OneBigSoup

On CommunityWiki, we no longer look at InterWiki as the ultimate in networking wiki. Rather, we believe in networking WikiEngines with all SocialSoftware so that wiki, with its inherent advantages for managing semi-stable collective text, replaces less effective content management paradigms.

When all communications software (see SocialSoftware) are integrated, we call the resulting mix: “OneBigSoup.”

Components of the OneBigSoup

There are many ways of envisioning how the network of software will be arranged and work. Current development is piecemeal and resembles biological systems in that cooperation between types of software is incidental and site-specific. Some ways to understand the components:

Here are some things we may see:

Notes on each:

Personal server

A personal server is like a PersonalLogServer, with additional information about you. It stores information attached to your person. So for example, it has your name, your email address, your public PGP key, your web address, your blog address, links to your friends, links to various resources about you. It has a PersonalLogServer built into it, so, whenever you post somewhere, that somewhere can automatically leave a note on your personal log server. When you talk by IM, or IRC, or whatever, and you’re not in stealth mode, then it leaves little notes about your online activities on your PersonalServer?. If someone comments on something you wrote, you get little notices in your personal log. It’s all there.

So that’s “you”- or more appropriately, your online persona.

Imagine that we have a big server somewhere, holding 100’s of these person information stores in them. That’s a “personal server.” It stores people’s bodies, you could say, if you are very confused about bodies versus addresses.

When you log into some forum, you just point to your personal server, and you say, “Hey! That’s me!” And, it can believe you, it can ask you for a password to verify, it can ask you “please encrypt this by your key,” and then verify it, whatever. Depending on how much they want to know that it’s you. Whatever.

It may be more effective to think of this as a vehicle rather than a person. Like a car with a license plate, it is associated with you and your intentions and purposes and you are generally responsible for what happens to it. But it can be stolen like a vehicle, sometimes may be lent like a vehicle with appropriate protections, and when a problem with it occurs it’s generally up to you to track down the cause and resolve it.

In other words, it’s safe for others to assume they’re dealing with the bodily you in that vehicle, and it’s your own responsibility if you hand the wheel over to some bot.

Local name servers

This is where we store the convenient names for things that we use in abbreviated forms, similar to nicknames (nickname server may be a much better name since “local” implies geo-locality). A brief treatment is at IntComm:LocalNames:

A Local Name Server maps a bunch of short names like WikiNames to full URIs, so one can type DannysBlog? instead of http://dannyayers.com/. T his saves typing time and reading time, allowing one to build a PatternLanguage of good pages, and if any URIs change the short name can stay the same. Attached to a Wiki, it acts rather like redirecting a bunch of Wiki pages to external sites.

Whereas a personal server isolates your personal online data & activities from wiki (and other online communities), the local name server isolates you and your groups’ links and language. The danger of this isolation is that you may overload a name very differently than someone else and make it impossible to integrate your own pages with theirs.

Short names however make it easy to refer to things, beyond just wiki pages. You can now talk about blogs, threaded conversations, pictures, static web pages, wiki pages, anything that can be linked to, and you can do it ubiquitously. You are not confined to talking about wiki about things in that wiki (or near wiki, or far wiki- provided you have the InterLink.) You can talk about anything, anywhere, with the convenience of local names.

We’ve isolated not just data about users, but we’ve isolated names and name-link binding of things proximal to individual users that they refer to all the time.

Group servers

For our next trick, we will pull RecentChanges out of the wiki. This is an easy step, because this is just RSS aggregation. But, we’re going to make it a little more interesting.

We’re going to suggest that newcomers to the group can add their own RSS feeds. They just say, “Hi, I’m a newbie. Here’s my RSS feed.” We AssumeGoodFaith, but if they are just here to pester us, or sell us viagra, we can filter them, kick them off, or something like that, if they are still in the probationary period. I don’t know exactly how it will work, but suppose that there will be something. Captcha may be on by default and then be turned off after a user passes several tests and has no spam complaints.

