Advertisement
(1) Opt-in for the person exporting file has affect on his/her friends
(2) When sending mail to friends of friends, answering requires a person in the middle to protect email - how can FOAF help here?
(3) When invite to Tribe and accepting on a different email address there are unification processes that need to be explicit
(4) Import implies skeleton accounts. How do people later claim them?
(5) Update/consistency - where is the master record and what happens when they get out of sync?
(2) When sending mail to friends of friends, answering requires a person in the middle to protect email - how can FOAF help here?
(3) When invite to Tribe and accepting on a different email address there are unification processes that need to be explicit
(4) Import implies skeleton accounts. How do people later claim them?
(5) Update/consistency - where is the master record and what happens when they get out of sync?
posted by:
|
Mountain View
|
Advertisement
Advertisement
-
Re: Top 5 Current Import/Export Issues
Sat, February 14, 2004 - 2:51 PMSpecifically regarding #1: As things stand today on tribe, any person on the internet may access the public portion of anyone else's profile. Thus, the information contained therein is already "public". In some sense, exposing any one person's "public profile" via FOAF is arguably the same as exposing it via HTML. Or is it??
I think that a great prototype for tribe.net's embracing of FOAF may be to just produce trivial FOAF files that contain nothing more than first name, main photo (if there is one) and the list of friends as resources which may be accessed by FOAFnaut or its companions. This way, we would be able to quickly let the network visualization tools go to town.
Anyone have any thoughts on this, specifically, the privacy issues around HTML vs. FOAF presentation? -
-
Re: Top 5 Current Import/Export Issues
Sun, February 15, 2004 - 7:47 AMYes - having public FOAF rdf URLs is the first step. Everything in that file will be public, so friends is the obvious choice.
All metadata exposed on one's peoplecard would also be a no brainer - since it's all public anyway. In other words up to and including a one-to-one mapping of HTML to FOAF would be reasonable - without expecting ANY flack from end-users on 'privacy'.
It's in the export process - that control can be exerted. And obviously cherry picking - on import. -
-
Re: Top 5 Current Import/Export Issues
Sun, February 15, 2004 - 1:54 PMMarc, can you clarify a bit on what you mean by your last sentence?
"It's in the export process - that control can be exerted. And obviously cherry picking - on import."
First, on the export process, you are referring to control that tribe.net can exert over what gets exported? Or are you referring to the user exerting control over what gets exported? This is indeed an interesting question, especially with respect to the HTML analogy I mentioned previously. In our HTML presentation, we are able to take into account the data of the viewer being viewed as well as the identity of the viewer of that data. It sounds like if we really want to take this "alternate presentation" (i.e. FOAF) to the next level, there would need to be a notion of identity of the retriever that tribe.net can understand and apply all of the same security rules to when determining which data to expose. Another fly in the ointment is around the "Mature content" feature that we have put in, which allows people to say that they aren't interested in seeing the more "edgy" stuff. This is not so much a question of who you are, but rather what you prefer to see.
Second, regarding the cherry picking on import, are you just referring to importing FOAF's INTO tribe for presentation? -
-
Re: Top 5 Current Import/Export Issues
Sun, February 15, 2004 - 8:56 PMHey Bri,
I'm sorry I thought I went over that with you when we met.
Easy one first - Import.
Yes - I was referring to WHO gets imported. So my list of 444 (well let's use as an example the 564 friends I have on Orkut.) So if I import into Tribe - I might not want (probably won't want!) ALL 564 coming in. So by me saying "cherry picking" - I was referring to a giant list of ALL the folks I'm importing - each with a checkbox next to their face. I can go through and decidce who gets brought in or not. That's the cherry picking. Got it?
The issue of exporting (to me) is all about what the end-user wishes to export. Obviously it's implied in the system (in your case - Tribe) to enable end-users to decide how much to export.
So I'd assume TypePad and Blogger would enable their blog posts to go out - or maybe just the most recent ones - not the entire archive, while web links and lists of faves might also be natural.
For teh Tribe - we've talked about one's friends and their Listings. Obviously YOu control - by enabling the end-user with only SO MUCH control - to decide what goes out. The perfect world would have every fucking thing - that's associated with the end-user's profile - to be able to be exported.
So utlimately the end-user decides - but the system (obviously) enables them - which is a form of control as well.
Am I making sense?
Remember swcraping programs can grab anything anyway - so you might as well make it easy on folks - and build their trust, admiration and loyalty!
-
-
-
-
Re: Top 5 Current Import/Export Issues
Tue, February 17, 2004 - 10:39 AMI just posted some thoughts on these issues under the "Plaxo is also working on this" discussion (since I'm at Plaxo), so this is my attempt to cross-list the posting here. :)
Thanks, js
