Re: Pymine!
Alec Muffett
That's a really interesting question; I have my own take on it but I will cc the maillist because I think it's important that these sort of questions get an airing.
First up: it's not really my idea, and I am really doing the coding legwork - which is very significant - but the kudos goes to Adriana Lukas for establishing the concepts the mine, and for trying to give people platforms that allow them to participate *in* the web; my job is just to try and give those concepts some kind of existence.
So it's not "my" work - it's someone else's idea, and a team effort. The code (so far) is my contribution. :-)
You mention Spotlight (which I know) and Bento (which I don't, but will briefly Google) - hmm, webpages for Bento seems to be:
http://www.macworld.com/article/61025/2007/11/bento.html
and
http://www.filemaker.com/products/bento/
Use Bento 3 to do all this and more:
• Organize contacts, clubs and mailing lists
• Track projects, tasks, and deadlines
• Plan special events, parties and weddings
• Link photos to contacts, projects, and events
• Manage students, classes, and lecture notes
• Search wine and movie collections
• Catalog inventory, assets, and equipment
• Record billable hours and payments due
• Keep a daily log of diet and exercise
• Store recipes and shopping lists
...a lot of which sounds very familiar, and which I think will excite the Mine UI team. I can't personally answer the question about "integration", but there certainly sounds like an overlap of interests. I see no barrier to various forms of integration and syncing between your Mine and your computer(s) - but I am too busy to address specifics right now. :-)
As an aside, I do clearly specifically intend my Mine to become the ultimate repository and source[1] for my blog, my delicious-like bookmarking, my recipe collection, (some) of my photos, and (eventually) my tweetstream.
Syncing it "dropbox"-style with Macintosh-apps had not really occurred to me except in the context of messing-around with "FUSE" filesystems, but it's clearly something to be investigated.
To your third paragraph: there have been times I have considered the Mine as a networked hard-disk where everything has automatic *sharing*; for me this contrasts strongly with the semantic web geek's pursuit of "everything on my computer having a URL" - because for me that has basically always been true, and it's not exciting.
Instead what is different and exciting is the concept that "everything on my computer has a URL which is different for every person on the planet, and I can control what that URL does" - that is a completely different proposition, a different lifestyle from the billions-of-broadcasters model we have on the web today.
Re: "the nagging fear that the effort of tagging and feed-management that Pymine! seems to imply", well for me that is kinda the point of the mine, although philosophically it has always been on the same level as "having a place to store your stuff and mess around with it"; however the "feeds" concept springs directly from "everything on my computer has a different URL for every person on the planet" concept - without which we would merely be reinventing either (a) the webserver or (b) WebDAV.
As for your pymine install I am most pleased you're playing with it - I assume you're using the version 403 package from the project site? If so, that's good, because every version since 409 inclusive has not been operation due to major core reworking.
I intend by Monday to have remedied this - I will confirm with an announcement when it happens - but I regret the resulting database code will not be compatible with your Mine, so please don't invest too heavily in it yet. :-)
- alec
[1] How can something be both an repository *and* a source? Hmm. Maybe I need to start using a word like "wellspring" or something?
On 17 Feb 2010, at 21:12, Michael James Hauan wrote:
> Thanks for your efforts in this vein, Alec. I'm eager to watch and participate in a new balance of control over our data, presence, and relationships on the web. I first heard of your work via Linux Journal and Doc Searls' column.
>
> I'm wondering how such an effort overall might integrate with two Mac-specific programs as a specific example of the more general problem. One program is Spotlight, which has an efficient index of everything on my Mac. The other is Bento which is an attempt at enabing the organization and inter-connection of data that might originate and reside in other programs (iCal, iPhoto, spreadsheets, Addressbook, et cetera).
>
> Do you imagine Pymine! using api's to such programs (like Bento itself does) and then providing the data tagging/feeding/management functions to help individuals build their relationships through the internet? For instance, I imagine everything on my computer having a URL -- either in the filesystem (e.g., via Spotlight) or as an object in an api-accessible program's data structure (e.g., via Bento) -- and then overlying that, a Pymine! layer interfacing with the world. One also thinks of the program Growl as an example where events (including data changes) in the underlying programs are organized for presentation to the user.
