Fluidinfo, Wikipedia For Databases

Slags writes "The idea behind Fluidinfo is that read-only information is just not as useful on the Web as openly writable information. Metadata is used routinely in the real world from name tags to post-it notes but it is much harder to apply metadata to information on the Internet. That is where Fluidinfo comes along. When information needs to be stored about an object the Fluidinfo database is queried. If the object exists in Fluidinfo, the information is appended to the object. If the object does not exist then it will be created and stored."

OK, so where's the data?

[maxwell demon]

I didn't find a link where I can actually access any data (not even to read it).

Re:

[telekon]

You have to sign up for API access. It looks liek this is early stages, plus they want to create an architecture for devs to build on, less than a site for browsing. So hopefully we'll see apps using their API soon. I'm going to play with it.

Re:

[terrycojones]

There are a few browsers for Fluidinfo data beginning to pop up. The most accessible is at http://explorer.fluidinfo.com/ [fluidinfo.com] which lets you do almost anything (make objects, add tags and values, query, change tag perms, etc) but its interface isn't crystal clear. You can also get at the data using http://shell-fish.appspot.com/ [appspot.com] which is a browser based shell for interacting with Fluidinfo (type help). You can get visualizations of Fluidinfo objects at http://abouttag.appspot.com/ [appspot.com] You can search Fluidinfo ab

Re:

[maxwell demon]

The first thing I notice: It's incredibly slow. To be of practical use, it must speed up at least by the factor 100 (probably more; I've still no result for my first query, and that was one of the example queries shown by the help command!).

Re:

[terrycojones]

Hi. You might want to try that again (btw, what was the query?). You have to keep in mind that we just got slashdotted :-) Sorry!

Re:

[maxwell demon]

Hi. You might want to try that again (btw, what was the query?). You have to keep in mind that we just got slashdotted :-) Sorry!

On shell-fish: show -q 'about = "Paris"' visited rating

It's a sample query from the help command.

Re:

[q0]

That's embarrassing: the documentation for fish was wrong.

The command should be:

show -q 'fluiddb/about = "Paris"' visited rating

If you try that, it should work in a reasonable time.

It's all my fault...I'll go and fix the documentation.

Re:

[maxwell demon]

Indeed, with that it works quick.
However, I'd also consider hanging on an incorrect query a bug.

Re:

[q0]

And, of course, you're right. It seems to fail a bit more gracefully now, if you try the original ("bad") query again.

I don't think it was Fluidinfo that was misbehaving: as far as I can see it was giving a 400 error, which is OK. I think shell-fish's AJAX magic got confused, which is bad (and my fault).

Apologies again. I'll try to figure out what made it hang and fix it.

Re:

[ntoll]

The first thing I notice: It's incredibly slow. To be of practical use, it must speed up at least by the factor 100 (probably more; I've still no result for my first query, and that was one of the example queries shown by the help command!).

Yup, we're a start-up losing our Slashdot virginity... ;-)

deletionist

[vlm]

just like Wikipedia has a web page for everything

So they're taking a bold anti-deletionist position. Good for them.

So what is it?

[sakdoctor]

Fluidinfo is an online information storage and search platform.

So what is it?

Fluidinfo provides a universal metadata engine because it has an object for everything imaginable, just like Wikipedia has a web page for everything.

So what is it?

Fluidinfo makes it possible for data to be social. It allows almost unlimited information personalization by individual users and applications, and also between them.

So what is it?

Re:So what is it?

[oscartheduck]

It's a slashvertisement; the page linked to in the article is just the front page for the product. No news, no editorial, no review, no discussion (as you pointed out) of what it is. Nothing.

Re:

[malakai]

Fluidinfo named as Tim O’Reilly’s favorite startup
Tuesday, March 15th, 2011

I've lost track since '00, does Tim O'Reilly have a hand in Geeknet/slash now?

Re:So what is it?

[Lord Grey]

Fluidinfo is a database of metadata. But since metadata is really just data, Fluidinfo is really a database of data. Which is to say, it's a database. But there's a twist. You can make new "objects" at will. Kind of like most other databases, actually. But with even more of a twist, anyone can do that! Like what happens when you forget to secure your firewall. Then the excitement starts: You can add arbitrary key/value data -- metadata! -- to the object! Like a JOIN with another SQL table but with different semantics. But since the actual usage of the key/value pairs is not governed, you will have to collaborate with other users and applications through some external channel. The shared keys could be coordinated in an external database, for example.

