Hi Dan,

OK, fair enough. I suspect my rant was in the hope that these
expectations could be exceeded. :^)

And I'd like to make clear, I think that the changes are for the
better - especially the ability to deliver content through SSL, and
avoid the warning on IE. I guess I'll just have to wait and hope that
some developer support comes along.

Thanks for your response,

String

On Apr 30, 5:50 pm, "Dan (Google)" <[email protected]> wrote:
> Hi String,
>
> I feel your pain, but the gadgets API in Gmail is, as you note, a
> Gmail Labs feature which bears the warning: "They may change, break or
> disappear at any time." From a formal standpoint, there isn't any
> documentation, nor was there a developer blog or group when you
> started using the API. I'd agree on calling out a communications
> failure if these channels existed and were silent, but they do not
> exist and I believe the Gmail team set expectations accordingly.
>
> Dan
>
> On Apr 30, 3:02 am, String <[email protected]> wrote:
>
>
>
> > Code-breaking change to gadget support on Gmail
>
> > Sometime in the last 48 hours, a change was made to the Gmail gadgets
> > platform that broke most of my gadgets. I've been digging into it
> > since I noticed the change (time permitting), and here's what I've
> > found.
>
> > Previously, gadgets were served from subdomains of gmodules.com, such
> > as:
>
> >http://1.gmail.gmodules.comhttp://2.gmail.gmodules.com
>
> > Where each gadget on the page had a different number at the beginning.
> > This is congruent with iGoogle gadgets generally, they've always been
> > served from gmodules.com, regardless of host platform.
>
> > Now, the host for Gmail gadget iframes look like this:
>
> >https://psglh63b1qffbcf6jeem59as0o0j9vro-a-gm-opensocial.googleuserco......
>
> > Again, the exact subdomain is different, but apparently the long
> > string at the beginning is a hash of the gadget URL - you can't serve
> > one gadget from another's subdomain, whereas before, the numbers were
> > generic. That's no big deal, and might actually enhance security.
>
> > But note the base domain is different. This in itself broke all my map
> > gadgets at a stroke. The GMaps API requires a domain-specific key, and
> > while it recognizes certain domains (like gmodules.com) as Google-
> > owned and not requiring a key, googleusercontent.com isn't one of
> > them. There is a workaround (register an API key for
> > googleusercontent.com), but it requires changing the gadget source
> > code.
>
> > The other issue, affecting non-map gadgets as well, is that the iframe
> > is now served via HTTPS. This is good, in that the mixed content
> > security warning can now be avoided on IE, but there's a serious
> > downside. For gadgets that retrieve additional content via XHR, those
> > retrievals now fail if they don't also use HTTPS. So again, a code
> > change is required to make any such gadgets work again on Gmail.
>
> > Here's the bottom line: This is abominable behavior from Google for a
> > published API, even one that is still in Labs. The very least they
> > could have done was warn us that this was coming, on the Gmail Gadgets
> > Developer's Group... what's that you say, there is no such group? Oh
> > yeah, there is that small problem. There isn't anywhere that support
> > is offered for this.
>
> > Come on Google, you know better than to treat developers like this.
> > Communication is vital.
>
> > String
>
> > -------------
> > Cross-posted at the Gmail Labs Help Add Any Gadget group and the
> > iGoogle Developer Forum because, like I said, the gadgets API on Gmail
> > is an illegitimate child and doesn't have a group of its own.
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"iGoogle Developer Forum" 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-Gadgets-API?hl=en
-~----------~----~----~----~------~----~------~--~---

Reply via email to