On Jul 23, 2009, at 12:00 AM, pamela (Google employee) wrote:

>
> Office Hours: 2009-07-23 [CLOSED]
>
> Welcome to API office hours!
> *Please top post your question or comment about the robot, gadget, or
> embed API to make sure we see it. You can do so by replying to this
> blip.*
> *A transcript of this wave will be made publicly available. If you
> don't want to appear in the transcript, delete your blip after getting
> a response. Be careful not to remove neighboring blips.*
> *And don't add robots to this wave!*
>
> 2:02 pm
>
> Bastian Hoyer:
> good morning everyone :)
> 2:28 pm
>
> David Byttow:
> Morning? It's 9:30pm! :)
> 2:29 pm
>
> James Purser:
> it's always obscenely early oclock somewhere
> 2:29 pm
>
> Bastian Hoyer:
> it's 6:20 am and I can count the days I was up that early this year on
> one hand I think ;)
> 2:03 pm
>
> me:
> hiya, got a question?
> 2:30 pm
>
> James Purser:
> Hey, was just wondering if anyone else was trying to get the WRFS
> working and had any success
> 2:05 pm
>
> Johannes Feulner:
> Good morning,
> I am trying to write a gadget using GWT in combination with RPC.
> However, RPC consistently fails.
> I found the GadgetRPC sample
> http://gwt.google.com/samples/GadgetRPC-1.0.2/com.google.gwt.gadgets.sample.gadgetrpc.client.GadgetRPC.gadget.xml
> but that one also fails on RPC. The Browser produces a SOP
> violation.How can I get that sample running?
>
> 2:11 pm
>
> Vadim Gerasimov:
> Do you have a non-compiled version of this?
> 2:14 pm
>
> Vadim Gerasimov:
> For GWT gadgets I recommend using our current libraries. I'll add you
> to the wave that has them as attachments.[Wave Syd Dev Day - GWT Wave
> Gadget]
> 2:15 pm
>
> Johannes Feulner:
> Good, what do mean by "non-compiled version"? I took the sources from
> the sample as included in the gwt-gadget package.
> I am not yet a participant in this wave ... now I amm tnx
> 2:15 pm
>
> Vadim Gerasimov:
> Try again.
> 2:20 pm
>
> Johannes Feulner:
> Hi Vadim, I took a brief look at the sources. Looks like a good
> wrapper for the gadget functions. But at first glance, it doesn't seem
> to address the proxy problem. In that wave there is the sample gadget
> of Alex from Lombardi. That one does not use RPC, which is what I am
> trying to do.
>
> Did you the GadgetRPC sample running?
> 2:21 pm
>
> Vadim Gerasimov:
> Let me take a look at the sample source. What is it supposed to do?
> 2:24 pm
>
> Johannes Feulner:
> The gadget shows a button. On pressing the button a RPC call is made
> to the server side which return some data like Version of server side,
> pretty much like the Hello World sampe of GWT.
> You can install an run it from the gadget manifest url above.
> 2:27 pm
>
> Vadim Gerasimov:
> The gadget incorrectly identifies the proxy. For some reason it is
> using a wrong host name.
> 2:32 pm
>
> Johannes Feulner:
> What would be the right one? I fiddled around with te name to get rid
> of SOP violation. But then the proxy returns 404 error : POST not
> supported.
> Trying the same in iGoogle produced the very same effect.
> 2:32 pm
>
> Vadim Gerasimov:
> It could make calls to /gadgets/proxy/
> refresh=60&container=wave&gadget=gadget/...
> 2:32 pm
>
> Vadim Gerasimov:
> Do not specify the host.
> 2:34 pm
>
> Johannes Feulner:
> You mean, simply strip of the host part?
> 2:34 pm
>
> Vadim Gerasimov:
> yes
> 2:34 pm
>
> Vadim Gerasimov:
> There maybe a more fundamental problem with the current gadget
> container compatibility with the GWT gadget library.
> 2:36 pm
>
> Johannes Feulner:
> http://www.ig.gmodules.com/gadgets/proxy/refresh=3600&container=ig&gadget=http%3A%2F%2Fgwt.google.com%2Fsamples%2FGadgetRPC-1.0.2%2Fcom.google.gwt.gadgets.sample.gadgetrpc.client.GadgetRPC.gadget.xml/http://gwt.google.com/samples/GadgetRPC-1.0.2/GadgetRPC
> This is the proxy URL I get. Does it look OK that the URL to be called
> is simply appended to the gadget URL? Shouldn't there  be someting
> liek &proxy_url=... ?
> 2:37 pm
>
> Vadim Gerasimov:
> This is ok if you run it from an iGoogle container.
> 2:38 pm
>
> Vadim Gerasimov:
> Yes, it's supposed to work this way.
> 2:38 pm
>
> Johannes Feulner:
> So I will try stripping the host part. That will take a while. May it
> be possible to reach you after office hours?
> 2:39 pm
>
> Vadim Gerasimov:
> Yes, just send me a wave to [email protected]
> 2:39 pm
>
> Johannes Feulner:
> Thanks.