Sarcasm aside, I'm sure this project is really cool and stuff, but the cynic in me thinks otherwise.

Re:

[malakai]

The Gist: http://blogs.fluidinfo.com/fluidinfo/2011/02/23/putting-domain-names-onto-data-with-fluidinfo/ [fluidinfo.com]
Example Object: http://blogs.fluidinfo.com/fluidinfo/wp-content/uploads/2011/02/book-object.png [fluidinfo.com]

My 15s appraisal:
They want to be the single-source OO database for 'everything'. Take all the data in wiki or any webpage ( assuming it's about an entity), extract any quantitative properties, ( Size, color, temperature, weight, Atomic Number... etc) and add them to Fluidinfo. Incorporate a way for domain names

Re:

[maxwell demon]

So it's basically semantic web, but concentrated in a single server?

Re:

[aix tom]

So because Web 2.0 has blown all it's buzzword potential they decided to give Usenet 2.0 a shot?

Re:

[terrycojones]

In a way, but it's much simpler than the semantic web stuff. It's more like post-it notes - walk up to anything and put a note containing anything onto any object you like. I don't think it should be a requirement that you have a PhD in CS to understand a data model (and in the case of the semantic web with its jungle of acronyms that's sometimes not enough). Fluidinfo has an object for everything, in a very simple way, and those things don't have to be URLs - they can be email addresses, zipcodes, DNA seq

Re:

[terrycojones]

The problem with your model of "I expect the same info to simply be published by data owners. Or, simply extracted by an app for us running in a Google data center." is that many people (and apps) don't have a place to put things. And if they do, they put it all separately and it's less valuable (you can't search across it easily). The problem with having Google extract the metadata (as with microformats) is that then only the owners can add to it. Fluidinfo lets you put your data alongside other related i

Re:

[no_opinion]

My interpretation is that it's freebase.com but with a different (easier?) API, loose (or absent) semantics, and no starting data.

Re:

[terrycojones]

There are some pretty big differences from Freebase. All their data was readable. There were no perms and therefore no way to build apps with private or shared data. Freebase was also very heavily into ontology (which I guess is what you're getting at with your reference to semantics), whereas Fluidinfo is heavily into evolution (of representation, convention, reputation & trust). There's actually quite a bit of data in Fluidinfo, but we've not done a good job of making that obvious yet. Hop onto #flui

Re:

[Infiniti2000]

Sarcasm aside, I'm sure this project is really cool and stuff, but the cynic in me thinks otherwise.

Same here. Why do I get the feeling that my name, address, and birth-date will be stored in this DB and all spammers will have easy and immediate access to it? More importantly, what's to keep them from doing that?

Re:

[terrycojones]

Hi. I know, it's easy to be a cynic - I'm great at it :-) I'm happy to help you understand if you like. There is a perms system on tags, so you don't have to collaborate with others over the data you add to objects. Your tags are in your namespace, like lordgrey/rating and you control them. And sure, it's a database (in fact it used to be called FluidDB). It's also a Turing Machine :-) The main point it that normally when people or apps have information about something (a URL, an email address, a zipcode

Re:

[WrongSizeGlass]

So what is it?

It's a database about data. Who's data? What data? What kind of data about data? Who knows.

Re:

[jamiesan]

Who's data?

I don't know.

THIRD BASE!

Re:

[terrycojones]

It's any data - yours or anyone who cares to add it. You can add tags to objects, and the tag values can be null, Boolean, numbers, strings, sets of strings. But they can also be anything you like: PDF, binaries, audio, etc., each with a MIME type (and you can make up your own MIME type if you like). So you can store HTML, CSS, JS tags onto objects in Fluidinfo and get at them through your browser. Or anything else you like.

Re:

[tenco]

Sounds like a *SQL server listening on 0.0.0.0

Re:

[gl4ss]

from the blurb it actually sounds like it's another spin on so called "context web" or "flock".

looking at the site, it seems it's that idea expanded to random strings that can have a random number of strings as attributes which can.. and then calling those object.

tags mentioned, of course. there's also some api, which I guess is the main entree here, actually. but I still fail to see the advantage and practicality and freshness of this... an article which would have done some performance testing etc etc on

Re:

[Slags]

I see what happened. The link to the article was cut out. See below for the explanation. There's a slideshow to give you a visual representation. http://matthewthetech.blogspot.com/2011/08/fluidinfocom-is-like-wikipedia-for.html [blogspot.com]

Re:

[migla]

It's a rent in the space-time continuum.

