The Mine! Project

open source project for online data and relationships logistics

Truly social software?

Tags: , , , ,

When describing how the Mine! could help me manage my data better than the existing applications I use, one of the examples I describe is uploading photos to Flickr and trying to structure them.

Flickr is a flow of photos and I am not really in charge of how they are organised. For example, I have 100+ wine photos as part of my wine interest. If I upload all of them, that’s what people subscribing to my Flickr will see as they all appear in my Flickr stream. I have a choice of a couple of combinations of friends & family settings but that does not solve my problem: I may want some people to see the wine photos and I may not want others to be bored by a bunch of wine bottle shots. Some people may not be on Flickr, so the privacy settings don’t help.

On another occasion, I needed to share photos with my mother who is not on Flickr. It was a practical need – we went shopping together for items for her apartment and the photo set was meant to help her remember and decide what to buy and exchange notes and comments on the photos. I couldn’t make it work. I tried setting a new account for only those photos but there were too many for a free account and I didn’t feel like paying $24 for this simple use. I tried signing her up to Flickr, as a family contact, and uploading the photos with the family setting. This was awkward as I don’t necessarily want other contacts with the same privacy setting to see those photos or me being forced to change their status permanently…

I needed my own space, not just for communication or publishing but for my own notes for future reference. Then I also wanted to share it with those who might be interested in my wine photos or window-shopping.

I want two basic functionalities from the online tools, which help me organise my life and connect me with people. First, I want to capture and sort out my data, upload photos, take notes, cross-reference information, etc. For that I need applications that are more analytical than the current social media/web tools. Once I organise my stuff, I want to go on sharing it in ways that are more social than the current web 2.0 tools allow me to be.

But isn’t social networking all about being social? Not quite. At the moment, I don’t drive who gets to see what beyond simple decisions about who is ‘in’ and who is ‘out’. Social interactions and relationships are far more granular than social networks allow them to be. Usually, this is seen as a privacy issues and results in a complicated access management e.g. Facebook privacy settings.

Why do we have our relationships pre-determined by others such as Facebook, Flickr, Plaxo etc.? Presumably to give us more ‘control’ over our social network and contacts in it. But how is lumping people into categories imposed by an application helping me to be social? By determining the types of relationships I am able to have – business contact or colleague, family or friend, I am not able to reflect relationships I already have. The best social software is not online, it is loaded on to my cortex. And no software can fully map the relationships, let alone replace our natural ability to create and maintain them.

Privacy is merely the other side of the coin of complexity in human relationships. My ‘privacy settings’ are inherent in my behaviour. My privacy policy should not be embedded in any software. In that sense, software cannot be social (or antisocial), though it can help me be more or less social. Software privacy settings limit my ability to be truly social i.e. capable of maintaining complex relationships and interactions with others – arguably the purpose of such tools.

The Mine! needs to satisfy both requirements of my online life – allow its users to organise data differently and support people’s relationships as defined by themselves.

Let’s deal with the data in this posting – I will do the privacy & relationships in a future one.

For some type of data a flow is just fine – e.g. Flickr, FriendFeed, PlaxoPulse, Facebook etc. Structure sometimes emerges – sets, rooms, groups. For other purposes I may need alternate data structures and new functionality to build them. A pool of tagged objects is a good start, it is flexible and not determining how the data is organised. The user who generates the data also owns the actual raw data, as opposed to its formatted representation – e.g. Facebook, Amazon reviews. There will therefore, be many more options how to manipulate them. (For more see Models of Data Imprisonment.)

Various ways of thinking about data structures…

  • created prior to data input as a skeleton for data with known or standard structures to be stored in later – perhaps medical or financial data or other complex data
  • created with input of new data – when you upload photos, you create sets; when you bookmark a link, you add tags and notes etc
  • created on retrieval – the hierarchy or structure emerges at click of button depending on what you are looking and on the flow or the dynamic of the data, e.g. I click on a tag in del.icio.us and get all articles tagged with it
  • shoe-horned into single vision – google reader and del.icio.us, pick your means of rendering – by tag, by who person, date, no tag at all
  • created from a pool of objects, with tags and meta-data, with functionality that helps you create whatever hierarchy you want

The Mine! needs to have a user-driven structure. Once the user has the option of putting the data under his ‘domain’ i.e. in the Mine!, he can create new data or import existing data. The user can then manipulate it; mash it up, trend it, analyse it, collage it. Extra functionality can come from application or plugins which will allow the data in the Mine! to be structured as the user sees fit.

A Mine! plugin would be a package of functionality that enables users to manage a particular topic or format of data. They can create groups or categories of their own, reflecting areas of interest – travel, restaurants, shopping, cars, wine, fashion, cosmetics, sport, etc. This would be not dissimilar to a scrapbook – not a hierarchy or a taxonomy but a patchwork of stuff that the owner of the scrapbook is interested in.

Hierarchy is often synonymous with order. A feature of hierarchy of information (taxonomy) is that it exists outside the user’s mind. The web has driven home the point that taxonomy is by far not the only order possible.

It is all very well to insist on user-driven structure, but where is the convenience if users have to determine not only the data flows but their underlying structure? To flip the way we tend to think – from the user’s point of view, the structure doesn’t have to precede the data. Emergent order is more user-friendly in the long run – think folksonomies and tags vs. directories and folders. In order words, complexity should come from usage, not design.

