits done in the near future,
> and
> > > >  > > combined with a good bit of testing i hope to have a 'patch'
> > > ready for
> > > >  > > submission soonish (and i need to formalize the permission to
> > > submit
> > > >  > > it under the apache license with my employer, but that is a
> mere
> > > >  > > formality).
> > > >  > >
> > > >  > > Feature wise it's completely compatible to the java cousin,
> > > including
> > > >  > > the 'features', message bundle, etc support and only the
> > > GadgetServer,
> > > >  > > web service and caching parts are different from the java
> version
> > > (in
> > > >  > > memory caching for instance couldn't work in a multi process
> php
> > > >  > > environment), and i think i'll soon add a few more php specific
> > > >  > > optimizations soon to speed the whole thing up a bit as well.
> > > >  > >
> > > >  > > The code i'm using for porting is from feb 14th, i'll evaluate
> > > the
> > > >  > > patches that happen in the meantime once i've got everything
> > > done,
> > > >  > > since doing both would be to much of a strain progress right
> now.
> > > >  > >
> > > >  > > A working demo can be seen here:
> > > >  > >
> > > >  > >
> > > >  >
> > >
> http://shindig.chabotc.com/ifr/?url=http://www.labpixies.com/campaigns/todo/todo.xml
> > > >  > >
> > > >  > > And the code-so-far for those who are interested in a sneak
> > > preview of
> > > >  > > it is here:
> > > >  > > http://www.chabotc.com/php-shindig.tar.gz
> > > >  > >
> > > >  > > Please do keep in mind that it's still work in progress, so i
> > > know
> > > >  > > there's no gadget container yet, no proxy implementation etc;
> If
> > > you
> > > >  > > just want a fully functional php shindig server, please wait a
> > > little
> > > >  > > more since this is nothing but a code preview :)
> > > >  > >
> > > >  > >        -- Chris
> > > >  > >
> > > >  > >
> > > >  >
> > > >
> > >
> >
> >
>

Reply via email to