Re:

[terrycojones]

Hi sakdoctor Sorry, you're right that it's hard to figure out what Fluidinfo is - and it's hard to make a website that is understandable to a wide range of people too. One summary is to say that Fluidinfo is like Wikipedia, but for applications and their data rather than for humans. Like Wikipedia has a page (URL) for everything, Fluidinfo has an object for everything - lazily created, of course. The differences from a wiki that you need to make a Wikipedia for data are: 1) a permissions system so apps c

Re:

[aix tom]

The Cat: [to Rimmer] What *is* it?
Rimmer: It's a rent in the space-time continuum.
The Cat: [to Lister] What *is* it?
Lister: The stasis room freezes time, you know, makes time stand still. So whenever you have a leak, it must preserve whatever it's leaked into, and it's leaked into this room.
The Cat: [to Rimmer] What *is* it?
Rimmer: It's singularity, a point in the Universe where the normal laws of space and time don't apply.
The Cat: [to Lister] What *is* it?
Lister: It's a hole back into the past.
The Cat: Oh

Re:

[whiteboy86]

a VC trap

Re:

[terrycojones]

LOL! I wish :-)

Re:

[ntoll]

Fluidinfo in a nutshell: http://www.slideshare.net/fluidinfo/fluidinfo-in-a-nutshell [slideshare.net] Happy to answer any questions (n.b. yeah, I work for Fluidinfo and I authored the presentation that I'm linking to).

Re:

[dchamp]

It's a big swirly thing in space.

Re:

[murakdar]

Welcome to fluidinfo. This is fluidinfo. Welcome. You can do anything at fluidinfo. Anything at all. The only limit is yourself. This is fluidinfo, welcome. The infinite is possible at fluidinfo. The unattainable is unknown at fluidinfo. This is fluidinfo. Welcome to fluidinfo. The OP reminded me of zombo.com. Where anything is possible.

Can anyone figure this out?

[vlm]

Can anyone express what this does in technical terms? Everything I could find was sorta like liberal arts major expounds on the new (to the liberal arts establishment) idea of memoization.

http://en.wikipedia.org/wiki/Memoization [wikipedia.org]

Their blog has a good example

[brokeninside]

A thought experiment how their service might be used to automatically confirm or reject friendship requests: http://blogs.fluidinfo.com/fluidinfo/2011/06/01/personalized-filtering-of-friend-requests-in-social-networks/ [fluidinfo.com]

If I understand what they want to do, I think it's a failure. They make a big deal about metadata being context dependent and, as such, it should stored in the context in which it is meaningful rather than in a single place. But, if I understand what they do correctly, their service is basical

Re:

[magarity]

Can anyone express what this does in technical terms?

Well, from this part:

If the object exists in Fluidinfo, the information is appended to the object. If the object does not exist then it will be created and stored.

It sounds like they've invented the MERGE command.

Re:

[terrycojones]

Yes, that's a good example use case. Then just apply the same thinking to anything (URLs, stocks, email addresses, names, zip codes, etc), and include the possibility that normal people (via apps) might want to store their data onto the same objects - I own it, I want it, here's my rating, whatever.

Re:

[terrycojones]

Hi. Love the memoization comment :-) I don't know how technical you want it, but there are some slide decks on http://www.slideshare.net/fluidinfo/presentations [slideshare.net] and some more technical ones in http://www.slideshare.net/terrycojones/presentations [slideshare.net] (back when we were still calling it FluidDB). If you have specific questions I'll try to answer them. Fluidinfo doesn't provide the actual storage layer (we build on a variety of things), it's more that it's an alternate interface to information - a bit like Wiki

WikiDB

[ihop0]

I've had some ideas for wiki-like databases for awhile now, if this were open people could create some really cool open information sets

Re:

[Hatta]

Of course, you'd run into the same problems Wikipedia has. How do you curate data from diverse sources and ensure integrity?

Re:

[HarrySquatter]

Duh. Build up a huge hierarchy of preferred "editors" of the database and have them camp out the bits of metadata that they wrote. Then when it gets successful have those same people start marking en masse lots of the metadata to be deleted because it's not notable.

Re:

[jedidiah]

The less contentious the subject is, the less likely that this will be a problem.

Even then, you can provide metrics that allow the individual to judge the information based on it's churn.

The fact that some data will be crap is not a good reason to be afraid of collecting data in general.

Re:

[ihop0]

Well if you're creating a db of measurements or values, like dimensions of various objects, or automobile specifications, there's much less bickering to be done. If the nature of the DB isn't open ended && subjective, you avoid lots of landmines.

Re:

[gridaphobe]

If the nature of the DB isn't open ended && subjective, you avoid lots of landmines.

You can also avoid the landmines if you have a mechanism for attributing content to specific users as Fluidinfo does via namespaces. There can only be one Wikipedia page for Transformers 3, so there might be a lot of contention about whether the article was fair/neutral, but in Fluidinfo you, RottenTomatoes, IMDB, and anyone else can place a rating or review tag on the object about Transformers 3, without any conflicts.

Re:

[Tablizer]

Over at C2.com, the very first web wiki, we have kicked around various ideas for things that are kind of catch-all dumping grounds for semi-structured info: part wiki, database, file-system, note-pad, content manager, and CRUD-stack. It wouldn't necessarily do any of these well (up front at least), but integrate various aspects of them all.

It could be useful for projects where you are not sure what you want yet and want to grow in an organic way.

Freebase

[Jimmy King]

It sounds like Freebase [freebase.com] to me, which has been around for years.

Re:

[ntoll]

It's similar to Freebase *but* rather than curating the organisation of data (as Freebase does) Fluidinfo is, er, "fluid" in that it expects conventions in tagging and organising data to emerge. Evolutionary pressure will make the best / most appropriate "schema" to survive (become conventions) in much the same way that hashtags is a bottom-up convention that emerged in Twitter.

Re:

[Jimmy King]

Hm, I see. Well, I'll definitely keep an eye on it. I do like the general concept of these sorts of things. There's a lot of good uses for them and I've used freebase for a project or two in the past.

Re:

[Jimmy King]

No idea. The biggest thing I used it for was pulling data on video game and game console release dates for a demo website. That was more due to the amount of data and convenience than anything else and pretty simple.

Actual Wikipedia for Databases

[Downchuck]

The OmegaWiki project provides a multilingual database, is based on MediaWiki and was authored in large part by a Wikimedia staff-member. It's an interesting re-imagining of Wiktionary.

Wikipedia already is a Database

[popo]

Wikipedia *is* a database. This is like proposing a search engine for "data"....oooh! Sounds amazing!

100% vaporware. But please don't tell the VC guys who got bamboozled. They're just figuring it out and hoping to pass the buck to a greater fool.

Re:

[gl4ss]

wikipedia is a service that runs some wiki-sw.
this is a tagging or databasing, data linked to data software.

once you put it like that, without the marketing talk, without the blog, without their non functioning demo.. yeah it seems a bit boring.

Re:

[terrycojones]

Wikipedia is built on a database. It provides an alternate interface to information, allowing anyone to contribute. Ten years ago that sounded pretty ridiculous, I think to just about everyone. What if you had the same thing, but for applications and their data (or metadata, if you like). Is that also ridiculous? The idea of openly writable storage (with a permissions system, typed data, and a query language - as in Fluidinfo) isn't as bad as you're making out. Re vaporware - actually not. I've spent about

Hello old friend?

[recharged95]

Reading the about and blog sections, sure sounds like the '11 version of the Semantic Web [wikipedia.org].

Flawed reasoning

[Kazymyr]

Quote from the discussion:

"The justification is simple. We're removing the Firefox version number
from all of the common user-visible locations because we don't believe
that users need to know what version they're on. We're moving to a model
that's more like the Web. What version of Gmail are you on?

We've removed it from all of our marketing materials. We're removing it
from the download button on the Website. We're removing it from how we
talk to users about Firefox. We're ending version numbers because
they're

How to think about Fluidinfo

[terrycojones]

Probably the simplest way to think about Fluidinfo is by analogy to Wikipedia. Wikipedia is not a storage system in & of itself. It sits on top of some form of storage and provides an alternate interface to information for normal people. Its most distinguishing characteristics are 1) that anyone can add information to any page, and 2) that there is a page for everything (you get to create it if it doesn't already exist). Ten years ago if you'd said you were going to build an encyclopedia by letting an

Re:

[terrycojones]

You're exaggerating wildly re "respond to EACH and EVERY post" and "respond to absolutely everyone individually". Have you counted? You're of course right that people are saying WTF, and with good reason. Fluidinfo is a bit abstract and our web site sucks. People are asking questions here and I'm trying to help. It's not about control, it's about trying to answer the questions of people who are trying to understand. If people were asking questions about a project you were involved in, what would you do? Y