>
> I ask this motivated by your avowed intent not to re-invent the wheel and by the nagging fear that the effort of tagging and feed-management that Pymine! seems to imply, might also involve a wholesale reorganization of one's current data. (In particular, I'm thinking of the work involved in adding items to Pymine! at present.) It would seem easiest if it just "rode on top" of the way an individual currently organizes their data. If Pymine!'s presence in this regard subsequently improved one's data organization, so much the better.
>
> In any event -- kudos on an excellent idea -- both in particulars but also in overall philosophy. I've installed the latest version on an Ubuntu box and am trying it out.
>
> mjh
>
> Michael James Hauan
> m...@hauan.org
> http://www.hauan.org
> 573-642-8150 home
> 573-823-7114 cell
--
alec.m...@gmail.com
http://www.crypticide.com/dropsafe/
Michael James Hauan
I listened to Adriana's presentation and was very impressed and, again, thankful that people, you in particular, are re-thinking what these technologies mean to our identity and our relationships. It's not just fascinating but critical that we think clearly about this before structures harden around us in unduly restrictive or even harmful ways.
Generally, the user-centric v. user-driven concept might well apply on a single computer as well as on the web. (BTW -- I think of the difference in terms more as the definition of a spectrum, on which programs lie more on one side than on the other, rather than as a categorical distinction.) A helpful illustrative question in this regard, among others, is "can I use my data objects outside this program?". For instance, "can I use my text document outside MS Word or WP or OO?" Or "can I use my media file outside iTunes or Songbird or VLC?" The more difficult or otherwise restrictive the use of data objects is outside a particular program, the less user-driven I'd say that program is likely to be.
Given that, I imagine a user-driven Mine! ideally remaining disinterested in what programs (including itself!) might work on data objects while focusing on 1) why they are important to the user, 2) what the user does with them, and 3) how they support the user's relationships with others.
I like how you situate the importance of URLs, not as some meaning in itself, but as a part of a meaningful action between people -- "here's the thing I want (for some reason) to share with you". That situates a URL in a relationship with a justification, a reason, a desire, that is under the control of *both* parties even if there is a sender-receiver polarity.
Everyone with a computer already has their stuff -- photos, lists, media, appointments, addresses, docs, links, etc. -- organized in some fashion according to what they do and how they do it. Some of it is raw on the filesystem (file://somedir/somefile) and some is buried inside a programmatic wrapper (iPhoto://somealbum/somephoto). They already have their sandbox. So does a Mine! squeeze in alongside all the other programs in there and need the user to add stuff to it or can the Mine! consider everything on the computer mineable? To it's already extant URL (duh!) one adds the rationale (e.g. tags, or some other expression) for the data object being offered by me (authentication scheme applied) to you (authentication scheme applied then the feed) which, as you say, really starts making it interesting.
One can imagine, as using a Mine! becomes a larger part of what a user does, that the organization of the data will adapt to take advantage. For instance, one might define a directory for Alec and place actual files or links to files there that one knows meet the rationale and relationship requirements. Likewise, define an album in iPhoto, a playlist in iTunes, a calendar in iCal, or a contact list in AddressBook that functions the same way. The Mine! would survey the sandbox (like Spotlight which gets inside both files and programs to an amazing degree) or get notified (as in Growl), see the change, and fill the feed.
In any event, my early inclination would be to think that, with a Mine!, the user could be defining and building relationships "out of the box" rather than "seeding" the Mine!
BTW -- I'm a physician trained in medical informatics and have longed to see (since I'm a full-time ER doc at present and not working on informatics) promising work along these lines with particular focus on enabling better patient management of their own information. Patient-centered medical care (in the sense of centered v. driven you all use) is quite distinct from patient-driven medical care. Unfortunately, the patient-driven care has been driven to the margins and the patient-centered care might better be called "profit-centered". Any kind of thinking that enriches the medical subset of relationships in this information age will get my support.
mjh
Michael James Hauan
m...@hauan.org
http://www.hauan.org
573-642-8150 home
573-823-7114 cell