Imagine that RecentChanges is like the saucer section of the Enterprise. You know that it can detach- right? In Star Trek: The Next Generation? Yes: The saucer section can detach, and become it’s own thing. That’s what we’re doing with RecentChanges here. Wow this paragraph is so nerdy it must stay in forever despite being useless. It stands as a shining beacon of nerdiness to all nerds who come after forevermore. It should never be deleted.

Imagine your Wiki’s RecentChanges, but it’s now detached, and that the group of people who were interacting on the wiki, and bonding around RecentChanges, are now floating away in RecentChanges. Now they are not just observing their wiki, but they are observing other things as well, like the main ship being blown up by Romulans.

We can call this detached and expanded RecentChanges a “Group server.” This is more or less a semantics-free name. It has absolutely no meaning whatsoever. Nope, none. It’s one abstraction tied to another with neither of them bound even to warp engines.

People can link in, say, the RSS feed for a blog of theirs to this group server. They can observe multiple wikis’ RSS feeds. They can suggest links for their peers to see, like WikiFeatures:RaisedChanges.

I don’t know the particulars: Perhaps the group members’ collected aggregated blog RSS summaries all appear in one visible page (like the KuroShin diaries page), and perhaps the sum of other things: What group members are doing, what’s happening to the pages, suggested links to look at, happenings on the groups’ local name servers- perhaps all that happens on a seperate page.

Regardless, this is a SharedAwarenessSystem. But you don’t really “post” on it. Rather, you post on this thing by posting somewhere else.

For example, if you post on a thread somewhere, and you do so operating as a member of this group, then the group will see it on it’s radar. If you post on a page on a wiki that the group’s observing, then it will appear on the groups’ radar. If you post a blog entry on a blog that the group monitors, it’ll appear on the groups’ collected blog observing page. Facebook-style or Twitter-style status reports could also appear, just one-line “what I am doing now” reports. Actions like “Mattis is writing a blog entry titled Star Trek Ate My Life” could be added into this log automatically.

You could make a group server without “detaching RecentChanges” using UnifiedClusters. It’ll take any RSS feed and pretend it’s part of a cluster. This seems to do what you’re describing, and isn’t fuzzy :)

::Oh wow another abstraction word tied to another abstraction word saying exactly nothing. How about some more bound terminology like associations, collaborations, teams or cliques? Save the ultra-abstract language for distributed computing papers.

Document, Thread, and Blog Servers

Now we’ve stripped wiki of it’s users, with the user server. We’ve stripped wiki of RecentChanges, with the group server. We’ve stripped wiki even of page names, having the local name server. And we’ve stripped us all of our dignity with the Star Trek saucer analogy. What’s left?

Wiki is “merely a document server.” It’s just a thing that stores and lets you edit documents. It probably just numbers them. You never notice, because you’re using local names the whole time.

Let’s consider how this works. You are on the Internet.

You are you: you have your personal server that you’re tied with.

You assume the role of a group member: You put on your “group suit,” and possibly several at once.

You go to one of the pages by typing in a local name in your web browser. You say, “OneBigSoup,” and Mozilla (or whatever) looks it up in it’s cache of your local name server, and says, “Ah, yes, http://www.emacswiki.org/cgi-bin/community?00003591,” and takes you here. You don’t know that this happens to be #3591, and you don’t care. As far as you are concerned, this is just “One Big Soup.”

You edit the page, and the document server, recognizing both your personal names for things, and your groups names for thigns, binds all links correctly.

You hit save, and it sends off notices to your personal logs, to your group log (which is observing the page,) and to the logs of other observers.

Now: I’m not saying that this is how it’s going to be. Very likely, it will be different, in major ways.

However, I’m saying: Things will work something like this. (I’ll present evidence for that in a moment.)

But now- see, we’re not constrained to just wiki.

Threads. You can have dedicated thread servers. If you are on your group website, and you say, “Make a new thread,” it’ll ask you for the local name of the thread you are creating, and then it’ll go to the threaded discussion server, and it’ll “check out a thread.” Suppose it gets threaded discussion #5349. It binds the name you gave for that discussion to the URL for discussion #5349, and away you go. You now have a threaded discussion that you can link to from just about anywhere.

