Forum Moderators: Robert Charlton & goodroi
Google to launch Friend Connect for the social Web Monday [webmasterworld.com]
Facebook Disconnects Google Friends Connect [webmasterworld.com]
But rather than focus on the social networking angle, I was wondering, what about the content? Do you have any concerns regarding "your" content?
Anytime a site starts publishing user generated content, I think there's always opportunity for trouble and abuse, along with all the good stuff. Just look at the history of blogs! I'd need to know a lot more about those management controls to feel totally good about adding Friend Connect. And I'm also uncomfortable with Google graphing out social connections.
I'm also thinking I'd like to see some kind of "section targeting" to scuplt the organic indexing of the social areas - similar to what Adsense pages allow.
Let me lay another layer down here ... let's say that the content required approval first, not only by Google and their TOS, but by you. Let's set aside the content approval, we will pretend that Google approves, and you approve. OK, now I have user-generated content on "my" site. It's not really stored on "my" site though now, is it? No, it's not. What concerns might that raise?
It's not really stored on "my" site though now, is it? No, it's not. What concerns might that raise?
coop - Are you concerned...
- about the nature of web apps in general (that your data and infrastructure are viable {and private} only as long as the web app company is viable {and private})?
- about Google as big bro?
- about both of the above?
- or about the general promiscuity of it all?
I haven't done much research either... but from the introductory video, it's not clear where the the data resides for a given developer's app. I'm guessing it's with the application developer, not with Google and not with you.
As for who's tracking what, that's also not clear... but I do know that last night as I was about to do a transaction online with a company I didn't know and was presented with the option to sign up for Google Checkout (which on the surface was much safer than going with a credit card), I thought about it for a while and then decided that I'd better know more about what Google does with my shopping data before I did sign up for Checkout.
With Friend Connect, I think Google could track... in addition to everything about me... what my friends are searching for and where they're going on the web... and maybe eventually what they're buying too. But this is just a top of my head-wearing-tinfoil-hat thought.
However, I'm looking at this whole concept for what it is but also for what it is not. Let's take a Facebook application as an example. Take a look at a Facebook application that a site owner develops because ultimately it benefits him by creating interest in the products/services that the site owner sells. Facebook hosts the pages and developers write their 3rd party applications using the Facebook API to hit their 3rd party servers and deliver content. Facebook is in full control of the site and ultimately the content and delivery (assuming the 3rd party server is up and running effectively). As the Facebook application owner, we realize that we are at the mercy of Facebook in regards to the application and content and it's effective delivery. We do everything we can on our end to keep up our end of the deal. It ultimately benefits me because my intention is to get people involved with my application to move them toward my products or services. I'm using Facebook as another form of marketing. It's not my primary domain or selling point, but merely another marketing resource.
Now take the Google model here. It is much the same, except it is my domain, my site. Now, rather than relying on my own server for content and it's effective delivery, I am at the mercy of Google services and/or a 3rd party developer and their services. It's not like I am developing for Facebook here, it is my site. I'm used to having control of my site. Yes, I seemingly still have control, but in many ways I have relinquished control. Perhaps not the control of the content itself (according to the BETA model today), but indeed I am at the mercy of others when it comes to indexing the content and the effective delivery of the content.
...I am at the mercy of others when it comes to indexing the content and the effective delivery of the content.
This is the problem I've had with web apps in general. In a way, though, it's even a problem with data on your local machine that depends on continuity of the application company, the operating system, and the long time viability of the data format itself.
In the late 90s, here in the SF Bay area at the height of the dot-com boom, there was a first wave of web apps... tools for things like building sites that allowed easy client editing of content and all sorts of really nice built-in functionality... but the question I always asked was "what happens to my site when you go out of business?" The tee-shirts and trade-show swag outlasted the companies, but now even those are in the rag bin.
So, you've got to choose your web apps carefully. Google's undoubtedly going to be in business for a while, and it looks like the APIs will become standards. Not sure what kinds of standards bodies there are behind these, or what sorts of agreements exist about things like privacy or spiderability.
I suppose, since the APIs will be public, you could build your own apps... or that a group concerned about the issues you're concerned about could do so. How long do you promise do be around, though, if I hook into any of your applications on my site? ;) Ditto on server up time at your end.
Even without remotely served apps, some big sites on the web that depend heavily on ad servers for full loading have become so slow they're almost unusable.
How would these apps, btw, interact with Google search? What happens when you've got a page of Friend Connect messages that appear on a dozen other social sites?
How long do you promise do be around, though, if I hook into any of your applications on my site? ;) Ditto on server up time at your end.
Exactly! The apps I develop are not on my server, they are on your server. There is a summary at the end of the Google Video Introduction that states "Function w/o coding" and the taglines on the index page for Friend Connect state:
No programming whatsoever. Just copy and paste snippets of code into your site, and Google Friend Connect does the rest.
Today, if you want features like this you insert open source code, you code them yourself or you hire somebody to code them. The end result is that the function resides on your server and belongs to you. If it is down it is because your server is down. Granted, you can also find services that you incorporate into your site, and Friend Connect is just another one of those services/features. And therein lies my question, what concerns arise?
Even without remotely served apps, some big sites on the web that depend heavily on ad servers for full loading have become so slow they're almost unusable.
Another question/concern that came to mind for me as well. I wrote them down as ...
Some search engines (including Google) will stop a crawl on some site pages if response times are not meeting certain requirements. Google WMT will let you know if you have any issues here. External content (ad servers) are one of the biggest issues, today. Now, if the service being provided reaches that threshold and I am counting on that information/content to be indexed, where do I stand? Do I stand to lose a little link love? Possibly, yes.
How would these apps, btw, interact with Google search? What happens when you've got a page of Friend Connect messages that appear on a dozen other social sites?
Great question, and one that will likely only be answered over time. I imagine the latter question here is in regards to possible duplicate content issues? Yes, yet another question on my mind. I suppose you attempt to create enough unique content on your page otherwise in order to avoid that possibility. But, we still don't know how Google is going to handle that on the backside! Perhaps they monitor the same thing and have a different algorithm for determining dupe content based on the linking/API being used.