> 2:06 pm
>
> Bastian Hoyer:
> which browser are you using ? did you try firefox ?
> 2:07 pm
>
> Johannes Feulner:
> Using firefox and Chrome, but from the error message I have the
> feeling, that the proxy of the wave container is not working or the
> name of the proxy is configured in wrong way. The name is returned by
> the cachedUrl() intrinsic function.
> 2:08 pm
>
> Johannes Feulner:
> Do you know of anyone who got a gadget with GWT+RPC running?
> 2:10 pm
>
> Bastian Hoyer:
> I did not try that because I'm not that firm in java and profer
> python ;) but when I played with some gadgets I got error mesages wen
> I used the <Require feature ...> and they were gone when I loaded the
> scripts in content field.
> 2:11 pm
>
> Bastian Hoyer:
> btw any eta on the new python api ? ;)
> 2:11 pm
>
> David Byttow:
> New python API? I just pushed out a new release a couple days ago and
> new zip file. What specifically were you referring to?
> 2:12 pm
>
> Bastian Hoyer:
> hmm.. i think there was something announced like api 0.2 ;)
> 2:12 pm
>
> David Byttow:
> Ah, yes, the wire protocol format. We're still working on that, the
> only announcement we made was that it would be available sometime
> before our extended preview on Sept. 30th. :)
> 2:14 pm
>
> Bastian Hoyer:
> Ah :) I hope the Python api gets more features like forms and gadget
> stuff.. or would it be possible to get a whitepaper about how the json
> output should look like for operations ? ;)
> 2:13 pm
>
> Douwe Osinga:
> What specifically are you missing? Gadgets should be supported now and
> I think forms should be workable
> 2:14 pm
>
> David Byttow:
> Most of this is supported, but the gadget stuff could use a little bit
> more code to support state modifications.
> 2:14 pm
>
> Douwe Osinga:
> THe thing with the JSON wireformat is that it is rather messy right
> now and we'd like to clean that up before describing it. But we're on
> it.
> 2:15 pm
>
> Bastian Hoyer:
> Will the current implementation still be supported when 0.2 api is
> released ?
> 2:16 pm
>
> Douwe Osinga:
> Yes.
>
> Gadget wise it is not the prettiest, but in Python you should be able
> to do:
>        gadget = document.Gadget('http://kitchensinky.appspot.com/
> public/embed.xml')
>        gadget.waveid = new_waveid
>        doc.AppendElement(gadget)
>
> 2:17 pm
>
> Bastian Hoyer:
> yes that works.. It's a bit harder to remove a Gadget :)
> 2:16 pm
>
> Bastian Hoyer:
> Will the 100k people invite still be the sandbox or will it move to a
> different domain? Ijust hope all people with sandbox account get in
> there ;)
> 2:17 pm
>
> David Byttow:
> I don't think we've announced that specifically yet. But, I will say
> that the sandbox is purely for development at this stage. :)
> 2:19 pm
>
> Bastian Hoyer:
> btw in on german keyboard you can't type an (at) sign in the editor
> unless you change that debug option with non-whitleisted combos.. and
> then it only works one time ;) it's altgr-q :)
> 2:21 pm
>
> Johannes Feulner:
> Want to confirm that.
> 2:45 pm
>
> Douwe Osinga:
> Filed a bug about this internally.
> 2:19 pm
>
> Douwe Osinga:
> Interesting. I'll report that to the i18n people
> 2:21 pm
>
> Karma Dordrak:
> Might be a little off topic, but I think need a generic way to contact
> robot authors for bug reports.  I think many of us have found issues
> with existing robots but there is no way to report the bugs.  Also
> there is this issue of robots crashing regularly... like Rosy never
> seems to be online these days.
> 2:22 pm
>
> David Byttow:
> We've added author support to our new extension manifest format.
> Unfortunately, people can create robots without providing a
> manifest... We're working on this, right now it's pretty much the wild
> west.
> 2:23 pm
>
> Bastian Hoyer:
> hmm... you could add a suggestion in the docs that robot authors could
> create a feedback page on their http://robot.appspot.com page :)
> 2:24 pm
>
> David Byttow:
> Yep -- or, perhaps, it could be auto-created for them via the API. For
> example, in Python API, we could automatically create a handler with a
> simple feedback form (e.g. http://my-robot.appspot.com/_wave/feedback)
> or something like that. Then the author of that would have access to
> the data that is stored. There are lots of little things like that
> which could be added (in fact anyone could do it and submit a
> patch). :)
> 2:23 pm
>
> Karma Dordrak:
> I understand.  I was thinking of something like a "report bug" on the
> robot contact that sends reports back to the author or their bug
> tracker.
> 2:23 pm
>
> David Byttow:
> Yes, this is do-able, and useful to support at the high-level. It's a
> good idea. :)
> 2:25 pm
>
> me:
> Hmm. (Nice Idea) Would be nice to do something that's consistent for
> gadgets as well, perhaps?
> 2:27 pm
>
> David Byttow:
> Unfortunately gadgets are client-side, so we'd need to support
> something in the client or... hmm, not sure really. It's a good idea.