Or pictures. If you upload a picture resource, you go through a similar process.

Or blogs. If you want to make a new blog, you can go to the blog server, manually, and ask for a blog # (“You have been issued blog#4354.”) or the machine can do it automatically- you go to your group page, and say, “I’d like to check out a blog, in my name, for presentation on this group site, that’ll be called ‘Lion’s Development Blog’” and the group server will fill out all the necessary paperwork for you.

Or whatever. Stuff we haven’t even thought of yet. “Code servers.” Whatever. Anything you want.

It all keys into the “groups,” “individuals,” and “names” framework, and everything can talk to everything else.

Hence the term, “One Big Soup.”

The OneBigSoup Project

LionKimbro used the term “OneBigSoup” (contributed by MattisManzel) to start talking about this idea here, and he started a project (“the OneBigSoup” project) to work towards this ideal. The first major project undertaken was LocalNames, which mixes InterWiki ideas with the larger software world.

Further information is available at: OneBigSoup Wiki

That said: There are hoards of projects working towards the general OneBigSoup ideal. The OneBigSoup project is just specially mentioned here, because it is has close social ties with CommunityWiki.

Evidence

Why do I think we’re going this way?

Because:

People leave KuroShin diaries to go start their own blogs, and their own RSS feeds. People are groping for technologies to make threaded conversations across the internet easy. The idea of “portable threaded conversations” is floating around.

These technologies all start as experiments on centralized servers, but they always end up floating away. That way, people can have independence, and they can do things their own way, and they can experiment with new technologies, and they can develop new things, without being tied to a particular server and a particular guy and a particular community.

Because:

Consider that we already feel the pressure to implement BuiltinThreading? in wiki. I think we all know it’s more or less a good idea; We just are weary of FeatureKarma. And right we should be. What do we do as developers when we feel we are putting in too much technology in one place? Well, we decouple things. Which is exactly what I’m talking about: Massive decoupling.

Having thought of the model in this way, it became clear to me: Wiki “should” also support blogs. Not just pages that we keep our blogs on, but- bona fida blogs. That you can just type away into. That have summaries, which are then aggregated into a central “Here’s the CommunityWiki blog aggregation,” collecting all our community’s CommunityWiki blogs together on.

“But we can already do this on CommunityWiki,” I hear you cry. Yes, we can. But, no, it’s not good enough, dammit. This is the same as BuiltinThreading?. We don’t want to have to edit a whole page, just to add an entry. We don’t want cheesy techniques and hacks. We don’t want to type [[new?]]. We want to just hit “Add Blog Entry,” type in a summary in the summary box, type in a full entry in the full entry box, and have that go to a special “this is blogs” section, and have that be it. I know that we can simulate all this by using pure wiki pages, but just because we can, doesn’t mean that that’s the best, and that that’s what we want to do.

We can simulate full-featured K5 or Slashdot style threading. But that’s not what we really want. (Then again, we don’t want to be constrained by K5’s and Slashdot’s threading either- we’d like to be able to erase threads, too. In most cases.)

So my points are:

The logical conclusion of these two forces is that: We’ll put more into wiki. And then it will all explode outwards. The architecture I have described as “One Big Soup” is an exploded architecture.

Why This is Cool

Once things have exploded out, they become much more… Open. Spacious. And even “Soupy.”

When you’re locked into LiveJournal, you’re locked into LiveJournal. You can only play link games with your LiveJournal friends.

When you’re locked into KuroShin, you’re locked into KuroShin. You can’t publicize your non-existant RSS feed for your K5 diary. You can’t aggregate non-K5 people into your K5 watch list. You're ''stuck'' with K5 people.

But once things explode out, there’s a faaaar vaster community, with much more going on in it, and much more choice, and much more diversity, and much more experimentation, than the GatedCommunity could ever provide.