Tags: , , , ,

7 Responses to “Truly social software?”


  1. VRM Hub » Truly social software?
    on Sep 15th, 2008
    @ 9:18 pm

    [...] have been thinking about how social software and social networking platforms actually limits my ability to be [...]


  2. Jim
    on Sep 19th, 2008
    @ 3:04 pm

    OK, I’m going to pitch in here as someone who’s not been privy to a lot of the discussion that’s going on, so my ideas are probably a little behind..

    As a coder, I’m trying to think how you could achieve this – perhaps each datum needs to be input with a raw minimum of the most general metadata we can possibly think of: dates of creation and modification, title, description, some ‘tags’.

    But surely there’s more to controlling your data that just labelling it with tags? For example, can we allow the user to create their own metadata fields? Simply by either defining (key,value) pairs for their objects on an ad hoc basis, the same way you add tags only with a value; or providing a UI interface for managing their own keys, to which they would be prompted to add values for the item types they specify (”You’re adding a photo, I’ll prompt you for the ‘PlaceTaken’ or ‘Person’ fields that you created”).

    Not sure if I’m making sense here, I’m stumbling around in the dark a bit, but you’ve got to start somewhere.


  3. alecm
    on Sep 22nd, 2008
    @ 9:53 pm

    @Jim – there’s nothing constraining the code to tags, but they are a good start and people are familiar with the concept; we’ve some ideas about tag matching which are interesting – eg: rather than label winebottle.jpg with both “wine” AND “merlot” instead you can choose to just stick “merlot” against it, and create a general rule somewhere else that “merlot” is a “wine”…

    …and then if there is a relationship rule saying Jim is interested in “wine” then you’ll get a feed of winebottle.jpg; unless, of course, your rule says “wine not:merlot” in which case you’ll get the picture of chardonnay.jpg instead…

    But limitations? No, there are none. Publication is via ATOM feed. What goes on in the Mine! itself is a matter of user interface and good design…


  4. Jim
    on Sep 23rd, 2008
    @ 9:18 am

    That’s what scares me. I can’t help but think you need to lay down some structure for all this to hang on. Tags are good, but you need (at least) something that extends them from the boolean into key/value pairs, as a basis at least – it’s nicely flexible, particularly if values can be feeds or elements of a feed.

    I’m not sure how flinging in inference rules might not complicate matters, but it can be implemented as a plugin layer on top of the basic database, so it doesn’t matter that much, I suppose. People can change it if they don’t like it.

    Hm.. in fact, provided you have a very simple base metadata system consisting of key/value pairs and, maybe, a publicly available and extensible list of keys and their meanings (kept on a Wiki or something), then everyone else can write plugins that add more meaning to that.

    I still feel like I’m missing some elements of your overall vision though, and I have a suspicion I might be missing the whole thing; but I suppose

    “a system which allows ordinary users to own, catalogue, mashup and manage their own data (of any type provided plugins exist) and also the data of other people who provide them with (possibly audited) feeds of their data”

    is one way of looking at it. It does unfortunately omit the VRM aspects, though; particularily the idea of providing your personal data as a feed to the companies with whom you have a relationship rather than them getting to keep and manage it. But maybe that’s good from an architectural point of view; the VRM aspects just fall out of the architecture of the Mine! naturally.

    Hope I’m not treading on anyone’s toes by playing around with these ideas; they’ve been sitting in my head taking up space and partying since we saw you both in Swansea.


  5. alecm
    on Sep 23rd, 2008
    @ 10:04 am

    @Jim

    Actually I think I am already there, but I misapprehended what you were trying to say.

    - Mine objects are tagged to allow the user to categorise objects

    - Mine relationships are reciprocally tagged to determine which objects appear in the feeds for what relationship.

    So far, so familiar…

    However, “under the hood” the objects have exactly what you say, a rich set of “key: value” pairs which can be used and extended by plugins, with a bunch of them pre-booked for database administration purposes; but this is all hidden from the user, hence it doesn’t appear in the user documentation:

    objectComment: this is a comment about mountains.jpg
    objectName: upload from mountains.jpg
    objectStatus: draft
    objectType: image/jpeg
    

    And yes, there is a concept of “an enumerated type”, eg: objectStatus can only be one of draft/final/public

    Once we cut a working beta, I’ll be bugging you to write a plugin for us. :-)


  6. Adriana
    on Sep 23rd, 2008
    @ 12:02 pm

    “Once we cut a working beta, I’ll be bugging you to write a plugin for us. :-)

    Jim, serves you right for doing some thinking on this and letting us know… :P

    Re the Mine! and VRM – the Mine! is intended as an infrastructural point of integration for user’s data, relationships and sharing. The method of sharing data from the Mine! is the way in which I try to redress the balance of power in relationships with vendors – I can create as granular a feed as I wish for each recipient including vendors, I can audit the feed and terminate it whenever I wish without disturbing other feeds/relationships.

    In fact the first paper, implicitly anticipating the Mine!, talks mainly about the mechanism of such ‘data logistics’.

    They can both be found here:
    http://themineproject.org/index.php/the-mine-papers/


  7. The Mine! project » Feeds, feeds everywhere
    on Oct 29th, 2008
    @ 1:23 pm

    [...] is getting at, this feed business is more than RSS. But I don’t think social networks are truly social when it comes to sharing. Newsfeed is but one way of sharing. It is like relationships via [...]

Leave a Reply

© 2009 The Mine! Project. All Rights Reserved.

This blog is powered by Wordpress and Magatheme by Bryan Helmig.