> But, also, if extensions were the only means of adding robots/gadgets,
> then we could do it all at that level (via the author tag).
>
> Actually bring up a good point... we could potentially introspect the
> author_email from the gadget spec and show or provide an option to
> email the author on the gadget drop-down menu.
> 2:29 pm
>
> Karma Dordrak:
> Or webpage resource.
> 2:25 pm
>
> Karma Dordrak:
> Do we know why the robots are crashing regularly yet?  Rosy is never
> working.
> 2:25 pm
>
> David Byttow:
> Rosy was disabled externally for now.
> 2:31 pm
>
> Karma Dordrak:
> If a bot is disabled that should really reflect somehow, maybe with a
> sleeping or grey icon or something.
> 2:32 pm
>
> David Byttow:
> Agreed!
> 2:25 pm
>
> David Byttow:
> How do you mean crashing? Simply just not working?
> 2:26 pm
>
> Karma Dordrak:
> By crashing I mean stops responding, and eventually even their icon
> disappears.   Rosy is the classic example at the moment.  I've yet to
> see it work at all.  One of the reasons I want to be able to report
> bugs :-)
> 2:27 pm
>
> David Byttow:
> bugs: http://code.google.com/p/google-wave-resources
> 2:28 pm
>
> Karma Dordrak:
> Is this for the google created bots too?
> 2:29 pm
>
> David Byttow:
> Sure. It's for anything you want to report.
> 2:22 pm
>
> Douwe Osinga:
> The other thing is that we haven't really supported the robot profile
> url bit in the spec which presumably would allow you to bring up the
> profile card for the robot and give more info there.
> 2:24 pm
>
> Bastian Hoyer:
> I think a way to subscribe/unsubscribe from the wave-discuss group
> would be great ;)
> 2:30 pm
>
> Bastian Hoyer:
> I guess you can't give any hints what's coming with the next
> update ? :)
> 2:30 pm
>
> David Byttow:
> With respect to robots?
> 2:31 pm
>
> Bastian Hoyer:
> or wave in general ;)
> 2:32 pm
>
> David Byttow:
> Here are some things coming soon:
> - Robot wire protocol.
> - Direct server/robot access to wave servers (ability to create new
> waves, get your waves, etc).
> - More robots/gadgets!
> ... and we'll be announcing more soon. :)
> 2:33 pm
>
> Bastian Hoyer:
> hmm.. anything of that until the next office hours next week ? or is
> it more a soon like more then a week ? ;)
> 2:34 pm
>
> David Byttow:
> Good question. Hard to say... We're getting our "roadmap" together for
> external publishing. One thing I can say though, is, we have an
> extremely packed schedule of features/fixes/etc to address before
> Sept. 30th, and we'll roll them out as they're ready.
>
> This includes changes across the board from API to the client.
> 2:36 pm
>
> Karma Dordrak:
> Is there a roadmap published anywhere with milestones?
> 2:37 pm
>
> David Byttow:
> Not yet, coming soon.
> 2:36 pm
>
> Bastian Hoyer:
> yeah some kind of changelog would be nice... I have seen a wave with
> that but I think it was destroyed by bots and gadgets...
> 2:36 pm
>
> David Byttow:
> I agree, we must get better about this. We've just been rushing
> through changes recently to test/fix, etc. I'm sure change logs, etc
> will solidify soon.
> 2:37 pm
>
> Bastian Hoyer:
> Maybe a Webpage would be great with links to welcome waves for
> starters
> 2:37 pm
>
> David Byttow:
> Or perhaps a page with an embedded wave. :)
> 2:37 pm
>
> Karma Dordrak:
> I'm used to using Trac. We log all features/bugs/tasks as tickets then
> assign them to milestones.  This then autogenerates a roadmap.  As you
> close tickers the milestones get ticked off.
> 2:38 pm
>
> David Byttow:
> Sure -- we do have various tracking methods internally. The problem is
> the line between what we publish internally and externally. But yes,
> process is improtant.
> 2:39 pm
>
> Karma Dordrak:
> Just thinking as a motivating factor.  I saw quite a few dev express
> concern that the Api wasnt stable or prepared enough.  When you can
> see what's in the pipeline it really helps.  this is most importand
> when developing free software since we need to invest our time wisely
> and not on an unstable API.
> 2:39 pm
>
> David Byttow:
> Definitely, and it's really not. Hence developer "preview." I do
> believe that a roadmap is important especially to business
> development, etc.
> 2:41 pm
>
> Bastian Hoyer:
> At least now I have another reason to learn more java besides the wish
> to write my own android application ;)
> 2:42 pm
>
> David Byttow:
> I'm off now (it's late here in the US). Take care guys!
> 2:42 pm
>
> Karma Dordrak:
> cya,. and thanks!
> 2:44 pm
>
> Bastian Hoyer:
> cu ;)
>
> btw.. I can't find this wave with the search function and it's not in
> my inbox ,)
>
> >


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"Google Wave API" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to 
[email protected]
For more options, visit this group at 
http://groups.google.com/group/google-wave-api?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to