See Also:

Discussion

Reworked the page only a little. Lots of it still needs reworking, but this is what I’m willing to do now.

Looking for wiki pages on this wiki or other wiki about the phenomenon of continuous disintegration. This is where some piece of software initially appears all integrated in a tight package, and then, over time, becomes atomized: Individual reusable software components that are no longer a central focus.

That way, I could comfortably huck a big long section in the middle of the page.

continuous disintegration.: Interesting choice of words; wouldn’t “decentralization” or “component-ization” be more canonical?

Component-ization is good. :)

blog post I had lost and now have found: http://ext337.blogspot.com/2004_09_01_ext337_archive.html and also link to a guy who’s using his blog to aggregate various parts of his life: http://www.willpate.org/ - people are clamoring for pieces of this one big soup

Dunno where to write, couldn’t really follow anymore, sry, local trouble, blah. But I just sat here with MarkoStevelic and it was one big soup what came to my mind. Just the term. One big soup. In the middleages (in the best case) someone took a horse and rode over to the next monastery and carried a roll of paper with a good idea written on it when there has been one in the first monastery {:)}. What we do nowadays is a wiki-node. And every monk can read it.
For making this electronic horse-ride effective - it takes us whereever in the time of setting a link - we need to make our tool compatible to multilinguality, as in most parts of the world where we can connect to in the time of setting a link other languages are spoken and our words can not be understood. But there are people who understand and can translate. This must be featured, as it connects. Learning languages must be featured. Having several, choosable languages on one page must be done easy. Editing such pages them must be done easy. There sind so etliche Wellen die gegen den Bug unseres Schiffchen laufen während der Halse, eine ist das vielsprachige Experiment, eine nur. Ich Affe krieg kaum mit von meimen Krähennest aus, wie Captain Alex - dessen Hut ich mir übrigens auf dem Kopf umdrehe - wild am Ruder dreht. Was ich mitkrieg’ ist die Bewegung des Schiffes, dessen allmähliches kränken, wies sich neigt. I can see far as I’m high up, communication is hard, a matter of plain talk, actually. TemporyVideoDeleteWhenSeen?

http://spectaclar.org/ – here’s something that’s working sympathetic :) – author would probably be interested in PICA, as well.

There is an Identity Mashups wiki that is looking interesting.

I have met Jon Ramer (been to his house a few times, actually,) and he’s an interesting person. We are connected with several other agents connected with the site, through the EvolutionarySalon, through the RecentChangesCamp, and so on.

If possible, I would want to delegate conversation on identity in that direction, since they are working very seriously on it, and doing good work tracking (and hosting!) the larger conversation.

That’s not to say we shouldn’t talk about it here; It is to say that we should keep our eye on what they’re doing, and think about working there if we’re working seriously on identity.

It’s too bad the wiki doesn’t have a blog. If it did, it’d be easier for us to casually track it, no? BlogControlledByWiki, and all. I really really really want to promote this idea.

I’m trying to put the one big soup on the cooker on [[kabo-wiki-hive_-_whoRthey-wiki_-_one_big_soup?]].

identi.ca and twitter #OneBigSoup

http://identi.ca/tag/onebigsoup/rss returned no data, or LWP::UserAgent is not available.

http://search.twitter.com/search.rss?q=OneBigSoup returned no data, or LWP::UserAgent is not available.

Feel free to consider the free Cmaps client (and free Cmaps server) to integrate into the OneBigSoup. What about leaving a link on my sandbox or Nathaniels sandbox to an arbitrary (friendly and constructive) Html page or even another Cmap, e.g. your sandbox. http://cmaps.cmappers.net/rid=1G5KYQS1R-236KJHZ-TP/Sandbox.cmap

Define external redirect: TemporyVideoDeleteWhenSeen DannysBlog kabo-wiki-hive - whoRthey-wiki - one big soup new BuiltinThreading PersonalServer

EditNearLinks: GatedCommunity FeatureKarma KuroShin WikiNames WikiEngines LiveJournal

Languages: